novdocx (en) 13 May 2009

AUTHORIZED DOCUMENTATION

Storage Administration Guide

Novell ®

SUSE Linux Enterprise Server

®

10 SP2
May 15, 2009

www.novell.com

SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Legal Notices
Novell, Inc. makes no representations or warranties with respect to the contents or use of this documentation, and
specifically disclaims any express or implied warranties of merchantability or fitness for any particular purpose.
Further, Novell, Inc. reserves the right to revise this publication and to make changes to its content, at any time,
without obligation to notify any person or entity of such revisions or changes.

Further, Novell, Inc. makes no representations or warranties with respect to any software, and specifically disclaims
any express or implied warranties of merchantability or fitness for any particular purpose. Further, Novell, Inc.
reserves the right to make changes to any and all parts of Novell software, at any time, without any obligation to
notify any person or entity of such changes.

Any products or technical information provided under this Agreement may be subject to U.S. export controls and the
trade laws of other countries. You agree to comply with all export control regulations and to obtain any required
licenses or classification to export, re-export or import deliverables. You agree not to export or re-export to entities on
the current U.S. export exclusion lists or to any embargoed or terrorist countries as specified in the U.S. export laws.
You agree to not use deliverables for prohibited nuclear, missile, or chemical biological weaponry end uses. See the
Novell International Trade Services Web page (http://www.novell.com/info/exports/) for more information on
exporting Novell software. Novell assumes no responsibility for your failure to obtain any necessary export
approvals.

Copyright © 2006-2008 Novell, Inc. All rights reserved. No part of this publication may be reproduced, photocopied,
stored on a retrieval system, or transmitted without the express written consent of the publisher.

Novell, Inc. has intellectual property rights relating to technology embodied in the product that is described in this
document. In particular, and without limitation, these intellectual property rights may include one or more of the U.S.
patents listed on theNovell Legal Patents Web page (http://www.novell.com/company/legal/patents/) and one or more
additional patents or pending patent applications in the U.S. and in other countries.

Novell, Inc.
404 Wyman Street, Suite 500
Waltham, MA 02451
U.S.A.
www.novell.com

Online Documentation: To access the latest online documentation for this and other Novell products, see
the Novell Documentation Web page (http://www.novell.com/documentation).
 novdocx (en) 13 May 2009
Novell Trademarks
For Novell trademarks, see the Novell Trademark and Service Mark list (http://www.novell.com/company/legal/
trademarks/tmlist.html).

Third-Party Materials
All third-party trademarks and copyrights are the property of their respective owners.
Some content in this document is copied, distributed, and/or modified from the following document under the terms
specified in the document’s license.
EVMS User Guide, January 18, 2005
Copyright © 2005 IBM
License Information
This document may be reproduced or distributed in any form without prior permission provided the copyright
notice is retained on all copies. Modified versions of this document may be freely distributed provided that they
are clearly identified as such, and this copyright is included intact.
 novdocx (en) 13 May 2009

4 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Contents

About This Guide 9

1 Overview of EVMS 11
1.1 Benefits of EVMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.2 Plug-In Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.3 Supported File Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.4 Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.5 Location of Device Nodes for EVMS Storage Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2 Using EVMS to Manage Devices 15

2.1 Configuring the System Device at Install to Use EVMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.1.1 Before the Install . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.1.2 During the Server Install . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.1.3 After the Server Install . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.2 Configuring an Existing System Device to Use EVMS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.2.1 Disable the boot.lvm and boot.md Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.2.2 Enable the boot.evms Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.2.3 Edit the /etc/fstab File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.2.4 Edit the Boot Loader File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.2.5 Force the RAM Disk to Recognize the Root Partition . . . . . . . . . . . . . . . . . . . . . . . . 25
2.2.6 Restart the Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.2.7 Verify that EVMS Manages the Boot, Swap, and Root Partitions . . . . . . . . . . . . . . . 26
2.3 Configuring LVM Devices to Use EVMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.4 Using EVMS with iSCSI Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.5 Using the ELILO Loader Files (IA-64) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.6 Starting EVMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.7 Starting the EVMS Management Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

3 Using UUIDs to Mount Devices 31

3.1 Naming Devices with udev . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.2 Understanding UUIDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.2.1 Using UUIDs to Assemble or Activate File System Devices . . . . . . . . . . . . . . . . . . . 32
3.2.2 Finding the UUID for a File System Device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.3 Using UUIDs in the Boot Loader and /etc/fstab File (x86) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.4 Using UUIDs in the Boot Loader and /etc/fstab File (IA64) . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.5 Additional Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

4 Managing EVMS Devices 35

4.1 Understanding Disk Segmentation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.1.1 Segment Managers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.1.2 Disk Segments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
4.2 Initializing Disks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
4.2.1 Before You Begin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
4.2.2 Guidelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.2.3 Adding a Segment Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.3 Removing the Segment Manager from a Device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

Contents 5
 novdocx (en) 13 May 2009
4.4 Creating Disk Segments (or Partitions) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.5 Configuring Mount Options for Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.6 What’s Next . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

5 Managing Multipath I/O for Devices 43

5.1 Understanding Multipathing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.1.1 What Is Multipathing? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.1.2 Benefits of Multipathing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.2 Planning for Multipathing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
5.2.1 Guidelines for Multipathing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
5.2.2 Using Multipathed Devices Directly or in EVMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
5.2.3 Using LVM2 on Multipath Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
5.2.4 Using mdadm with Multipath Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
5.2.5 Partitioning Multipath Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
5.2.6 Supported Architectures for Multipath I/O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
5.2.7 Supported Storage Arrays for Multipathing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
5.3 Multipath Management Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
5.3.1 Device Mapper Multipath Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
5.3.2 Multipath I/O Management Tools. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.3.3 Using mdadm for Multipathed Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.3.4 The Linux multipath(8) Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
5.4 Configuring the System for Multipathing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
5.4.1 Preparing SAN Devices for Multipathing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
5.4.2 Partitioning Multipathed Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
5.4.3 Configuring the Server for Multipathing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
5.4.4 Adding multipathd to the Boot Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
5.4.5 Creating and Configuring the /etc/multipath.conf File . . . . . . . . . . . . . . . . . . . . . . . . 55
5.5 Enabling and Starting Multipath I/O Services. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
5.6 Configuring Path Failover Policies and Priorities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
5.6.1 Configuring the Path Failover Policies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
5.6.2 Configuring Failover Priorities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
5.6.3 Using a Script to Set Path Priorities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
5.6.4 Configuring ALUA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
5.6.5 Reporting Target Path Groups. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.7 Tuning the Failover for Specific Host Bus Adapters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.8 Configuring Multipath I/O for the Root Device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.9 Configuring Multipath I/O for an Existing Software RAID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
5.10 Scanning for New Devices without Rebooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
5.11 Scanning for New Partitioned Devices without Rebooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
5.12 Viewing Multipath I/O Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
5.13 Managing I/O in Error Situations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
5.14 Resolving Stalled I/O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
5.15 Additional Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
5.16 What’s Next . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

6 Managing Software RAIDs with EVMS 75

6.1 Understanding Software RAIDs on Linux. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
6.1.1 What Is a Software RAID? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
6.1.2 Overview of RAID Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
6.1.3 Comparison of RAID Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
6.1.4 Comparison of Disk Fault Tolerance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
6.1.5 Configuration Options for RAIDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
6.1.6 Interoperability Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78

6 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
6.1.7 Guidelines for Component Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
6.1.8 RAID 5 Algorithms for Distributing Stripes and Parity . . . . . . . . . . . . . . . . . . . . . . . . 80
6.1.9 Multi-Disk Plug-In for EVMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
6.1.10 Device Mapper Plug-In for EVMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
6.2 Creating and Configuring a Software RAID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
6.3 Expanding a RAID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
6.3.1 Adding Mirrors to a RAID 1 Device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
6.3.2 Adding Segments to a RAID 4 or 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
6.4 Adding or Removing a Spare Disk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
6.4.1 Do You Need a Spare Disk? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
6.4.2 Adding a Spare Disk When You Create the RAID . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
6.4.3 Adding a Spare Disk to an Existing RAID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
6.4.4 Removing a Spare Disk from a RAID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
6.5 Managing Disk Failure and RAID Recovery. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
6.5.1 Understanding the Disk Failure and RAID Recovery. . . . . . . . . . . . . . . . . . . . . . . . . 87
6.5.2 Identifying the Failed Drive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
6.5.3 Replacing a Failed Device with a Spare . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
6.5.4 Removing the Failed Disk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
6.6 Monitoring Status for a RAID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
6.6.1 Monitoring Status with EVMSGUI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
6.6.2 Monitoring Status with /proc/mdstat. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
6.6.3 Monitoring Status with mdadm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
6.6.4 Monitoring a Remirror or Reconstruction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
6.6.5 Configuring mdadm to Send an E-Mail Alert for RAID Events. . . . . . . . . . . . . . . . . . 93
6.7 Deleting a Software RAID and Its Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

7 Managing Software RAIDs 6 and 10 with mdadm 97

7.1 Creating a RAID 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
7.1.1 Understanding RAID 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
7.1.2 Creating a RAID 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
7.2 Creating Nested RAID 10 Devices with mdadm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
7.2.1 Understanding Nested RAID Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
7.2.2 Creating Nested RAID 10 (1+0) with mdadm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
7.2.3 Creating Nested RAID 10 (0+1) with mdadm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
7.3 Creating a Complex RAID 10 with mdadm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
7.3.1 Understanding the mdadm RAID10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
7.3.2 Creating a RAID10 with mdadm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
7.4 Creating a Degraded RAID Array. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104

8 Resizing Software RAID Arrays with mdadm 107

8.1 Understanding the Resizing Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
8.1.1 Guidelines for Resizing a Software RAID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
8.1.2 Overview of Tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
8.2 Increasing the Size of a Software RAID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
8.2.1 Increasing the Size of Component Partitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
8.2.2 Increasing the Size of the RAID Array . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
8.2.3 Increasing the Size of the File System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
8.3 Decreasing the Size of a Software RAID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
8.3.1 Decreasing the Size of the File System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
8.3.2 Decreasing the Size of Component Partitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
8.3.3 Decreasing the Size of the RAID Array . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

Contents 7
 novdocx (en) 13 May 2009
9 Installing and Managing DRBD Services 117
9.1 Understanding DRBD. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
9.2 Installing DRBD Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
9.3 Configuring the DRBD Service. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
9.4 Testing the DRBD Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
9.5 Troubleshooting DRBD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
9.5.1 Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
9.5.2 Host Names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
9.5.3 TCP Port 7788 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
9.5.4 The --do-what-i-say Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
9.6 Additional Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

10 Troubleshooting Storage Issues 123

10.1 Is DM-MP Available for the Boot Partition? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
10.2 Rescue System Cannot Find Devices That Are Managed by EVMS . . . . . . . . . . . . . . . . . . . 123
10.3 Volumes on EVMS Devices Do Not Appear After Reboot . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
10.4 Volumes on EVMS Devices Do Not Appear When Using iSCSI . . . . . . . . . . . . . . . . . . . . . . 124
10.5 Device Nodes Are Not Automatically Re-Created on Restart . . . . . . . . . . . . . . . . . . . . . . . . 124

A Documentation Updates 125

A.1 May 15, 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
A.1.1 Managing Multipath I/O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
A.2 November 24, 2008 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
A.2.1 Managing Multipath I/O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
A.2.2 Using UUIDs to Mount Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
A.3 June 10, 2008 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
A.3.1 Managing Multipath I/O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
A.4 March 20, 2008 (SLES 10 SP2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
A.4.1 Managing Multipath I/O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

8 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
About This Guide

This guide provides information about how to manage storage devices on a SUSE® Linux
Enterprise Server 10 Support Pack 2 server with an emphasis on using the Enterprise Volume
Management System (EVMS) 2.5.5 or later to manage devices. Related storage administration
issues are also covered as noted below.
Š Chapter 1, “Overview of EVMS,” on page 11
Š Chapter 2, “Using EVMS to Manage Devices,” on page 15
Š Chapter 3, “Using UUIDs to Mount Devices,” on page 31
Š Chapter 4, “Managing EVMS Devices,” on page 35
Š Chapter 5, “Managing Multipath I/O for Devices,” on page 43
Š Chapter 6, “Managing Software RAIDs with EVMS,” on page 75
Š Chapter 7, “Managing Software RAIDs 6 and 10 with mdadm,” on page 97
Š Chapter 8, “Resizing Software RAID Arrays with mdadm,” on page 107
Š Chapter 9, “Installing and Managing DRBD Services,” on page 117
Š Chapter 10, “Troubleshooting Storage Issues,” on page 123

Audience

This guide is intended for system administrators.

Feedback

We want to hear your comments and suggestions about this manual and the other documentation
included with this product. Please use the User Comments feature at the bottom of each page of the
online documentation, or go to www.novell.com/documentation/feedback.html and enter your
comments there.

Documentation Updates
For the most recent version of the SUSE Linux Enterprise Server 10 Storage Administration Guide
for EVMS, visit the Novell Documentation Web site for SUSE Linux Enterprise Server 10 (http://
www.novell.com/documentation/sles10).

Additional Documentation

For information about managing storage with the Linux Volume Manager (LVM), see the SUSE
Linux Enterprise Server 10 Installation and Administration Guide (http://www.novell.com/
documentation/sles10).

For information about iSNS (Internet Storage Name Service), see “iSNS” (http://www.novell.com/
documentation/sles10/sles_admin/index.html?page=/documentation/sles10/sles_admin/data/
cha_isns.html) in the SUSE Linux Enterprise Server 10 Installation and Administration Guide (http:/
/www.novell.com/documentation/sles10).

About This Guide 9

 novdocx (en) 13 May 2009
Documentation Conventions

In Novell documentation, a greater-than symbol (>) is used to separate actions within a step and
items in a cross-reference path.

A trademark symbol (®, TM, etc.) denotes a Novell trademark. An asterisk (*) denotes a third-party
trademark.

10 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Overview of EVMS
1

The Enterprise Volume Management System (EVMS) 2.5.5 management tool for Linux* is an
1
extensible storage management tool that integrates all aspects of volume management, such as disk
partitioning, the Logical Volume Manager (LVM), the Multiple-Disk (MD) manager for software
RAIDs, the Device Mapper (DM) for multipath I/O configuration, and file system operations.
Š Section 1.1, “Benefits of EVMS,” on page 11
Š Section 1.2, “Plug-In Layers,” on page 11
Š Section 1.3, “Supported File Systems,” on page 12
Š Section 1.4, “Terminology,” on page 12
Š Section 1.5, “Location of Device Nodes for EVMS Storage Objects,” on page 13

1.1 Benefits of EVMS

EVMS provides the following benefits:

Š An open source volume manager

Š A plug-in framework for flexible extensibility and customization
Š Plug-ins to extend functionality for new or evolving storage managers
Š Support for foreign partition formats
Š Cluster-aware

1.2 Plug-In Layers

EVMS abstracts the storage objects in functional layers to make storage management more user-
friendly. The following table describes the current EVMS plug-in layers for managing storage
devices and file systems:

Table 1-1 EVMS Plug-In Layers

Storage Managers Description Plug-Ins

Device Manages the physical and logical Device Mapper (DM)

devices

Segment Manages the partitioning of Uses Device Mapper (DM)

physical and logical devices into
smaller segments of free space. Segment managers include DOS,
GPT, System/390* (S/390),
Segment managers can be Cluster, BSD, Mac, and BBR
stacked. For example, a cluster
segment can contain other For more information, see
storage objects or volumes. Section 4.1, “Understanding Disk
Segmentation,” on page 35.

Overview of EVMS 11
 novdocx (en) 13 May 2009
Storage Managers Description Plug-Ins

Regions Manages the combination of LVM/LVM2 for containers and

multiple storage objects region, MD for RAIDs, and DM for
multipath I/O

EVMS Features Manages EVMS features Drive linking (linear

concatenation), Bad Block
Relocation (BBR), and Snapshot

File System Interface Modules Manages the interface between For information, see Section 1.3,
(FSIM) the file system managers and the “Supported File Systems,” on
segment managers page 12.

Cluster Manager Interface Manages the interface between HeartBeat 2

Modules the cluster manager and the file
systems and devices

1.3 Supported File Systems

EVMS supports the following Linux file systems:

Š EXT3
Š ReiserFS
Š XFS
Š OCFS2
Š JFS
Š EXT2
Š Swap
Š NTFS (read only)
Š FAT (read only)

For more information about file systems supported in SUSE® Linux Enterprise Server 10, see the
SUSE Linux Enterprise Server 10 Installation and Administration Guide. (http://www.novell.com/
documentation/sles10).

The File System Primer (http://wiki.novell.com/index.php/File_System_Primer) describes the

variety of file systems available on Linux and which ones are the best to use for which workloads
and data.

1.4 Terminology
EVMS uses the following terminology in the EVMS user interface:

Table 1-2 EVMS Terms

Term Description

Sector The lowest level that can be addressed on a block device.

Disk A physical disk or a logical device.

12 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Term Description

Segment An ordered set of physically contiguous sectors on a single device. It is similar

to traditional disk partitions.

Region An ordered set of logically contiguous sectors that might or might not be
physically contiguous. The underlying mapping can be to logical disks, disk
segments, or other storage regions.

Feature A logically contiguous address space created from one or more disks,
segments, regions, or other feature objects through the use of an EVMS
(Feature Object, EVMS feature.
Feature, EVMS Object)

Storage Object Any storage structure in EVMS that is capable of being a block device. Disks,
segments, regions, and feature objects are all storage objects.

Container A collection of devices that is managed as a single pool of storage.

(Storage Container) Private Storage Container: A storage container that is exclusively owned and
accessed by only one server.

Cluster Storage Container: A storage container managed by the Cluster

Resource Manager. It is accessible to all nodes of a cluster. An administrator
can configure the storage objects in the cluster container from any node in the
cluster. Cluster containers can be private, shared, or deported.

Š Private: The cluster container is exclusively owned and accessed by only

one particular node of a cluster at any given time. The ownership can be
reassigned by failover policies or the administrator.
Š Shared: The cluster container is concurrently owned and accessed by all
nodes of a cluster. Shared containers are preferred for distributed
databases, clustered file systems, and cluster-aware applications that can
coordinate safe access to shared volumes.
Š Deported: The cluster container is not owned or accessed by any node of
the cluster.

Volume A mountable storage object. Logical volumes can be EVMS volumes or

compatibility volumes.
(Logical Volume)
Š EVMS Volume: Volumes that contain EVMS metadata and support all
EVMS features. Device nodes for EVMS volumes are stored in the /dev/
evms directory. For example: /dev/evms/my_volume
Š Compatibility Volume: Volumes that are backward-compatible to other
volume managers. They do not contain EVMS metadata and cannot
support EVMS features.

1.5 Location of Device Nodes for EVMS Storage

Objects
EVMS creates a unified namespace for the logical volumes on your system in the /dev/evms
directory. It detects the storage objects actually present on a system, and creates an appropriate
device node for each one, such as those shown in the following table.

Overview of EVMS 13
 novdocx (en) 13 May 2009
Table 1-3 Device Node Location

Storage Object Standard Location the Device Node EVMS Location of the Device Node

A disk segment of disk /dev/sda5 /dev/evms/sda5

A software RAID device /dev/md1 /dev/evms/md/md1

An LVM volume /dev/lvm_group/lvm_volume /dev/evms/lvm/lvm_group/

lvm_volume

14 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Using EVMS to Manage Devices
2

This section describes how to configure EVMS as the volume manager of your devices.
2
Š Section 2.1, “Configuring the System Device at Install to Use EVMS,” on page 15
Š Section 2.2, “Configuring an Existing System Device to Use EVMS,” on page 22
Š Section 2.3, “Configuring LVM Devices to Use EVMS,” on page 27
Š Section 2.4, “Using EVMS with iSCSI Volumes,” on page 27
Š Section 2.5, “Using the ELILO Loader Files (IA-64),” on page 28
Š Section 2.6, “Starting EVMS,” on page 28
Š Section 2.7, “Starting the EVMS Management Tools,” on page 28

2.1 Configuring the System Device at Install to

Use EVMS
This section describes how to configure the system device during the Linux install to use EVMS as
the volume manager instead of the current default of Linux Volume Manager (LVM).
Š Section 2.1.1, “Before the Install,” on page 15
Š Section 2.1.2, “During the Server Install,” on page 17
Š Section 2.1.3, “After the Server Install,” on page 20

2.1.1 Before the Install

Š “System Device” on page 15
Š “Device Size Limits” on page 16
Š “Data Loss Considerations for the System Device” on page 16
Š “Storage Deployment Considerations for the System Device” on page 16

System Device
For the purposes of this install documentation, a system device is any device that contains the Linux
/boot, swap, or root (/) partitions for your Linux computer.

The install instructions assume the following:

Š All three system partitions are on the same physical disk.

If you use different disks for any of the system partitions, make sure to modify the install
instructions for your deployment scenario so that all of the system partitions are managed by
EVMS.
Š You must configure the boot partition within the BIOS-addressable space (such as 2 GB for x86
or 8 GB for x86-64) of the first disk recognized by the system.
If this restriction is not required for your hardware, you can modify the location of the /boot
partition as desired.

Using EVMS to Manage Devices 15

 novdocx (en) 13 May 2009
Š Your system uses the Grub or LILO boot loaders.
If you have an IA64 system, you must modify these install instructions to use the ELILO boot
loader (/boot/efi/elilo.conf) instead.

WARNING: Whenever you manually alter the kernel or initrd on your system, make sure to
run /sbin/elilo before shutting down the computer. If you leave out this step, your system
might not be bootable.

Device Size Limits

Version 2.3 and later of mdadm supports component devices up to 4 TB in size each. Earlier versions
support component devices up to 2 TB in size.

IMPORTANT: If you have a local disk, external disk arrays, or SAN devices that are larger than the
supported device size, use a third-party disk partitioner to carve the devices into smaller logical
devices.

You can combine up to 28 component devices to create the RAID array. The md RAID device you
create can be up to the maximum device size supported by the file system you plan to use. For
information about file system limits for SUSE® Linux Enterprise Server 10, see “Large File System
Support” in the SUSE Linux Enterprise Server 10 Installation and Administration Guide. (http://
www.novell.com/documentation/sles10).

Data Loss Considerations for the System Device

This install requires that you delete the default partitioning settings created by the install, and create
new partitions to use EVMS instead. This destroys all data on the disk.

WARNING: To avoid data loss, it is best to use the EVMS install option only on a new device.

If you have data volumes on the system device, take one or more of the following precautionary
measures:
Š Move the data volumes from the system device to another device.
Š If you cannot move the volumes, make a backup copy of the data, so you can restore the data
volumes later from a backup copy.

Storage Deployment Considerations for the System Device

By default, the YaST install for SUSE Linux Enterprise Server uses the Linux Volume Manager to
manage the system device. The install procedures in this section describe how to install SUSE Linux
Enterprise Server with EVMS as the volume manager of the system device. The instructions assume
the following:
Š You want to use EVMS to manage the system device.
Š Only the system device is to be configured during the install.
Š Other devices on the system are not configured during the install, or are attached to the server
later. These additional devices are configured only after the system is operating and performing
as expected.

16 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
2.1.2 During the Server Install
To install Linux with EVMS as the volume manager for your boot and system partitions, you must
modify the Partitioning configuration in the Installation Settings.

WARNING: The following procedure destroys all data on the system device.

1 Begin the install, according to the instructions provided in Deployment (http://

www.novell.com/documentation/sles10/sles_admin/data/part_setup.html) in the SUSE Linux
Enterprise 10 Installation and Administration Guide (http://www.novell.com/documentation/
sles10/sles_admin/data/bookinfo_book_sles_admin.html).
2 When the installation reaches the Installations Settings screen, delete the proposed LVM-based
partioning solution. This deletes the proposed partitions and the partition table on the system
device so that the device can be marked to use EVMS as the volume manager instead of LVM.
2a In the list of Installation Settings, select Partitioning.
2b In the Partitioning menu, select Create Custom Partition Setup, then click Next.
2c Select Custom Partition - for Experts, then click Next to open the Expert Partitioner
dialog box.
2d Select Expert > Delete Partition Table and Disk Label, then click Yes twice to continue
through the Warning advisories.
This deletes the recommended partitions and the partition table on the system disk.
3 Create a primary partition on the system disk to use as the boot partition:
3a Click Create.
3b From the list of devices, select the device you want to use for the boot partition, such as /
dev/hda, then click OK.
If you have a single system disk, only one device is available, and you are not prompted to
choose the device.
3c Select Primary Partition, then click OK.
3d Select Format, then select the native Linux file system you want to use, such as Ext3.

IMPORTANT: In a paravirtualized environment, use Ext2 as the file system for the boot
device.

3e In Size (End Value) field, specify 200 MB or larger.

For example, to set the size at 300 MB, type 300M.
3f Set the mount point to /boot.
3g Click OK.
The partition appears as a logical device in the devices list, such as /dev/hda1.
4 Create a second primary partition on the system disk to use for both the swap and system
volumes:
4a Click Create.
4b From the list of devices, select the device you want to use for the second primary partition,
such as /dev/hda, then click OK.
If you have a single system disk, only one device is available and you are not prompted to
choose the device.

Using EVMS to Manage Devices 17

 novdocx (en) 13 May 2009
4c Select Primary Partition, then click OK.
4d Select Do Not Format, then select Linux LVM (0x8E) from the list of file system IDs.
4e In Size (End Value field), set the cylinder End value to 5 GB or larger, depending on the
combined partition size you need to contain your system and swap volumes.

IMPORTANT: Do not make the system partition larger than necessary. The remaining
space on the system disk can be used to create NSS volumes or native Linux volumes that
are managed by EVMS.

To determing how much space to use, consider the following recommendations:

Š For your system volume, allow 2 GB (minimum) to 10 GB (recommended),
depending on the OES services that you intend to install.
Š If you intend to create additional NSS volumes on the same physical disk, you must
leave unpartitioned space available.
Š Set aside 128 MB or larger for the swap volume.
Swap management is different for Linux kernel 2.4.10 and later. How much swap to
add depends on the RAM size, the tasks that are planned for the system, and whether
you want to make more virtual memory available than the RAM provides.
Some swap (at least 128 MB) is good to have to minimize the risk of losing data
when active processes run out of RAM space. Swap is not required for systems with
more than 1 GB of RAM. You must have at least 1 GB of virtual memory (RAM plus
swap) during the install, but if the swap is more than 2 GB, you might not be able to
install on some machines.
Š The total size should be the size you need for your system volume plus the size you
need for your swap volume.
For example, if you have a 20 GB hard drive with 2 GB of RAM and plan to install all of
the OES services on the system volume, your system partition should be at least 11 GB.
The remaining 9 GB should remain as free unpartitioned space that can be used for NSS
volumes or other Linux partitions that you might want to create later.
4f Click OK.
The partition appears as a logical device in the devices list, such as /dev/hda2.
5 Modify the volume management type from LVM to EVMS for the second primary partition
you created in Step 4:
5a At the bottom of the page, click EVMS.
Available partitions for EVMS appear as devices under /dev/evms, such as /dev/evms/
hda2.
5b In the EVMS Configurator, select the LVM partition created in Step 4, then click Create
Container.
5c In the Create EVMS Container dialog box, select the partition, specify the container name
(such as system), then click Add Volume to create the lvm/system container, where
system is the container name.
5d Click OK.
The EVMS Configurator displays the lvm/system container you just created, its size, and
free space.

18 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
6 Create the swap volume in the lvm/system container:
6a Select lvm/system, then click Add.
6b In the Create Logical Volume dialog box, select Format, then select Swap from the File
System drop-down menu.
6c Specify swap as the volume name.
6d Specify 1 GB (recommended) for the swap volume.
The swap size should be 128 MB or larger, with a recommended size of 1 GB. For an
explanation of this recommendation, see Step 4e.
6e Specify the mount point as swap.
6f Click OK.
7 Create the system volume in the lvm/system container:
7a Select lvm/system, then click Add.
7b In the Create Logical Volume dialog box, select Format, then select the file system to use
from the File System drop-down menu, such as Reiser or Ext3.
7c In the Volume Name field, specify a volume name, such as sys_lx.
7d In the Size field, click Max to set the size of the system volume as the remaining space
available in the lvm/system partition.
7e Specify the mount point as / (root volume).
7f Click OK.
8 Click Next to return to the list of devices.
Below is an example of the physical and logical devices that should be configured on your
system. Your setup depends on the number of devices in the server and the sizes you choose for
your partitions.

Device Size F Type Mount Start End Used By

/dev/hda 149.0 GB 6Y160p0 0 19456

/dev/hda1 305.9 MB F Linux Native /boot 0 38

(Reiser)

/dev/hda2 20.0 GB Linux LVM 39 2649 EVMS

lvm/
system
/dev/hdb 111.8 GB SP1203N 0 14595

/dev/evms/lvm/system/ 10.0 GB F EVMS / - -

sys_lx

/dev/evms/lvm/system/ 1.0 GB F EVMS swap - -

swap

9 Click Next to return to the Installation Settings page.

You can dismiss the message warning that you should not mix EVMS and non-EVMS
partitions on the same device.
10 Continue with the SUSE Linux Enterprise Server installation.

Using EVMS to Manage Devices 19

 novdocx (en) 13 May 2009
IMPORTANT: After the install is complete, make sure to perform the mandatory post-install
configuration of the related system settings to ensure that the system device functions properly
under EVMS. Otherwise, the system fails to boot properly.
For information, see “After the Server Install” on page 20.

2.1.3 After the Server Install

After the SUSE Linux Enterprise Server 10 install is complete, you must perform the following
tasks to ensure that the system device functions properly under EVMS:
Š “Edit the /etc/fstab File” on page 20
Š “Make a New initrd” on page 21
Š “Disable the boot.lvm and boot.md Services” on page 21
Š “Enable the boot.evms Service” on page 21
Š “Restart the Server” on page 22
Š “Verify the System Services” on page 22

Edit the /etc/fstab File

When you boot the system, the kernel reads the /etc/fstab file to identify which file systems
should be mounted and then mounts them. This file contains a table of file system information about
the root (/), /boot, and swap partitions plus other partitions and file systems you want to mount.
The /boot partition is separate from the EVMS container where you placed the root (/) and swap
partitions and is not managed by EVMS at this time. However, in the following steps, you disable
boot.lvm and boot.md, then enable boot.evms. In effect, this forces EVMS to scan all the
partitions at boot time, including the /boot partition, and it activates /boot under the /dev/evms
directory. Therefore, this makes /boot a partition that is discovered by EVMS at startup, and
requires that the device be listed under /dev/evms in the fstab file so it can be found when booting
with boot.evms. You must edit the /etc/fstab file to modify the location of the /boot partition
so it is under the /dev/evms directory.
In fstab, the entry for the boot device might present the boot device by the device node name (such
as /dev/sda1) or by the UUID pathname (such as /dev/disk/by-id/scsi-
SServeRA_Drive_1_600BC00000-part1). In ether case, that name for the boot device must be
changed to include evms in the path, such as /dev/evms/sda1.
The procedure in this section shows how to change /dev/sda1 to /dev/evms/sda1. Replace sda1
with the device name of the device you used for your /boot partition.

IMPORTANT: When working in the /etc/fstab file, do not leave any stray characters or spaces
in the file. This is a configuration file, and it is highly sensitive to such mistakes.

To modify the path of the boot device in the /etc/fstab file, complete the following procedure:

1 Open the /etc/fstab file in a text editor.

2 Locate the line that contains the /boot partition.
For example, if your /boot partition uses device sda1 and the Reiser file system, look for a
line similar to this:

20 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
/dev/sda1 /boot reiser defaults 1 1
3 In the Device Name column, modify the location of the /boot partition from /dev to /dev/
evms so it can be managed by EVMS. Modify only the device name by adding /evms to the
path:
/dev/evms/sda1 /boot reiser defaults 1 1
4 Save the file.
The changes do not take effect until the server is restarted. Do not restart at this time.
5 Continue with “Make a New initrd” on page 21.

Make a New initrd

1 Open a terminal console, and log in as the root user.

2 At the console prompt, enter
mkinitrd
This creates a new initrd file with the correct settings for the boot device. The changes do not
take effect until the server is restarted. Do not restart at this time.
3 Continue with “Disable the boot.lvm and boot.md Services” on page 21.

Disable the boot.lvm and boot.md Services

Disable the boot.lvm and boot.md services so they do not run at boot time (runlevel B). EVMS
now handles the boot.

1 In YaST, click System > System Services (Runlevel) > Expert Mode.
2 Select boot.lvm.
3 Click Set/Reset > Disable the Service.
4 Select boot.md.
5 Click Set/Reset > Disable the Service.
6 Click Finish, then click Yes.
The changes do not take effect until the server is restarted. Do not restart at this time.
7 Continue with “Enable the boot.evms Service” on page 21.

Enable the boot.evms Service

The boot.evms service should be enabled automatically after the install, but you should verify that
it is enabled.

1 In YaST, click System > System Services (Runlevel) > Expert Mode.
2 Select boot.evms.
3 Click Set/Reset > Enable the Service.
The B runlevel option is automatically selected.
4 Click Finish, then click Yes.
The changes do not take effect until the server is restarted.

Using EVMS to Manage Devices 21

 novdocx (en) 13 May 2009
NOTE: Effective in SUSE Linux Enterprise 10, the /dev directory is on tmpfs, and the device
nodes are automatically re-created on boot. It is no longer necessary to modify the /etc/
init.d/boot.evms script to delete the device nodes on system restart, as was required for
previous versions of SUSE Linux.

5 Continue with “Restart the Server” on page 22.

Restart the Server

1 Restart the server to apply the post-install configuration settings.

2 On restart, if the system device does not appear, it might be because EVMS has not been
activated. At the prompt, enter
evms_activate

Verify the System Services

After the post-install configuration is complete and you have restarted the server, make sure the
server is operating as expected.

2.2 Configuring an Existing System Device to

Use EVMS
If you have already installed Linux with a different volume manager for the system device (that is,
the devices where you installed the /boot, swap, or root (/) partitions), you can optionally
configure the device for EVMS at any time after the install.

If you do not configure the device to use EVMS, you must manage the device and all of its volumes
with its current volume manager (the default is LVM), and free space on the device cannot be used
for volumes you want to create using EVMS. Beginning with the Linux 2.6 kernel, a given device
cannot be managed by multiple volume managers. However, you can have different volume
managers for different devices.

The following procedures assume that you installed Linux with three partitions on a single SCSI
device named sda:
/dev/sda1 reiserfs /boot

/dev/sda2 swap swap

/dev/sda3 reiserfs /

IMPORTANT: Make sure to modify the following procedures as necessary for your specific setup.

Š Section 2.2.1, “Disable the boot.lvm and boot.md Services,” on page 23

Š Section 2.2.2, “Enable the boot.evms Service,” on page 23
Š Section 2.2.3, “Edit the /etc/fstab File,” on page 23
Š Section 2.2.4, “Edit the Boot Loader File,” on page 24
Š Section 2.2.5, “Force the RAM Disk to Recognize the Root Partition,” on page 25
Š Section 2.2.6, “Restart the Server,” on page 26
Š Section 2.2.7, “Verify that EVMS Manages the Boot, Swap, and Root Partitions,” on page 26

22 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
2.2.1 Disable the boot.lvm and boot.md Services
You need to disable boot.lvm (handles devices for Linux Volume Manager) and boot.md (handles
multiple devices in software RAIDs) so they do not run at boot time. In the future, you want
boot.evms to run at boot time instead.

1 In YaST, click System > Runlevel Editor > Expert Mode.

2 Select boot.lvm.
3 Click Set/Reset > Disable the Service.
4 Select boot.md.
5 Click Set/Reset > Disable the Service.
6 Click Finish, then click Yes.
The changes do not take effect until the server is restarted. Do not restart at this time.
7 Continue with Section 2.2.2, “Enable the boot.evms Service,” on page 23.

2.2.2 Enable the boot.evms Service

You need to enable the boot.evms service so that it boots devices when you restart the server.

1 In YaST, click System > Runlevel Editor > Expert Mode.

2 Select boot.evms.
3 Click Set/Reset > Enable the Service.
The B runlevel option is automatically selected.
4 Click Finish, then click Yes.
The changes do not take affect until the server is restarted. Do not restart at this time.

NOTE: Effective in SUSE Linux Enterprise 10, the /dev directory is on tmpfs and the device
nodes are automatically re-created on boot. It is no longer necessary to modify the /etc/
init.d/boot.evms script to delete the device nodes on system restart as was required for
previous versions of SUSE Linux.

5 Continue with “Edit the /etc/fstab File” on page 23.

2.2.3 Edit the /etc/fstab File

When you boot the system, the kernel reads the /etc/fstab file to identify which file systems
should be mounted and then mounts them. This file contains a table of file system information about
the /boot, swap, and root (/) partitions plus other partitions and file systems you want to mount.
You must edit the /etc/fstab file to modify the mount location of these three partitions so they are
mounted under the /dev/evms directory. For example, change /dev/sda1 to /dev/evms/sda1.
Although the /boot partition is not managed by EVMS, the boot.evms script forces EVMS to scan
all the partitions at boot time, including the /boot partition, and it activates /boot under the /dev/
evms directory. Therefore, this makes /boot a partition that is discovered by EVMS at startup, and
requires that the device’s path be listed under /dev/evms in the fstab file so it can be found when
booting with boot.evms.
Make sure to replace sda1, sda2, and sda3 with the device names you used for your partitions.

Using EVMS to Manage Devices 23

 novdocx (en) 13 May 2009
IMPORTANT: When working in the /etc/fstab file, do not leave any stray characters or spaces
in the file. This is a configuration file, and it is highly sensitive to such mistakes.

1 Open the /etc/fstab file in a text editor.

2 Locate the line that contains the /boot partition.
For example, if your /boot partition uses device sda1 and the Reiser file system, look for a line
similar to this:
/dev/sda1 /boot reiser defaults 1 1
3 In the Device Name column, modify the mount location of the /boot partition from /dev to /
dev/evms so it can be managed by EVMS. Modify only the device name by adding /evms to
the path:
/dev/evms/sda1 /boot reiser defaults 1 1
4 Repeat Step 2 and Step 3 to edit the Device Name entry in the lines for the swap and root (/)
partitions.
For example, change /dev/sda2 to /dev/evms/sda2, and change /dev/sda3 to /dev/evms/
sda3.
5 Save the file.
The changes do not take effect until the server is restarted. Do not restart at this time.
6 Continue with Section 2.2.4, “Edit the Boot Loader File,” on page 24.

2.2.4 Edit the Boot Loader File

When you boot the system, the kernel reads the boot loader file for information about your system.
For Grub, this is the /boot/grub/menu.1st file. For LILO, this is the /etc/lilo.conf file.
You must edit the boot loader file to modify the mount location of partitions so they are mounted
under the /dev/evms directory. For example, change /dev/sda1 to /dev/evms/sda1. Make sure to
replace the path for all lines that contain device paths in the files. You can modify the boot loader
file by editing fields in YaST, or use a text editor to modify the file directly.

IMPORTANT: When working in the boot loader file, do not leave any stray characters or spaces in
the file. This is a configuration file, and it is highly sensitive to such mistakes.

Using YaST

To modify the boot loader file in the YaST Control Center:

1 Log in as the root user or equivalent.

2 In Yast, select System > Boot Loader.
3 Modify the boot loader image so that the root file system is mounted as /dev/evms/ instead of
/dev/.
3a Select the boot loader image file, then click Edit.
3b Edit the device path in the Root Device field.
For example, change the Root Device value from
/dev/sda2

24 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
to
/dev/evms/sda2
Replace sda2 with the actual device on your machine.
3c Edit any device paths in the Other Kernel Parameters field.
3d Click OK to save the changes and return to the Boot Loader page.
4 Modify the failsafe image so that the failsafe root file system is mounted as /dev/evms/
instead of /dev/.
4a Select the failsafe image file, then click Edit.
4b Edit the device path in the Root Device field.
4c Check the Other Kernel Parameters field and make changes if needed.
4d Click OK to save the change and return to the Boot Loader page.
5 Click Finish.
6 Continue with Section 2.2.5, “Force the RAM Disk to Recognize the Root Partition,” on
page 25.

Using a Text Editor

To edit the boot loader file in a text editor:

1 Log in as the root user or equivalent.

2 Open the boot loader file in a text editor.
For Grub, this is the /boot/grub/menu.1st file. For LILO, this is the /etc/lilo.conf file.
3 Locate the line that contains the root= parameter.
For example, if your root file system uses device sda1, look for a line similar to this:
kernel (sd0,0)/vmlinuz root=/dev/sda1 vga=0x31a splash=silent showopts
4 Modify the mount location from /dev to /dev/evms so it can be managed by EVMS.
For example, after the change, the line looks like this:
kernel (sd0,0)/vmlinuz root=/dev/evms/sda1 vga=0x31a splash=silent
showopts
5 Repeat Step 3 and Step 4 to locate other lines in the file that need to be similarly modified.
6 Save the file.
The changes do not take effect until the server is restarted. Do not restart at this time.
7 Continue with Section 2.2.5, “Force the RAM Disk to Recognize the Root Partition,” on
page 25.

2.2.5 Force the RAM Disk to Recognize the Root Partition

The mkinitrd(8) command creates file system images for use as initial RAM disk (initrd) images.
These RAM disk images are often used to preload the block device modules (SCSI or RAID) needed
to access the root file system.

You might need to force the RAM to update its device node information so that it loads the root (/)
partition from the /dev/evms path.

Using EVMS to Manage Devices 25

 novdocx (en) 13 May 2009
NOTE: Recent patches to mkinitrd might resolve the need to do this task. For the latest version of
mkinitrd, see Recommended Updates for mkinitrd (http://support.novell.com/techcenter/psdb/
24c7dfbc3e0c183970b70c1c0b3a6d7d.html) at the Novell Technical Support Center.

1 At a terminal console prompt, enter the EVMS Ncurses command as the root user or
equivalent:
evmsn
2 Review the output to verify that EVMS shows only the /boot and swap partitions as active in
EVMS.
You should see the following devices mounted (with your own partition names, of course) for
these two partitions:
/dev/evms/sda1
/dev/evms/sda2
3 At a terminal console prompt, enter the following to update the initrd image with the /dev/
evms path information for the root (/) partition:
/sbin/mkinitrd -f evms
This does not take effect until you restart the server.
4 Continue with Section 2.2.6, “Restart the Server,” on page 26.

2.2.6 Restart the Server

1 Restart the server to apply the post-install configuration settings.
When your system restarts, the kernel loads the init-ramdisk, which runs the EVMS tools to
activate your volumes and mount your root file system. Then your boot scripts run the EVMS
tools once more to make sure your /dev/evms/ directory correctly reflects the current state of
your volumes. Finally, the remaining EVMS volumes are mounted as specified in your /etc/
fstab file. Everything else on your system should start up as you would normally expect.
2 Continue with Section 2.2.7, “Verify that EVMS Manages the Boot, Swap, and Root
Partitions,” on page 26.

2.2.7 Verify that EVMS Manages the Boot, Swap, and Root
Partitions
1 At a terminal prompt, enter the EVMS Ncurses command as the root user or equivalent:
evmsn
2 Review the output to verify that EVMS shows the /boot, swap, and root (/) partitions as active
in EVMS.
You should see the following devices mounted (with your own partition names, of course)
under the /dev/evms directory:
/dev/evms/sda1
/dev/evms/sda2
/dev/evms/sda3

26 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
2.3 Configuring LVM Devices to Use EVMS
Use the following post-installation procedure to configure data devices (not system devices) to be
managed by EVMS. If you need to configure an existing system device for EVMS, see Section 2.2,
“Configuring an Existing System Device to Use EVMS,” on page 22.

1 In a terminal console, run the EVMSGUI by entering the following as the root user or
equivalent:
evmsgui
2 In the Volumes panel, review the names that EVMS reports as compatibility volumes, find the
devices that represent the devices you want to manage using EVMS, then write down the
names for future reference.
For example, /dev/sdb1.
3 In a text editor, edit the /etc/fstab file to use the EVMS volume names.
For example, change the following entry for an LVM2 volume from this
/dev/sdb1 / reiserfs defaults 1 2
to this
/dev/evms/lvm2/sdb1 / reiserfs defaults 1 2

IMPORTANT: Make sure not to leave any stray characters or spaces in the line.

With these changes, each time your system boots, your file system is mounted using EVMS as
the volume manager.
4 Update the boot scripts as follows:
Š The command evms_activate must be run from your boot scripts in order to activate
your volumes so they can be mounted.
Š If you run software-RAID (boot.md) or LVM (boot.lvm) boot files in your boot scripts,
and if you are moving all devices to EVMS, remove or disable those commands.
5 If you have not already done so, enable the boot.evms service.
For information, see “Enable the boot.evms Service” on page 21.
6 Restart your system.

2.4 Using EVMS with iSCSI Volumes

If your EVMS devices, RAIDs, and volumes use storage devices from an iSCSI SAN, make sure
that your system starts iSCSI before EVMS so that the SAN and its disks are available to EVMS on
system startup. iSCSI must be started and running before any disks or volumes on the iSCSI SAN
can be accessed. If EVMS starts before iSCSI, EVMS cannot see or access the devices in the iSCSI
SAN to mount the storage objects they contain, so the EVMS devices, RAIDs, and volumes might
not be visible or accessible.

If EVMS starts before iSCSI on your system so that your EVMS devices, RAIDs, and volumes are
not visible or accessible, you must correct the order in which iSCSI and EVMS are started. Enter the
chkconfig command at the Linux server console of every server that is part of your iSCSI SAN.

1 At a terminal console prompt, enter either

chkconfig evms on

Using EVMS to Manage Devices 27

 novdocx (en) 13 May 2009
or
chkconfig boot.evms on
This ensures that EVMS and iSCSI start in the proper order each time your servers restart.

2.5 Using the ELILO Loader Files (IA-64)

On a SUSE Linux Enterprise Server boot device EFI System Partition, the full paths to the loader
and configuration files are:
/boot/efi/SuSE/elilo.efi

/boot/efi/SuSE/elilo.conf

When configuring partitioning during the install on IA64 systems, set the file system type for the /
boot partition to vfat, then choose Fstab Options and set the Arbitrary option value to umask=077
to ensure that the partition is accessible only to administrators.

WARNING: Whenever you manually alter the kernel or initrd on your system, make sure to run /
sbin/elilo before shutting down the computer. If you leave out this step, your system might not be
bootable.

2.6 Starting EVMS

If EVMS does not start during the system boot, you must activate it manually.

1 Open a terminal console, then log in as the root user or equivalent.

2 At the terminal console prompt, enter
evms_activate

2.7 Starting the EVMS Management Tools

Use the following procedure to start the EVMS management tools.

IMPORTANT: When you are done, make sure to exit the EVMS UI tool. When it is running, the
EVMS UI tool locks the EVMS engine, potentially blocking other EVMS actions from taking place.

1 Open a terminal console, then log in as the root user or equivalent.

2 Enter one of the following commands to open the desired EVMS UI:

Command Description

evmsgui Starts the graphical interface for EVMS GUI. For information about features in
this interface, see ”EVMS GUI” (http://evms.sourceforge.net/user_guide/#GUI)
in the EVMS User Guide at the EVMS project on SourceForge.net.

evmsn Starts the text-mode interface for EVMS Ncurses. For information about
features in this interface, see the “EVMS Ncurses Interface” (http://
evms.sourceforge.net/user_guide/#NCURSES) in the EVMS User Guide at the
EVMS project on SourceForge.net.

28 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Command Description

evms Starts the EVMS commandline interpreter (CLI) interface. For information about
command options, see “EVMS Command Line Interpreter” (http://
evms.sourceforge.net/user_guide/#COMMANDLINE) in the EVMS User Guide
at the EVMS project on SourceForge.net.

To stop evmsgui from running automatically on restart:

1 Close evmsgui.
2 Do a clean shutdown (not a restart).
3 Start the server.
When the server comes back up, evmsgui is not automatically loaded on restart.

Using EVMS to Manage Devices 29

 novdocx (en) 13 May 2009

30 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Using UUIDs to Mount Devices
3

This section describes the optional use of UUIDs instead of device names to identify file system
3
devices in the boot loader file and the /etc/fstab file.
Š Section 3.1, “Naming Devices with udev,” on page 31
Š Section 3.2, “Understanding UUIDs,” on page 31
Š Section 3.3, “Using UUIDs in the Boot Loader and /etc/fstab File (x86),” on page 32
Š Section 3.4, “Using UUIDs in the Boot Loader and /etc/fstab File (IA64),” on page 33
Š Section 3.5, “Additional Information,” on page 34

3.1 Naming Devices with udev

In the Linux 2.6 and later kernel, udev provides a userspace solution for the dynamic /dev directory,
with persistent device naming. As part of the hotplug system, udev is executed if a device is added
or removed from the system.

A list of rules is used to match against specific device attributes. The udev rules infrastructure
(defined in the /etc/udev/rules.d directory) provides stable names for all disk devices,
regardless of their order of recognition or the connection used for the device. The udev tools
examine every appropriate block device that the kernel creates to apply naming rules based on
certain buses, drive types, or file systems. For information about how to define your own rules for
udev, see Writing udev Rules (http://reactivated.net/writing_udev_rules.html).

Along with the dynamic kernel-provided device node name, udev maintains classes of persistent
symbolic links pointing to the device in the /dev/disk directory, which is further categorized by
the by-id, by-label, by-path, and by-uuid subdirectories.

NOTE: Other programs besides udev, such as LVM or md, might also generate UUIDs, but they are
not listed in /dev/disk.

3.2 Understanding UUIDs

A UUID (Universally Unique Identifier) is a 128-bit number for a file system that is unique on both
the local system and across other systems. It is a randomly generated with system hardware
information and time stamps as part of its seed. UUIDs are commonly used to uniquely tag devices.
Š Section 3.2.1, “Using UUIDs to Assemble or Activate File System Devices,” on page 32
Š Section 3.2.2, “Finding the UUID for a File System Device,” on page 32

Using UUIDs to Mount Devices 31

 novdocx (en) 13 May 2009
3.2.1 Using UUIDs to Assemble or Activate File System
Devices
The UUID is always unique to the partition and does not depend on the order in which it appears or
where it is mounted. With certain SAN devices attached to the server, the system partitions are
renamed and moved to be the last device. For example, if root (/) is assigned to /dev/sda1 during
the install, it might be assigned to /dev/sdg1 after the SAN is connected. One way to avoid this
problem is to use the UUID in the boot loader and /etc/fstab files for the boot device.
The device ID assigned by the manufactuer for a drive never changes, no matter where the device is
mounted, so it can always be found at boot. The UUID is a property of the filesystem and can
change if you reformat the drive. In a boot loader file, you typically specify the location of the
device (such as /dev/sda1 or /dev/evms/sda1) to mount it at system boot. The boot loader can
also mount devices by their UUIDs and administrator-specified volume labels. However, if you use
a label and file location, you cannot change the label name when the partition is mounted.

You can use the UUID as criterion for assembling and activating software RAID devices. When a
RAID is created, the md driver generates a UUID for the device, and stores the value in the md
superblock.

3.2.2 Finding the UUID for a File System Device

You can find the UUID for any block device in the /dev/disk/by-uuid directory. For example, a
UUID looks like this:
e014e482-1c2d-4d09-84ec-61b3aefde77a

3.3 Using UUIDs in the Boot Loader and /etc/

fstab File (x86)
After the install, you can optionally use the following procedure to configure the UUID for the
system device in the boot loader and /etc/fstab files for your x86 system.
1 Install the SUSE® Linux Enterprise Server for x86 with no SAN devices connected.
2 After the install, boot the system.
3 Open a terminal console as the root user or equivalent.
4 Navigate to the /dev/disk/by-uuid directory to find the UUID for the device where you
installed /boot, /root, and swap.
4a At the terminal console prompt, enter
cd /dev/disk/by-uuid
4b List all partitions by entering
ll
4c Find the UUID, such as
e014e482-1c2d-4d09-84ec-61b3aefde77a —> /dev/sda1
5 Edit /boot/grub/menu.1st file, using the Boot Loader option in YaST2 or using a text editor.
For example, change
kernel /boot/vmlinuz root=/dev/sda1

32 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
to
kernel /boot/vmlinuz root=/dev/disk/by-uuid/e014e482-1c2d-4d09-84ec-
61b3aefde77a

IMPORTANT: Make a copy of the original boot entry, then modify the copy. If you make a
mistake, you can boot the server without the SAN connected, and fix the error.

If you use the Boot Loader option in YaST, there is a defect where it adds some duplicate lines
to the boot loader file when you change a value. Use an editor to remove the following
duplicate lines:
color white/blue black/light-gray
default 0
timeout 8
gfxmenu (sd0,1)/boot/message
When you use YaST to change the way that the root (/) device is mounted (such as by UUID or
by label), the boot loader configuration needs to be saved again to make the change effective
for the boot loader.
6 As the root user or equivalent, do one of the following to place the UUID in the /etc/fstab
file:
Š Open YaST to System > Partitioner, select the device of interest, then modify Fstab
Options.
Š Edit the /etc/fstab file to modify the system device from the location to the UUID.
For example, if the root (/) volume has a device path of /dev/sda1 and its UUID is
e014e482-1c2d-4d09-84ec-61b3aefde77a, change line entry from
/dev/sda1 / reiserfs acl,user_xattr 1 1
to
UUID=e014e482-1c2d-4d09-84ec-61b3aefde77a / reiserfs
acl,user_xattr 1 1

IMPORTANT: Make sure to make a backup copy of the fstab file before you begin, and
do not leave stray characters or spaces in the file.

3.4 Using UUIDs in the Boot Loader and /etc/

fstab File (IA64)
After the install, use the following procedure to configure the UUID for the system device in the
boot loader and /etc/fstab files for your IA64 system. IA64 uses the EFI BIOS. Its file system
configuration file is /boot/efi/SuSE/elilo.conf instead of /etc/fstab.
1 Install the SUSE Linux Enterprise Server for IA64 with no SAN devices connected.
2 After the install, boot the system.
3 Open a terminal console as the root user or equivalent.
4 Navigate to the /dev/disk/by-uuid directory to find the UUID for the device where you
installed /boot, /root, and swap.
4a At the terminal console prompt, enter
cd /dev/disk/by-uuid

Using UUIDs to Mount Devices 33

 novdocx (en) 13 May 2009
4b List all partitions by entering
ll
4c Find the UUID, such as
e014e482-1c2d-4d09-84ec-61b3aefde77a —> /dev/sda1
5 Edit the boot loader file, using the Boot Loader option in YaST2.
For example, change
root=/dev/sda1
to
root=/dev/disk/by-uuid/e014e482-1c2d-4d09-84ec-61b3aefde77a
6 Edit the /boot/efi/SuSE/elilo.conf file to modify the system device from the location to
the UUID.
For example, change
/dev/sda1 / reiserfs acl,user_xattr 1 1
to
UUID=e014e482-1c2d-4d09-84ec-61b3aefde77a / reiserfs acl,user_xattr
1 1

IMPORTANT: Make sure to make a backup copy of the /boot/efi/SuSE/elilo.conf file

before you begin, and do not leave stray characters or spaces in the file.

3.5 Additional Information

For more information about using udev(8) for managing devices, see “Dynamic Kernel Device
Management with udev” (http://www.novell.com/documentation/sles10/sles_admin/data/
cha_udev.html) in the SUSE® Linux Enterprise Server 10 Installation and Administration Guide.
For more information about udev(8) commands, see its man page. Enter the following at a terminal
console prompt:
man 8 udev

34 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Managing EVMS Devices
4

This section describes how to initialize a disk for EVMS management by adding a segment
4
management container to manage the partitions that you later add to the disk.
Š Section 4.1, “Understanding Disk Segmentation,” on page 35
Š Section 4.2, “Initializing Disks,” on page 36
Š Section 4.3, “Removing the Segment Manager from a Device,” on page 38
Š Section 4.4, “Creating Disk Segments (or Partitions),” on page 38
Š Section 4.5, “Configuring Mount Options for Devices,” on page 39
Š Section 4.6, “What’s Next,” on page 41

4.1 Understanding Disk Segmentation

In EVMS, you initialize a disk by assigning a segment manager to it. The segment manager creates
metadata for the disk and exposes its free space so you can subdivide it into one or multiple
segments (also called partitions).
Š Section 4.1.1, “Segment Managers,” on page 35
Š Section 4.1.2, “Disk Segments,” on page 36

4.1.1 Segment Managers

The most commonly used segment manager is the DOS Segment Manager. The following table
describes the segment managers available in EVMS.

Table 4-1 EVMS Segment Managers

Segment Manager Description

DOS The standard MS-DOS* disk partitioning scheme. It is the most commonly used
partitioning scheme for Linux, NetWare®, Windows*, OS/2*, BSD, Solaris* X86,
and UnixWare*.

GPT (Globally Unique A partitioning scheme used for IA-64 platforms, as defined in the Intel*
Identifier (GUID) Extensible Firmware Interface (EIF) Specification. It is not compatible with
Partitioning Table) DOS, Windows, or OS/2 systems.

The GUID is also known as Universally Unique Identifier (UUID). The GPT
combines time and space descriptors to create this unique 128-bit tag for the
disk and its segments.

S/390 A partitioning scheme used exclusively for System/390 mainframes.

Cluster A partitioning scheme for high-availability clusters. It provides a GUID for the
disk, creates an EVMS container for the shared cluster devices, and specifies a
node ID for the node that owns the device and the cluster ID.

BSD A partitioning scheme for BSD UNIX*.

Managing EVMS Devices 35

 novdocx (en) 13 May 2009
Segment Manager Description

MAC A partitioning scheme for Mac-OS partitions.

4.1.2 Disk Segments

After you initialize the disk by adding a segment manager, you see metadata and free space
segments on the disk. You can then create one or multiple data segments in a disk segment.

Table 4-2 Disk Segment Types

Segment Type Description

Metadata A set of contiguous sectors that contain information needed by the segment
manager.

Free Space A set of contiguous sectors that are unallocated or not in use. Free space can
be used to create a segment.

Data A set of contiguous sectors that has been allocated from a disk. The segment
might be in use for a volume or a software RAID.

4.2 Initializing Disks

You must initialize new disks and disks that you want to reformat. After the disk is initialized, you
can subdivide, or carve, the device into one or more disk segments for your file systems.
Š Section 4.2.1, “Before You Begin,” on page 36
Š Section 4.2.2, “Guidelines,” on page 37
Š Section 4.2.3, “Adding a Segment Manager,” on page 37

4.2.1 Before You Begin

If you use large disks or disk arrays, use the vendor’s tools to carve them into the sizes that are
usable for the management tools you plan to use. For example, the md driver recognizes disks only
up to 2 TB in size, so the limit also applies to the md plug-in for EVMS. Software RAID devices you
create with EVMS can be larger than 2 TB, of course, because the md driver plug-in manages the
disks underneath that storage structure.

When you boot the server, EVMS scans and recognizes all devices it manages. If you add a new
device to the server or create a device using mkfs, EVMS automatically mounts it on reboot under /
dev/evms as a compatibility volume, such as /dev/evms/sdb.

IMPORTANT: If you cannot find a new disk, device, or volume, look under /dev/evms in a file
browser, or look for compatibility volumes in the Volumes Manager in the EVMS GUI (evmsgui).

36 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
4.2.2 Guidelines
Consider the following guidelines when initializing a disk:

Š EVMS might allow you to create segments without first adding a segment manager for the
disk, but it is best to explicitly add a segment manager to avoid problems later.

IMPORTANT: You must add a Cluster segment manager if you plan to use the devices for
volumes that you want to share as cluster resources.

Š When you initialize an existing disk that is already formatted, the process of adding a segment
manager destroys all data on the disk. If you want to keep the data on the disk, make sure to
back up the data before you begin this process.
Š For existing disks on the system or disks that you move from another system, you must delete
any existing volume management structures, and remove any segment managers. This removes
the device’s metadata and data, and destroys all data on the disk.

WARNING: Do not initialize the device that contains your current system disk or any device
that contains the /boot, swap, or root (/) volumes.

Š If a new disk does not show up in the list of Available Objects, look for it in the Volumes list to
see if the disk shows up as a compatibility volume. For example, a new disk sdb would show
up as /dev/evms/sdb. Delete it from the Volumes list to force the disk to show up in Available
Objects, then create segments as desired.

4.2.3 Adding a Segment Manager

Use the following procedure to assign a segment manager to device for servers using x86, x64, and
IA64 controllers. This option is not available for S390 platforms, so simply continue with
configuring software RAIDs or file system partitions, as desired.

WARNING: Adding a segment manager initializes the disk, completely removing all the segments
it contains. All the data stored on the device is lost.

1 If the disk has any existing volume management structures or an existing segment manager,
remove them.
1a Select Actions > Delete > Volume to view the Volumes list.
1b Select any existing volume management structures on the device, then click Delete.
1c Select Actions > Remove > Segment Manager from Storage Object.
1d Select the type of Segment Manager in use, then click Next.
1e Select the device, then click Remove.
2 If the disk is a new one that is listed as a compatibility volume in the Volumes list, delete it as a
compatibility volume.
2a Select Actions > Delete > Volume to view the Volumes list.
2b Select the device, then click Delete.
3 Add the Segment Manager.
3a In the list of Availability Objects, select the device, then click Actions > Add > Segment
Manager to Storage Object.

Managing EVMS Devices 37

 novdocx (en) 13 May 2009
3b From the list, select one of the following types of segment manager, then click Next.
Š DOS Segment Manager (the most common choice)
Š GPT Segment Manager (for IA-64 platforms)
Š Cluster Segment Manager (available only if it is a viable option for the selected disk)
3c Select the device from the list of Plugin Acceptable Objects, then click Next.
3d If required, specify the disk type as Linux.
3e Click Add to create the segment management container for the disk, then click OK to
dismiss the confirmation message.

4.3 Removing the Segment Manager from a

Device
1 If the disk has any existing volume management structures, remove them.
1a Select Actions > Delete > Volume to view the Volumes list.
1b Select any existing volume management structures on the device, then click Delete.
2 Select Actions > Remove > Segment Manager from Storage Object.
3 Select the type of Segment Manager in use, then click Next.
4 Select the device, then click Remove.

4.4 Creating Disk Segments (or Partitions)

1 In EVMS, select Actions > Create > Segment to see a list of segment managers.
2 From the list, select the segment manager for the device you want to manage, then click Next.
Š DOS Segment Manager (the most common choice)
Š GPT Segment Manager (for IA-64 platforms)
Š Cluster Segment Manager (available only if it is a viable option for the selected disk)
For information about these and other segment managers available, see “Segment Managers”
on page 35.
3 Select the storage object that you want to segment, then click Next.
4 Complete the required configuration options for the segment, and modify default values as
desired.
Š Size (MB): Specify the amount of space (in MB) that you want to use. Use the arrows or
type a value. The interface corrects the value to the lower or upper size limit if you specify
a size that is too small or that exceeds the amount of free space available.
Š Offset (Sectors): Specify the number of sectors to skip before beginning this partition if
you want to leave free space in front of it.
Š Partition Type: From the drop-down list, select Linux (default), Linux Swap, Linux LVM,
NTFS, HPFS, FAT16, or Other Partition Type.
Š Partition Type ID: This value changes automatically based on the Partition Type value,
except for the Other Partition Type option, where you must manually enter a value.
Š Bootable: Click Yes to make a primary partition active so that you can boot from it, or
click No to make it unbootable. No is the only option if you are creating a logical partition.

38 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Š Primary Partition: Click Yes for a primary partition, or click No for a logical partition.
Required settings are denoted in the page by an asterisk (*). All required fields must be
completed to make the Create button active.
5 Click Create to create the segment.
6 Verify that the new segment appears in the Segment list.

4.5 Configuring Mount Options for Devices

The following table describes the Fstab Options that are configurable in YaST. The values are
written to the /etc/fstab file and are applied upon reboot.

Table 4-3 Fstab Options in YaST

Fstab Option Description

Mount by Š Device name (K) (default, such as /dev/sda2)

Š Volume label (L)
Make sure to also specify a value in Volume Label.
Š UUID (U)
For information about why you might want to discover partitions and
devices by UUID, see Section 3.2.1, “Using UUIDs to Assemble or
Activate File System Devices,” on page 32.
Š Device ID (I)
Š Device Path (P)
Volume label A useful name to help you easily identify the volume on the server. By default,
this field is empty.

Mount read-only Select the check box to enable this option. It is deselected (disabled) by default.

If this option is enabled, files and directories cannot be modified or saved on the
volume.

No access time Select the check box to enable this option. It is deselected (disabled) by default.

By default, the Linux open(2) command updates the access time whenever a
file is opened. The No Access Time option disables the updating of access
time, so that reading a file does not update its access time. Enabling the No
Access Time option allows you to back up a volume without modifying the
access times of its files.

Mountable by user Select the check box to enable this option. It is deselected (disabled) by default.

If this option is enabled, the volume can be mounted by any user; root
privileges are not required.

Do Not Mount at Select the check box to enable this option. It is deselected (disabled) by default.
System Start-up
The system volumes such as /boot, swap, and root (/) should all be mounted
at system start. For other volumes, enable this option for a volume if you want
to mount it manually later using the mount command at a terminal console
prompt.

Managing EVMS Devices 39

 novdocx (en) 13 May 2009
Fstab Option Description

Data journaling mode For journaling file systems, select the preferred journaling mode:

Š Ordered: Writes data to the file system, then enters the metadata in the
journal. This is the default.
Š Journal: Writes data twice; once to the journal, then to the file system.
Š Writeback: Writes data to the file system and writes metadata in the
journal, but the writes are performed in any order.

Access Control LIsts Select this option to enable access control lists on the file system. It is enabled
(ACL) by default.

Extended user Select this option to enable extended user attributes on the file system. It is
attributes enabled by default.

Arbitrary option value Specify any mount option that is legal for the Mount Options column for a
device entry in the /etc/fstab file. Use a comma with no spaces to separate
multiple options.

You can modify these values for each entry by editing the /etc/fstab file, or use the following
procedure to modify the mount options for a volume in the /etc/fstab file from YaST.
1 Open YaST, then click System > Partitioning.
2 Select the device you want to modify, then click Fstab Options.

3 Modify the settings as desired, then click OK to accept your changes.

40 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
4.6 What’s Next
If multiple paths exist between your host bus adapters (HBAs) and the storage devices, configure
multipathing for the devices before creating software RAIDs or file system volumes on the devices.
For information, see Chapter 5, “Managing Multipath I/O for Devices,” on page 43.

If you want to configure software RAIDs, do it before you create file systems on the devices. For
information, see Chapter 6, “Managing Software RAIDs with EVMS,” on page 75.

Managing EVMS Devices 41

 novdocx (en) 13 May 2009

42 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Managing Multipath I/O for
5

Devices 5
This section describes how to manage failover and path load balancing for multiple paths between
the servers and block storage devices.
Š Section 5.1, “Understanding Multipathing,” on page 43
Š Section 5.2, “Planning for Multipathing,” on page 44
Š Section 5.3, “Multipath Management Tools,” on page 49
Š Section 5.4, “Configuring the System for Multipathing,” on page 53
Š Section 5.5, “Enabling and Starting Multipath I/O Services,” on page 59
Š Section 5.6, “Configuring Path Failover Policies and Priorities,” on page 59
Š Section 5.7, “Tuning the Failover for Specific Host Bus Adapters,” on page 67
Š Section 5.8, “Configuring Multipath I/O for the Root Device,” on page 67
Š Section 5.9, “Configuring Multipath I/O for an Existing Software RAID,” on page 68
Š Section 5.10, “Scanning for New Devices without Rebooting,” on page 69
Š Section 5.11, “Scanning for New Partitioned Devices without Rebooting,” on page 70
Š Section 5.12, “Viewing Multipath I/O Status,” on page 71
Š Section 5.13, “Managing I/O in Error Situations,” on page 72
Š Section 5.14, “Resolving Stalled I/O,” on page 73
Š Section 5.15, “Additional Information,” on page 73
Š Section 5.16, “What’s Next,” on page 74

5.1 Understanding Multipathing

Š Section 5.1.1, “What Is Multipathing?,” on page 43
Š Section 5.1.2, “Benefits of Multipathing,” on page 43

5.1.1 What Is Multipathing?

Multipathing is the ability of a server to communicate with the same physical or logical block
storage device across multiple physical paths between the host bus adapters in the server and the
storage controllers for the device, typically in Fibre Channel (FC) or iSCSI SAN environments. You
can also achieve multiple connections with direct attached storage when multiple channels are
available.

5.1.2 Benefits of Multipathing

Linux multipathing provides connection fault tolerance and can provide load balancing across the
active connections. When multipathing is configured and running, it automatically isolates and
identifies device connection failures, and reroutes I/O to alternate connections.

Managing Multipath I/O for Devices 43

 novdocx (en) 13 May 2009
Typical connection problems involve faulty adapters, cables, or controllers. When you configure
multipath I/O for a device, the multipath driver monitors the active connection between devices.
When the multipath driver detects I/O errors for an active path, it fails over the traffic to the device’s
designated secondary path. When the preferred path becomes healthy again, control can be returned
to the preferred path.

5.2 Planning for Multipathing

Š Section 5.2.1, “Guidelines for Multipathing,” on page 44
Š Section 5.2.2, “Using Multipathed Devices Directly or in EVMS,” on page 45
Š Section 5.2.3, “Using LVM2 on Multipath Devices,” on page 46
Š Section 5.2.4, “Using mdadm with Multipath Devices,” on page 46
Š Section 5.2.5, “Partitioning Multipath Devices,” on page 46
Š Section 5.2.6, “Supported Architectures for Multipath I/O,” on page 47
Š Section 5.2.7, “Supported Storage Arrays for Multipathing,” on page 47

5.2.1 Guidelines for Multipathing

Use the guidelines in this section when planning your multipath I/O solution.

Š “Prerequisites” on page 44
Š “Vendor-Provided Multipath Solutions” on page 44
Š “Disk Management Tasks” on page 45
Š “Software RAIDs” on page 45
Š “High-Availability Solutions” on page 45
Š “Volume Managers” on page 45
Š “Virtualization Environments” on page 45

Prerequisites

Š Multipathing is managed at the device level.

Š The storage array you use for the multipathed device must support multipathing. For more
information, see Section 5.2.7, “Supported Storage Arrays for Multipathing,” on page 47.
Š You need to configure multipathing only if multiple physical paths exist between host bus
adapters in the server and host bus controllers for the block storage device. You configure
multipath for the logical device as seen by the server.

Vendor-Provided Multipath Solutions

For some storage arrays, the vendor will provide its own multipathing software to manage
multipathing for the array’s physical and logical devices. In this case, you should follow the
vendor’s instructions for configuring multipathing for those devices.

44 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Disk Management Tasks

Perform the following disk management tasks before you attempt to configure multipathing for a
physical or logical device that has multiple paths:
Š Use third-party tools to carve physical disks into smaller logical disks.
Š Use third-party tools to partition physical or logical disks. If you change the partitioning in the
running system, the Device Mapper Multipath (DM-MP) module does not automatically detect
and reflect these changes. DM-MP must be reinitialized, which usually requires a reboot.
Š Use third-party SAN array management tools to create and configure hardware RAID devices.
Š Use third-party SAN array management tools to create logical devices such as LUNs. Logical
device types that are supported for a given array depend on the array vendor.

Software RAIDs

The Linux software RAID management software runs on top of multipathing. For each device that
has multiple I/O paths and that you plan to use in a software RAID, you must configure the device
for multipathing before you attempt to create the software RAID device. Automatic discovery of
multipathed devices is not available. The software RAID is not aware of the multipathing
management running underneath.

High-Availability Solutions

High-availability solutions for clustering typically run on top of the multipathing server. For
example, the Distributed Replicated Block Device (DRBD) high-availability solution for mirroring
devices across a LAN runs on top of multipathing. For each device that has multiple I/O paths and
that you plan to use in a DRDB solution, you must configure the device for multipathing before you
configure DRBD.

Volume Managers

Volume managers such as LVM2 and EVMS run on top of multipathing. You must configure
multipathing for a device before you use LVM2 or EVMS to create segment managers and file
systems on it.

Virtualization Environments

When using multipathing in a virtualization environment, the multipathing is controlled in the host
server environment. Configure multipathing for the device before you assign it to a virtual guest
machine.

5.2.2 Using Multipathed Devices Directly or in EVMS

If you want to use the entire LUN directly (for example, if you are using the SAN features to
partition your storage), you can simply use the /dev/disk/by-id/xxx names directly for mkfs,
fstab, your application, etc.

If the user-friendly names option is enabled in the /etc/multipath.conf file, you can optionally
use the /dev/mapper/mpathn device name because this name is aliased to the devices ID. For
information, see “Configuring User-Friendly Names or Alias Names in /etc/multipath.conf” on
page 57.

Managing Multipath I/O for Devices 45

 novdocx (en) 13 May 2009
5.2.3 Using LVM2 on Multipath Devices
By default, LVM2 does not recognize multipathed devices. To make LVM2 recognize the
multipathed devices as possible physical volumes, you must modify /etc/lvm/lvm.conf. It is
important to modify it in a way that it does not scan and use the physical paths, but only accesses the
multipath I/O storage through the multipath I/O layer.

To modify /etc/lvm/lvm.conf for multipath use:

1 Open the /etc/lvm/lvm.conf file in a text editor.
2 Change the filter and types entry in /etc/lvm/lvm.conf as follows:
filter = [ "a|/dev/disk/by-id/.*|", "r|.*|" ]
This allows LVM2 to scan only the by-id paths and reject everything else.
3 If you are also using LVM2 on non-multipathed devices, make the necessary adjustments to
suit your setup.
4 Save the file.
5 Add dm-multipath to /etc/sysconfig/kernel:INITRD_MODULES.
6 Make a new initrd to ensure that the Device Mapper Multipath services are loaded with the
changed settings. Enter
mkinitrd -f mpath
7 Reboot the server to apply the changes.

5.2.4 Using mdadm with Multipath Devices

The mdadm tool requires that the devices be accessed by the ID rather than by the device node path.
Therefore, the DEVICE entry in /etc/mdadm.conf should be set as follows:
DEVICE /dev/disk/by-id/*

This is the default handling in SUSE Linux Enterprise Server 10 and later.

5.2.5 Partitioning Multipath Devices

SUSE Linux Enterprise Server 10

In SUSE Linux Enterprise Server 10, the kpartx software is used in the /etc/init.d/
boot.multipath to add symlinks to the /dev/dm-* line in the multipath.conf configuration file
for any newly created partitions without requiring a reboot. This triggers udevd to fill in the /dev/
disk/by-* symlinks. The main benefit is that you can call kpartx with the new parameters without
having to reboot the server.

SUSE Linux Enterprise Server 9

In SUSE Linux Enterprise Server 9, it is not possible to partition multipath I/O devices themselves.
If the underlying physical device is already partitioned, the multipath I/O device reflects those
partitions and the layer provides /dev/disk/by-id/<name>p1 ... pN devices so you can access
the partitions through the multipath I/O layer. As a consequence, the devices need to be partitioned

46 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
prior to enabling multipath I/O. If you change the partitioning in the running system, DM-MP does
not automatically detect and reflect these changes. The device must be reinitialized, which usually
requires a reboot.

5.2.6 Supported Architectures for Multipath I/O

The multipathing drivers and tools in SUSE® Linux Enterprise Server 10 support all seven of the
supported processor architectures: IA32, AMD64/EM64T, IPF/IA64, p-Series (32-bit/64-bit), z-
Series (31-bit and 64-bit).

5.2.7 Supported Storage Arrays for Multipathing

The multipathing drivers and tools in SUSE Linux Enterprise Server 10 support most storage arrays.
The storage array that houses the multipathed device must support multipathing in order to use the
multipathing drivers and tools. Some storage array vendors provide their own multipathing
management tools. Consult the vendor’s hardware documentation to determine what settings are
required.
Š “Storage ArraysThat Are Automatically Detected for Multipathing” on page 47
Š “Tested Storage Arrays for Multipathing Support” on page 48
Š “Storage Arrays that Require Specific Hardware Handlers” on page 48

Storage ArraysThat Are Automatically Detected for Multipathing

The multipath-tools package automatically detects the following storage arrays:

3PARdata VV
Compaq* HSV110
Compaq MSA1000
DDN SAN MultiDirector
DEC* HSG80
EMC* CLARiiON* CX
EMC Symmetrix
FSC CentricStor*
Hewlett Packard* (HP*) A6189A
HP HSV110
HP HSV210
HP Open-
Hitachi* DF400
Hitachi DF500
Hitachi DF600
IBM* 3542
IBM ProFibre 4000R
NetApp*
SGI* TP9100
SGI TP9300
SGI TP9400
SGI TP9500

Managing Multipath I/O for Devices 47

 novdocx (en) 13 May 2009
STK OPENstorage DS280
Sun* StorEdge 3510
Sun T4

In general, most other storage arrays should work. When storage arrays are automatically detected,
the default settings for multipathing apply. If you want non-default settings, you must manually
create and configure the /etc/multipath.conf file. For information, see Section 5.4.5, “Creating
and Configuring the /etc/multipath.conf File,” on page 55.
Hardware that is not automatically detected requires an appropriate entry for configuration in the
DEVICES section of the /etc/multipath.conf file. In this case, you must manually create and
configure the configuration file. For information, see Section 5.4.5, “Creating and Configuring the /
etc/multipath.conf File,” on page 55.

Consider the following caveats:

Š Not all of the storage arrays that are automatically detected have been tested on SUSE Linux
Enterprise Server. For information, see “Tested Storage Arrays for Multipathing Support” on
page 48.
Š Some storage arrays might require specific hardware handlers. A hardware handler is a kernel
module that performs hardware-specific actions when switching path groups and dealing with
I/O errors. For information, see “Storage Arrays that Require Specific Hardware Handlers” on
page 48.
Š After you modify the /etc/multipath.conf file, you must run mkinitrd to re-create the
INITRD on your system, then reboot in order for the changes to take effect.

Tested Storage Arrays for Multipathing Support

The following storage arrays have been tested with SUSE Linux Enterprise Server:

EMC
Hitachi
Hewlett-Packard/Compaq
IBM
NetApp
SGI

Most other vendors’ storage arrays should also work. Consult your vendor’s documentation for
guidance. For a list of the default storage arrays recognized by the multipath-tools package, see
“Storage ArraysThat Are Automatically Detected for Multipathing” on page 47.

Storage Arrays that Require Specific Hardware Handlers

Storage arrays that require special commands on failover from one path to the other or that require
special nonstandard error handling might require more extensive support. Therefore, the Device
Mapper Multipath service has hooks for hardware handlers. For example, one such handler for the
EMC CLARiiON CX family of arrays is already provided.

IMPORTANT: Consult the hardware vendor’s documentation to determine if its hardware handler
must be installed for Device Mapper Multipath.

48 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
The multipath -t command shows an internal table of storage arrays that require special handling
with specific hardware handlers. The displayed list is not an exhaustive list of supported storage
arrays. It lists only those arrays that require special handling and that the multipath-tools
developers had access to during the tool development.

IMPORTANT: Arrays with true active/active multipath support do not require special handling, so
they are not listed for the multipath -t command.

A listing in the multipath -t table does not necessarily mean that SUSE Linux Enterprise Server
was tested on that specific hardware. For a list of tested storage arrays, see “Tested Storage Arrays
for Multipathing Support” on page 48.

5.3 Multipath Management Tools

The multipathing support in SUSE Linux Enterprise Server 10 is based on the Device Mapper
Multipath module of the Linux 2.6 kernel and the multipath-tools userspace package. You can
use mdadm to view the status of multipathed devices.
Š Section 5.3.1, “Device Mapper Multipath Module,” on page 49
Š Section 5.3.2, “Multipath I/O Management Tools,” on page 51
Š Section 5.3.3, “Using mdadm for Multipathed Devices,” on page 51
Š Section 5.3.4, “The Linux multipath(8) Command,” on page 52

5.3.1 Device Mapper Multipath Module

The Device Mapper Multipath (DM-MP) module provides the multipathing capability for Linux.
DM-MP is the preferred solution for multipathing on SUSE Linux Enterprise Server 10. It is the
only multipathing option shipped with the product that is completely supported by Novell® and
SUSE.

DM-MP features automatic configuration of the multipathing subsystem for a large variety of
setups. Configurations of up to 8 paths to each device are supported. Configurations are supported
for active/passive (one path active, others passive) or active/active (all paths active with round-robin
load balancing).

The DM-MP framework is extensible in two ways:

Š Using specific hardware handlers. For information, see “Storage Arrays that Require Specific
Hardware Handlers” on page 48.
Š Using more sophisticated load-balancing algorithms than round-robin

The user-space component of DM-MP takes care of automatic path discovery and grouping, as well
as automated path retesting, so that a previously failed path is automatically reinstated when it
becomes healthy again. This minimizes the need for administrator attention in a production
environment.

DM-MP protects against failures in the paths to the device, and not failures in the device itself. If
one of the active paths is lost (for example, a network adapter breaks or a fiber-optic cable is
removed), I/O is redirected to the remaining paths. If the configuration is active/passive, then the
path fails over to one of the passive paths. If you are using the round-robin load-balancing

Managing Multipath I/O for Devices 49

 novdocx (en) 13 May 2009
configuration, the traffic is balanced across the remaining healthy paths. If all active paths fail,
inactive secondary paths must be waked up, so failover occurs with a delay of approximately 30
seconds.

If a disk array has more than one storage processor, make sure that the SAN switch has a connection
to the storage processor that owns the LUNs you want to access. On most disk arrays, all LUNs
belong to both storage processors, so both connections are active.

NOTE: On some disk arrays, the storage array manages the traffic through storage processors so
that it presents only one storage processor at a time. One processor is active and the other one is
passive until there is a failure. If you are connected to the wrong storage processor (the one with the
passive path) you might not see the expected LUNs, or you might see the LUNs but get errors when
trying to access them.

Table 5-1 Multipath I/O Features of Storage Arrays

Features of Storage Arrays Description

Active/passive controllers One controller is active and serves all LUNs. The second controller acts as
a standby. The second controller also presents the LUNs to the multipath
component so that the operating system knows about redundant paths. If
the primary controller fails, the second controller takes over, and it serves
all LUNs.

In some arrays, the LUNs can be assigned to different controllers. A given

LUN is assigned to one controller to be its active controller. One controller
does the disk I/O for any given LUN at a time, and the second controller is
the standby for that LUN. The second controller also presents the paths,
but disk I/O is not possible. Servers that use that LUN are connected to
the LUN’s assigned controller. If the primary controller for a set of LUNs
fails, the second controller takes over, and it serves all LUNs.

Active/active controllers Both controllers share the load for all LUNs, and can process disk I/O for
any given LUN. If one controller fails, the second controller automatically
handles all traffic.

Load balancing The Device Mapper Multipath driver automatically load balances traffic
across all active paths.

Controller failover When the active controller fails over to the passive, or standby, controller,
the Device Mapper Multipath driver automatically activates the paths
between the host and the standby, making them the primary paths.

Boot/Root device support Multipathing is supported for the root (/) device in SUSE Linux Enterprise
Server 10 and later. The host server must be connected to the currently
active controller and storage processor for the boot device. The /boot
partition must be on a separate, non-multipathed partition. Otherwise, no
boot loader is written. For information, see Section 5.8, “Configuring
Multipath I/O for the Root Device,” on page 67.

Device Mapper Multipath detects every path for a multipathed device as a separate SCSI device.
The SCSI device names take the form /dev/sdN, where N is an autogenerated letter for the device,
beginning with a and issued sequentially as the devices are created, such as /dev/sda, /dev/sdb,
and so on. If the number of devices exceeds 26, the letters are duplicated so that the next device after
/dev/sdz will be named /dev/sdaa, /dev/sdab, and so on.

50 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
If multiple paths are not automatically detected, you can configure them manually in the /etc/
multipath.conf file. The multipath.conf file does not exist until you create and configure it.
For information, see Section 5.4.5, “Creating and Configuring the /etc/multipath.conf File,” on
page 55.

5.3.2 Multipath I/O Management Tools

The multipath-tools user-space package takes care of automatic path discovery and grouping. It
automatically tests the path periodically, so that a previously failed path is automatically reinstated
when it becomes healthy again. This minimizes the need for administrator attention in a production
environment.

Table 5-2 Tools in the multipath-tools Package

Tool Description

multipath Scans the system for multipathed devices and assembles them.

multipathd Waits for maps events, then executes multipath.

devmap-name Provides a meaningful device name to udev for device maps (devmaps).

kpartx Maps linear devmaps to partitions on the multipathed device, which makes
it possible to create multipath monitoring for partitions on the device.

For a list of files included in this package, see the multipath-tools Package Description (http://
www.novell.com/products/linuxpackages/suselinux/multipath-tools.html).

1 Ensure that the multipath-tools package is installed by entering the following at a terminal
console prompt:
rpm -q multipath-tools
If it is installed, the response repeats the package name and provides the version information,
such as:
multipath-tools-04.7-34.23
If it is not installed, the response reads:
package multipath-tools is not installed

5.3.3 Using mdadm for Multipathed Devices

In SUSE Linux Enterprise Server 10, Udev is the default device handler, and devices are
automatically known to the system by the Worldwide ID instead of by the device node name. This
resolves problems in previous releases where mdadm.conf and lvm.conf did not properly
recognize multipathed devices.

Just as for LVM2, mdadm requires that the devices be accessed by the ID rather than by the device
node path. Therefore, the DEVICE entry in /etc/mdadm.conf should be set as follows:
DEVICE /dev/disk/by-id/*

This is the default handling in SUSE Linux Enterprise Server 10 and later, as noted above.

Managing Multipath I/O for Devices 51

 novdocx (en) 13 May 2009
To verify that mdadm is installed:

1 Ensure that the mdadm package is installed by entering the following at a terminal console
prompt:
rpm -q mdadm
If it is installed, the response repeats the package name and provides the version information.
For example:
mdadm-2.6-0.11
If it is not installed, the response reads:
package mdadm is not installed

For information about modifying the /etc/lvm/lvm.conf file, see Section 5.2.3, “Using LVM2 on
Multipath Devices,” on page 46.

5.3.4 The Linux multipath(8) Command

Use the Linux multipath(8) command to configure and manage multipathed devices.

General syntax for the multipath(8) command:

multipath [-v verbosity] [-d] [-h|-l|-ll|-f|-F] [-p failover | multibus |
group_by_serial | group_by_prio| group_by_node_name ]

General Examples

Configure multipath devices:

multipath

Configure a specific multipath device:

multipath devicename

Replace devicename with the device node name such as /dev/sdb (as shown by udev in the
$DEVNAME variable), or in the major:minor format.

Selectively suppress a multipath map, and its device-mapped partitions:

multipath -f

Display potential multipath devices, but do not create any devices and do not update device maps
(dry run):
multipath -d

Configure multipath devices and display multipath map information:

multipath -v2 <device>

multipath -v3

The -v2 option in multipath -v2 -d shows only local disks. Use the -v3 option to show the full path
list. For example:
multipath -v3 -d

Display the status of all multipath devices, or a specified multipath device:

52 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
multipath -ll

multipath -ll <device>

Flush all unused multipath device maps (unresolves the multiple paths; it does not delete the
device):
multipath -F

multipath -F <device>

Set the group policy:

multipath -p
[failover|multibus|group_by_serial|group_by_prio|group_by_node_name]

Specify one of the group policy options that are described in Table 5-3 on page 53:

Table 5-3 Group Policy Options for the multipath -p Command

Policy Option Description

failover One path per priority group. You can use only one path at a time.

multibus All paths in one priority group.

group_by_serial One priority group per detected SCSI serial number (the controller node
worldwide number).

group_by_prio One priority group per path priority value. Paths with the same priority are in the
same priority group. Priorities are determined by callout programs specified as
a global, per-controller, or per-multipath option in the /etc/multipath.conf
configuration file.

group_by_node_name One priority group per target node name. Target node names are fetched in the
/sys/class/fc_transport/target*/node_name location.

5.4 Configuring the System for Multipathing

Š Section 5.4.1, “Preparing SAN Devices for Multipathing,” on page 53
Š Section 5.4.2, “Partitioning Multipathed Devices,” on page 54
Š Section 5.4.3, “Configuring the Server for Multipathing,” on page 54
Š Section 5.4.4, “Adding multipathd to the Boot Sequence,” on page 55
Š Section 5.4.5, “Creating and Configuring the /etc/multipath.conf File,” on page 55

5.4.1 Preparing SAN Devices for Multipathing

Before configuring multipath I/O for your SAN devices, prepare the SAN devices, as necessary, by
doing the following:
Š Configure and zone the SAN with the vendor’s tools.
Š Configure permissions for host LUNs on the storage arrays with the vendor’s tools.

Managing Multipath I/O for Devices 53

 novdocx (en) 13 May 2009
Š Install the Linux HBA driver module. Upon module installation, the driver automatically scans
the HBA to discover any SAN devices that have permissions for the host. It presents them to
the host for further configuration.

NOTE: Ensure that the HBA driver you are using does not have native multipathing enabled.

See the vendor’s specific instructions for more details.

Š After the driver module is loaded, discover the device nodes assigned to specific array LUNs or
partitions.

If the LUNs are not seen by the HBA driver, lsscsi can be used to check whether the SCSI devices
are seen correctly by the operating system. When the LUNs are not seen by the HBA driver, check
the zoning setup of the SAN. In particular, check whether LUN masking is active and whether the
LUNs are correctly assigned to the server.

If the LUNs are seen by the HBA driver, but there are no corresponding block devices, additional
kernel parameters are needed to change the SCSI device scanning behavior, such as to indicate that
LUNs are not numbered consecutively. For information, see Options for SCSI Device Scanning
(http://support.novell.com/techcenter/sdb/en/2005/06/drahn_scsi_scanning.html) in the Novell
Support Knowledgebase.

5.4.2 Partitioning Multipathed Devices

Partitioning devices that have multiple paths is not recommended, but it is supported.

SUSE Linux Enterprise Server 10

In SUSE Linux Enterprise Server 10, you can use the kpartx tool to create partitions on
multipathed devices without rebooting. You can also partition the device before you attempt to
configure multipathing by using the Partitioner function in YaST2 or by using a third-party
partitioning tool.

SUSE Linux Enterprise Server 9

In SUSE Linux Enterprise Server 9, if you want to partition the device, you should configure its
partitions before you attempt to configure multipathing by using the Partitioner function in YaST2 or
by using a third-party partitioning tool. This is necessary because partitioning an existing
multipathed device is not supported. Partitioning operations on multipathed devices fail if
attempted.

If you configure partitions for a device, DM-MP automatically recognizes the partitions and
indicates them by appending p1 to pn to the device’s ID, such as
/dev/disk/by-id/26353900f02796769p1

To partition multipathed devices, you must disable the DM-MP service, partition the normal device
node (such as /dev/sdc), then reboot to allow the DM-MP service to see the new partitions.

5.4.3 Configuring the Server for Multipathing

The system must be manually configured to automatically load the device drivers for the controllers
to which the multipath I/O devices are connected within the INITRD. Therefore add the needed
driver module to the variable INITRD_MODULES in the file /etc/sysconfig/kernel.

54 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
For example, if your system contains a RAID controller accessed by the cciss driver and
multipathed devices connected to a QLogic* controller accessed by the driver qla2xxx, this entry
would look like:
INITRD_MODULES="cciss"

Because the QLogic driver is not automatically loaded on start-up, add it here:
INITRD_MODULES="cciss qla23xx"

After having changed /etc/sysconfig/kernel, you must re-create the INITRD on your system
with the command mkinitrd, then reboot in order for the changes to take effect.
When you are using LILO as a boot manager, reinstall it with the command /sbin/lilo. No further
action is required if you are using GRUB.

5.4.4 Adding multipathd to the Boot Sequence

Use either of the methods in this section to add multipath I/O services (multipathd) to the boot
sequence.
Š “YaST” on page 55
Š “Command Line” on page 55

YaST

1 In YaST, click System > System Services (Runlevel) > Simple Mode.
2 Select multipathd, then click Enable.
3 Click OK to acknowledge the service startup message.
4 Click Finish, then click Yes.
The changes do not take affect until the server is restarted.

Command Line

1 Open a terminal console, then log in as the root user or equivalent.

2 At the terminal console prompt, enter
insserv multipathd

5.4.5 Creating and Configuring the /etc/multipath.conf File

The /etc/multipath.conf file does not exist unless you create it. The /usr/share/doc/
packages/multipath-tools/multipath.conf.synthetic file contains a sample /etc/
multipath.conf file that you can use as a guide for multipath settings. See /usr/share/doc/
packages/multipath-tools/multipath.conf.annotated for a template with extensive
comments for each of the attributes and their options.
Š “Creating the multipath.conf File” on page 56
Š “Verifying the Setup in the etc/multipath.conf File” on page 56
Š “Configuring User-Friendly Names or Alias Names in /etc/multipath.conf” on page 57
Š “Blacklisting Non-Multipathed Devices in /etc/multipath.conf” on page 57

Managing Multipath I/O for Devices 55

 novdocx (en) 13 May 2009
Š “Configuring Default Multipath Behavior in /etc/multipath.conf” on page 58
Š “Applying Changes Made to the /etc/multipath.conf File” on page 58

Creating the multipath.conf File

If the /etc/multipath.conf file does not exist, copy the example to create the file:

1 In a terminal console, log in as the root user.

2 Enter the following command (all on one line, of course) to copy the template:
cp /usr/share/doc/packages/multipath-tools/multipath.conf.synthetic /etc/
multipath.conf
3 Use the /usr/share/doc/packages/multipath-tools/multipath.conf.annotated file
as a reference to determine how to configure multipathing for your system.
4 Make sure there is an appropriate device entry for your SAN. Most vendors provide
documentation on the proper setup of the device section.
The /etc/multipath.conf file requires a different device section for different SANs. If you
are using a storage subsystem that is automatically detected (see “Tested Storage Arrays for
Multipathing Support” on page 48), the default entry for that device can be used; no further
configuration of the /etc/multipath.conf file is required.
5 Save the file.

Verifying the Setup in the etc/multipath.conf File

After having set up the configuration, you can perform a “dry run” by entering
multipath -v3 -d

This command scans the devices, then displays what the setup would look like. The output is similar
to the following:
26353900f02796769
[size=127 GB]
[features="0"]
[hwhandler="1 emc"]

\_ round-robin 0 [first]
\_ 1:0:1:2 sdav 66:240 [ready ]
\_ 0:0:1:2 sdr 65:16 [ready ]

\_ round-robin 0
\_ 1:0:0:2 sdag 66:0 [ready ]
\_ 0:0:0:2 sdc 8:32 [ready ]

Paths are grouped into priority groups. Only one priority group is ever in active use. To model an
active/active configuration, all paths end up in the same group. To model active/passive
configuration, the paths that should not be active in parallel are placed in several distinct priority
groups. This normally happens automatically on device discovery.

The output shows the order, the scheduling policy used to balance I/O within the group, and the
paths for each priority group. For each path, its physical address (host:bus:target:lun), device node
name, major:minor number, and state is shown.

56 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Configuring User-Friendly Names or Alias Names in /etc/multipath.conf

A multipath device can be identified by either its WWID or an alias that you assign for it. The
WWID (World Wide Identifier) is an identifier for the multipath device that is guaranteed to be
globally unique and unchanging. The default name used in multipathing is the ID of the logical unit
as found in the /dev/disk/by-id directory. Because device node names in the form of /dev/sdn
and /dev/dm-n can change on reboot, referring to multipath devices by their ID is preferred.
The multipath device names in the /dev/mapper directory reference the ID of the LUN and are
always consistent because they use the /var/lib/multipath/bindings file to track the
association. These device names are user-friendly names such as mpath0.
You can specify your own device names to use via the ALIAS directive in the /etc/
multipath.conf file. Alias names override the use of ID and /dev/mapper/mpathN names.

IMPORTANT: We recommend that you do not use aliases for the root device, because the ability to
seamlessly switch off multipathing via the kernel command line is lost because the device name
differs.

For an example of multipath.conf settings, see the /usr/share/doc/packages/multipath-

tools/multipath.conf.synthetic file.

1 In a terminal console, log in as the root user.

2 Open the /etc/multipath.conf file in a text editor.
3 Uncomment the Defaults directive and its ending bracket.
4 Uncomment the user_friendly_names option, then change its value from No to Yes.
For example:
## Use user friendly names, instead of using WWIDs as names.
defaults {
user_friendly_names yes
}
5 Optionally specify your own user-friendly names for devices using the alias directive in the
multipath section.
For example:
multipath {
wwid 26353900f02796769
alias sdd4l0
}
6 Save your changes, then close the file.

Blacklisting Non-Multipathed Devices in /etc/multipath.conf

The /etc/multipath.conf file should contain a blacklist section where all non-multipathed
devices should be listed. For example, local IDE hard drives and floppy drives are not normally
multipathed. If you have single-path devices that multipath is trying to manage and you want
multipath to ignore them, put them in the blacklist section to resolve the problem.

For example, to blacklist local devices and all arrays from the cciss driver from being managed by
multipath, the blacklist section looks like this:

Managing Multipath I/O for Devices 57

 novdocx (en) 13 May 2009
devnode_blacklist {
wwid 26353900f02796769
devnode "^(ram|raw|loop|fd|md|dm-|sr|scd|st|sda)[0-9]*"
devnode "^hd[a-z][0-9]*"
devnode "^cciss!c[0-9]d[0-9].*"
}

You can also blacklist only the partitions from a driver instead of the entire array. For example, using
the following regular expression would blacklist only partitions from the cciss driver and not the
entire array:
^cciss!c[0-9]d[0-9]*[p[0-9]*]

After you modify the /etc/multipath.conf file, you must run mkinitrd to re-create the INITRD
on your system, then reboot in order for the changes to take effect.
Afterwards, the local devices should no longer be listed in the multipath maps when you issue the
multipath -ll command.

Configuring Default Multipath Behavior in /etc/multipath.conf

The /etc/multipath.conf file should contain a defaults section where you can specify default
behaviors. If the field is not otherwise specified in a device section, the default setting is applied for
that SAN configuration.

The following defaults section specifies a simple failover policy:

defaults {
multipath_tool "/sbin/multipath -v0"
udev_dir /dev
polling_interval 10
default_selector "round-robin 0"
default_path_grouping_policy failover
default_getuid_callout "/sbin/scsi_id -g -u -s /block/%n"
default_prio_callout "/bin/true"
default_features "0"
rr_min_io 100
failback immediate

Applying Changes Made to the /etc/multipath.conf File

Changes to the /etc/multipath.conf file cannot take effect when multipathd is running. After
you make changes, save and close the file, then do the following to apply the changes:

1 Stop the multipathd service.

2 Clear old multipath bindings by entering
/sbin/multipath -F
3 Create new multipath bindings by entering
/sbin/multipath -v2 -l
4 Start the multipathd service.
5 Run mkinitrd to re-create the INITRD on your system, then reboot in order for the changes to
take effect.

58 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
5.5 Enabling and Starting Multipath I/O Services
To start multipath services and enable them to start at reboot:

1 Open a terminal console, then log in as the root user or equivalent.

2 At the terminal console prompt, enter
chkconfig multipathd on
chkconfig boot.multipath on

If the boot.multipath service does not start automatically on system boot, do the following:

1 Open a terminal console, then log in as the root user or equivalent.

2 Enter
/etc/init.d/boot.multipath start
/etc/init.d/multipathd start

5.6 Configuring Path Failover Policies and

Priorities
In a Linux host, when there are multiple paths to a storage controller, each path appears as a separate
block device, and results in multiple block devices for single LUN. The Device Mapper Multipath
service detects multiple paths with the same LUN ID, and creates a new multipath device with that
ID. For example, a host with two HBAs attached to a storage controller with two ports via a single
unzoned Fibre Channel switch sees four block devices: /dev/sda, /dev/sdb, /dev/sdc, and /
dev/sdd. The Device Mapper Multipath service creates a single block device, /dev/mpath/
mpath1 that reroutes I/O through those four underlying block devices.

This section describes how to specify policies for failover and configure priorities for the paths.

Š Section 5.6.1, “Configuring the Path Failover Policies,” on page 59

Š Section 5.6.2, “Configuring Failover Priorities,” on page 60
Š Section 5.6.3, “Using a Script to Set Path Priorities,” on page 65
Š Section 5.6.4, “Configuring ALUA,” on page 65
Š Section 5.6.5, “Reporting Target Path Groups,” on page 67

5.6.1 Configuring the Path Failover Policies

Use the multipath command with the -p option to set the path failover policy:
multipath devicename -p policy

Replace policy with one of the following policy options:

Table 5-4 Group Policy Options for the multipath -p Command

Policy Option Description

failover One path per priority group.

Managing Multipath I/O for Devices 59

 novdocx (en) 13 May 2009
Policy Option Description

multibus All paths in one priority group.

group_by_serial One priority group per detected serial number.

group_by_prio One priority group per path priority value. Priorities are determined by callout
programs specified as a global, per-controller, or per-multipath option in the /
etc/multipath.conf configuration file.
group_by_node_name One priority group per target node name. Target node names are fetched in the
/sys/class/fc_transport/target*/node_name location.

5.6.2 Configuring Failover Priorities

You must manually enter the failover priorities for the device in the /etc/multipath.conf file.
Examples for all settings and options can be found in the /usr/share/doc/packages/
multipath-tools/multipath.conf.annotated file.

Š “Understanding Priority Groups and Attributes” on page 60

Š “Configuring for Round-Robin Load Balancing” on page 64
Š “Configuring for Single Path Failover” on page 64
Š “Grouping I/O Paths for Round-Robin Load Balancing” on page 64

Understanding Priority Groups and Attributes

A priority group is a collection of paths that go to the same physical LUN. By default, I/O is
distributed in a round-robin fashion across all paths in the group. The multipath command
automatically creates priority groups for each LUN in the SAN based on the path_grouping_policy
setting for that SAN. The multipath command multiplies the number of paths in a group by the
group’s priority to determine which group is the primary. The group with the highest calculated
value is the primary. When all paths in the primary group are failed, the priority group with the next
highest value becomes active.

A path priority is an integer value assigned to a path. The higher the value, the higher is the priority.
An external program is used to assign priorities for each path. For a given device, its paths with the
same priorities belong to the same priority group.

Table 5-5 Multipath Attributes

Multipath Attribute Description Values

user_friendly_names Specifies whether to use IDs yes Autogenerate user-friendly names as

or to use the /var/lib/ aliases for the multipath devices instead of the
multipath/bindings file actual ID.
to assign a persistent and
unique alias to the multipath no Default. Use the WWIDs shown in the /dev/
devices in the form of /dev/ disk/by-id/ location.
mapper/mpathN.

60 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Multipath Attribute Description Values

blacklist Specifies the list of device For an example, see “Blacklisting Non-
names to ignore as non- Multipathed Devices in /etc/multipath.conf” on
multipathed devices, such as page 57.
cciss, fd, hd, md, dm, sr, scd,
st, ram, raw, loop.

blacklist_exceptions Specifies the list of device For an example, see the /usr/share/doc/
names to treat as multipath packages/multipath-tools/
devices even if they are multipath.conf.annotated file.
included in the blacklist.

getuid_callout The default program and /lib/udev/scsi_id -g -u -s

argumentss to callout to
obtain a unique path identifier. This is the default location and arguments.
Should be specified with an
absolute path.

path_grouping_policy Specifies the path grouping failover One path is assigned per priority group
policy for a multipath device so that only one path at a time is used.
hosted by a given controller.
multibus (Default) All valid paths are in one
priority group. Traffic is load-balanced across all
active paths in the group.

group_by_prio One priority group exists for

each path priority value. Paths with the same
priority are in the same priority group. Priorities
are assigned by an external program.

group_by_serial Paths are grouped by the

SCSI target serial number (controller node
WWN).

group_by_node_name One priority group is

assigned per target node name. Target node
names are fetched in /sys/class/
fc_transport/target*/node_name.

Managing Multipath I/O for Devices 61

 novdocx (en) 13 May 2009
Multipath Attribute Description Values

path_checker Determines the state of the directio (Default in multipath-tools version

path. 0.4.8 and later) Reads the first sector that has
direct I/O. This is useful for DASD devices. Logs
failure messages in /var/log/messages.

readsector0 (Default in multipath-tools

version 0.4.7 and earlier) Reads the first sector
of the device. Logs failure messages in /var/
log/messages.

tur Issues a SCSI test unit ready command to

the device. This is the preferred setting if the
LUN supports it. The command does not fill up /
var/log/messages on failure with messages.

Some SAN vendors provide custom

path_checker options:

Š emc_clariion Queries the EMC Clariion

EVPD page 0xC0 to determine the path
state.
Š hp_sw Checks the path state (Up, Down, or
Ghost) for HP storage arrays with Active/
Standby firmware.
Š rdac Checks the path state for the LSI/
Engenio RDAC storage controller.

path_selector Specifies the path-selector round-robin 0 (Default) The load-balancing

algorithm to use for load- algorithm used to balance traffic across all active
balancing. paths in a priority group.

This is currently the only algorithm available.

pg_timeout Specifies path group timeout NONE (internal default)

handling.

62 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Multipath Attribute Description Values

prio_callout Specifies the program and If no prio_callout attribute is used, all paths
arguments to use to are equal. This is the default.
determine the layout of the
multipath map. /bin/true Use this value when the
group_by_priority is not being used.
When queried by the
multipath command, the The prioritizer programs generate path priorities
specified mpath_prio_* when queried by the multipath command. The
callout program returns the program names must begin with mpath_prio_
priority for a given path in and are named by the device type or balancing
relation to the entire multipath method used. Current prioritizer programs
layout. include the following:

When it is used with the /sbin/mpath_prio_alua %n Generates path

path_grouping_policy of priorities based on the SCSI-3 ALUA settings.
group_by_prio, all paths with
/sbin/mpath_prio_balance_units Generates
the same priority are grouped
the same priority for all paths.
into one multipath group. The
group with the highest /sbin/mpath_prio_emc %n Generates the path
aggregate priority becomes priority for EMC arrays.
the active group.
/sbin/mpath_prio_hds_modular %b
When all paths in a group fail,
the group with the next Generates the path priority for Hitachi HDS
highest aggregate priority Modular storage arrays.
becomes active. Additionally,
a failover command (as /sbin/mpath_prio_hp_sw %n Generates the
determined by the hardware path priority for Compaq/HP controller in active/
handler) might be send to the standby mode.
target. /sbin/mpath_prio_netapp %n Generates the
The mpath_prio_* program path priority for NetApp arrays.
can also be a custom script /sbin/mpath_prio_random %n Generates a
created by a vendor or random priority for each path.
administrator for a specified
setup. /sbin/mpath_prio_rdac %n Generates the path
priority for LSI/Engenio RDAC controller.
A %n in the command line
expands to the device name /sbin/mpath_prio_tpc %n You can optionally
in the /dev directory. use a script created by a vendor or administrator
that gets the priorities from a file where you
A %b expands to the device specify priorities to use for each path.
number in major:minor format
in the /dev directory. /usr/local/sbin/mpath_prio_spec.sh %n

A %d expands to the device Provides the path of a user-created script that

ID in the /dev/disk/by-id generates the priorities for multipathing based on
directory. information contained in a second data file. (This
path and filename are provided as an example.
If devices are hot-pluggable, Specify the location of your script instead.) The
use the %d flag instead of script can be created by a vendor or
%n. This addresses the short administrator. The script’s target file identifies
time that elapses between the each path for all multipathed devices and
time when devices are specifies a priority for each path. For an
available and when udev example, see Section 5.6.3, “Using a Script to
creates the device nodes. Set Path Priorities,” on page 65.

Managing Multipath I/O for Devices 63

 novdocx (en) 13 May 2009
Multipath Attribute Description Values

rr_min_io Specifies the number of I/O n (>0) Specify an integer value greater than 0.
transactions to route to a path
before switching to the next 1000 Default.
path in the same path group,
as determined by the
specified algorithm in the
path_selector setting.
rr_weight Specifies the weighting uniform Default. All paths have the same
method to use for paths. round-robin weightings.

priorities Each path’s weighting is determined

by the path’s priority times the rr_min_io setting.

no_path_retry Specifies the behaviors to use n (> 0) Specifies the number of retries until
on path failure. multipath stops the queuing and fails the path.
Specify an integer value greater than 0.

fail Specified immediate failure (no queuing).

queue Never stop queuing (queue forever until

the path comes alive).

failback Specifies whether to monitor immediate When a path recovers, enable the
the failed path recovery, and path immediately.
indicates the timing for group
failback after failed paths n (> 0) When the path recovers, wait n seconds
return to service. before enabling the path. Specify an integer
value greater than 0.
When the failed path
recovers, the path is added manual (Default) The failed path is not
back into the multipath monitored for recovery. The administrator runs
enabled path list based on the multipath command to update enabled
this setting. Multipath paths and priority groups.
evaluates the priority groups,
and changes the active
priority group when the
priority of the primary path
exceeds the secondary
group.

Configuring for Round-Robin Load Balancing

All paths are active. I/O is configured for some number of seconds or some number of I/O
transactions before moving to the next open path in the sequence.

Configuring for Single Path Failover

A single path with the highest priority (lowest value setting) is active for traffic. Other paths are
available for failover, but are not used unless failover occurs.

Grouping I/O Paths for Round-Robin Load Balancing

Multiple paths with the same priority fall into the active group. When all paths in that group fail, the
device fails over to the next highest priority group. All paths in the group share the traffic load in a
round-robin load balancing fashion.

64 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
5.6.3 Using a Script to Set Path Priorities
You can create a script that interacts with DM-MP to provide priorities for paths to the LUN when
set as a resource for the prio_callout setting.
First, set up a text file that lists information about each device and the priority values you want to
assign to each path. For example, name the file /usr/local/etc/primary-paths. Enter one line
for each path in the following format:
host_wwpn target_wwpn scsi_id priority_value

Return a priority value for each path on the device. Make sure that the variable
FILE_PRIMARY_PATHS resolves to a real file with appropriate data (host wwpn, target wwpn,
scsi_id and priority value) for each device.

The contents of the primary-paths file for a single LUN with eight paths each might look like this:
0x10000000c95ebeb4 0x200200a0b8122c6e 2:0:0:0 sdb
3600a0b8000122c6d00000000453174fc 50

0x10000000c95ebeb4 0x200200a0b8122c6e 2:0:0:1 sdc

3600a0b80000fd6320000000045317563 2

0x10000000c95ebeb4 0x200200a0b8122c6e 2:0:0:2 sdd

3600a0b8000122c6d0000000345317524 50

0x10000000c95ebeb4 0x200200a0b8122c6e 2:0:0:3 sde

3600a0b80000fd6320000000245317593 2

0x10000000c95ebeb4 0x200300a0b8122c6e 2:0:1:0 sdi

3600a0b8000122c6d00000000453174fc 5

0x10000000c95ebeb4 0x200300a0b8122c6e 2:0:1:1 sdj

3600a0b80000fd6320000000045317563 51

0x10000000c95ebeb4 0x200300a0b8122c6e 2:0:1:2 sdk

3600a0b8000122c6d0000000345317524 5

0x10000000c95ebeb4 0x200300a0b8122c6e 2:0:1:3 sdl

3600a0b80000fd6320000000245317593 51

To continue the example mentioned in Table 5-5 on page 60, create a script named /usr/local/
sbin/path_prio.sh. You can use any path and filename. The script does the following:

Š On query from multipath, grep the device and its path from the /usr/local/etc/primary-
paths file.
Š Return to multipath the priority value in the last column for that entry in the file.

5.6.4 Configuring ALUA

The mpath_prio_alua(8) command is used as a priority callout for the Linux multipath(8)
command. It returns a number that is used by DM-MP to group SCSI devices with the same priority
together. This path priority tool is based on ALUA (Asynchronous Logical Unit Access).
Š “Syntax” on page 66
Š “Prerequisite” on page 66

Managing Multipath I/O for Devices 65

 novdocx (en) 13 May 2009
Š “Options” on page 66
Š “Return Values” on page 66

Syntax
mpath_prio_alua [-d directory] [-h] [-v] [-V] device [device...]

Prerequisite

SCSI devices

Options

-d directory
Specifying the Linux directory path where the listed device node names can be found. The
default directory is /dev. When used, specify the device node name only (such as sda) for the
device or devices you want to manage.

-h
Displays help for this command, then exits.

-v
Turns on verbose output to display status in human-readable format. Output includes
information about which port group the specified device is in and its current state.

-V
Displays the version number of this tool, then exits.

device
Specifies the SCSI device you want to manage. The device must be a SCSI device that supports
the Report Target Port Groups (sg_rtpg(8)) command. Use one of the following formats for
the device node name:
Š The full Linux directory path, such as /dev/sda. Do not use with the -d option.
Š The device node name only, such as sda. Specify the directory path using the -d option.
Š The major and minor number of the device separated by a colon (:) with no spaces, such as
8:0. This creates a temporary device node in the /dev directory with a name in the format
of tmpdev-<major>:<minor>-<pid>. For example, /dev/tmpdev-8:0-<pid>.

Return Values

On success, returns a value of 0 and the priority value for the group. Table 5-6 shows the priority
values returned by the mpath_prio_alua command.

Table 5-6 ALUA Priorities for Device Mapper Multipath

Priority Value Description

50 The device is in the active, optimized group.

10 The device is in an active but non-optimized group.

66 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Priority Value Description

1 The device is in the standby group.

0 All other groups.

Values are widely spaced because of the way the multipath command handles them. It multiplies
the number of paths in a group with the priority value for the group, then selects the group with the
highest result. For example, if a non-optimized path group has six paths (6 x 10 = 60) and the
optimized path group has a single path (1 x 50 = 50), the non-optimized group has the highest score,
so multipath chooses the non-optimized group. Traffic to the device uses all six paths in the group in
a round-robin fashion.

On failure, returns a value of 1 to 5 indicating the cause for the command’s failure. For information,
see the man page for mpath_prio_alua.

5.6.5 Reporting Target Path Groups

Use the SCSI Report Target Port Groups (sg_rtpg(8)) command. For information, see the man
page for sg_rtpg(8).

5.7 Tuning the Failover for Specific Host Bus

Adapters
When using multipath I/O, you want any host bus adapter (HBA) failure or cable failures to be
reported faster than when multipathing is not in use. Configure time-out settings for your HBA to
disable failover at the HBA level to allow the failure to propogate up to the multipath I/O layer as
fast as possible where the I/O can be redirected to another, healthy path.

To disable the HBA handling of failover, modify the driver’s options in the /etc/
modprobe.conf.local file. Refer to the HBA vendor’s documentation for information about how
to disable failover settings for your driver.

For example, for the QLogic qla2xxx family of host bus adapters, the following setting is
recommended:
options qla2xxx qlport_down_retry=1

5.8 Configuring Multipath I/O for the Root Device

In the SUSE Linux Enterprise Server 10, the root partition (/) on multipath is supported only if the /
boot partition is on a separate, non-multipathed partition. Otherwise, no boot loader is written.

IMPORTANT: If you apply all online updates, the DM-MP is available but is not supported for /
boot and /root in SUSE Linux Enterprise Server 10 SP1 and later. More specifically, you need
mkinitrd 1.2-106.61 and multipath-tools 0.4.7-34.23 or later. However, if you install the
packages and set up the configuration, you might run into update issues later.
Full multipath support is available in SUSE Linux Enterprise Server 11.

Managing Multipath I/O for Devices 67

 novdocx (en) 13 May 2009
To enable multipathing on the existing root device:

1 Install Linux with only a single path active, preferably one where the by-id symlinks are listed
in the partitioner.
2 Mount the devices by using the /dev/disk/by-id path used during the install.
3 After installation, add dm-multipath to /etc/sysconfig/kernel:INITRD_MODULES.
4 Re-run /sbin/mkinitrd to update the initrd image.
5 Reboot the server.

To disable multipathing on the root device:

1 Add multipath=off to the kernel command line.

This affects only the root device. All other devices are not affected.

5.9 Configuring Multipath I/O for an Existing

Software RAID
Ideally, you should configure multipathing for devices before you use them as components of a
software RAID device. If you add multipathing after creating any software RAID devices, the DM-
MP service might be starting after the multipath service on reboot, which makes multipathing
appear not to be available for RAIDs. You can use the procedure in this section to get multipathing
running for a previously existing software RAID.

For example, you might need to configure multipathing for devices in a software RAID under the
following circumstances:
Š If you create a new software RAID as part of the Partitioning settings during a new install or
upgrade.
Š If you did not configure the devices for multipathing before using them in the software RAID
as a member device or spare.
Š If you grow your system by adding new HBA adapters to the server or expanding the storage
subsystem in your SAN.

NOTE: The following instructions assume the software RAID device is /dev/mapper/mpath0,
which is its device name as recognized by the kernel. Make sure to modify the instructions for the
device name of your software RAID.

1 Open a terminal console, then log in as the root user or equivalent.

Except where otherwise directed, use this console to enter the commands in the following
steps.
2 If any software RAID devices are currently mounted or running, enter the following commands
for each device to dismount the device and stop it.
umount /dev/mapper/mpath0
mdadm --misc --stop /dev/mapper/mpath0
3 Stop the boot.md service by entering
/etc/init.d/boot.md stop
4 Start the boot.multipath and multipathd services by entering the following commands:

68 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
/etc/init.d/boot.multipath start
/etc/init.s/multipathd start
5 After the multipathing services are started, verify that the software RAID’s component devices
are listed in the /dev/disk/by-id directory. Do one of the following:
Š Devices Are Listed: The device names should now have symbolic links to their Device
Mapper Multipath device names, such as /dev/dm-1.
Š Devices Are Not Listed: Force the multipath service to recognize them by flushing and
rediscovering the devices.
To do this, enter the following commands:
multipath -F
multipath -v0
The devices should now be listed in /dev/disk/by-id, and have symbolic links to their
Device Mapper Multipath device names. For example:
lrwxrwxrwx 1 root root 10 Jun 15 09:36 scsi-mpath1 -> ../../dm-1
6 Restart the boot.md service and the RAID device by entering
/etc/init.d/boot.md start
7 Check the status of the software RAID by entering
mdadm --detail /dev/mapper/mpath0
The RAID’s component devices should match their Device Mapper Multipath device names
that are listed as the symbolic links of devices in the /dev/disk/by-id directory.
8 Make a new initrd to ensure that the Device Mapper Multipath services are loaded before the
RAID services on reboot. Enter
mkinitrd -f mpath
9 Reboot the server to apply these post-install configuration settings.
10 Verify that the software RAID array comes up properly on top of the multipathed devices by
checking the RAID status. Enter
mdadm --detail /dev/mapper/mpath0
For example:
Number Major Minor RaidDevice State
0 253 0 0 active sync /dev/dm-0
1 253 1 1 active sync /dev/dm-1
2 253 2 2 active sync /dev/dm-2

5.10 Scanning for New Devices without

Rebooting
If your system has already been configured for multipathing and you later need to add more storage
to the SAN, use the following procedure to scan the devices and make them available to
multipathing without rebooting the system.

1 On the storage subsystem, use the vendor’s tools to allocate the device and update its access
control settings to allow the Linux system access to the new storage. Refer to the vendor’s
documentation for details.

Managing Multipath I/O for Devices 69

 novdocx (en) 13 May 2009
2 On the Linux system, scan the SAN at a low level to discover the new devices. At a terminal
console prompt, enter
echo 1 > /sys/class/fc_host/host<number>/issue_lip
For example, to probe the HBA on host1, enter
echo 1 > /sys/class/fc_host/host1/issue_lip
At this point, the newly added device is not known to the higher layers of the Linux kernel's
SCSI subsystem and is not yet usable.
3 Scan all targets for a host to make its new device known to the middle layer of the Linux
kernel's SCSI subsystem. At a terminal console prompt, enter
echo "- - -" > /sys/class/scsi_host/host<number>/scan
For example, to probe the HBA on host1, enter
echo "- - -" > /sys/class/scsi_host/host1/scan
4 Check for scanning progress in the system log (the /var/log/messages file). At a terminal
console prompt, enter
tail -30 /var/log/messages
This command displays the last 30 lines of the log. For example:
# tail -30 /var/log/messages
. . .
Feb 14 01:03 kernel: SCSI device sde: 81920000
Feb 14 01:03 kernel: SCSI device sdf: 81920000
Feb 14 01:03 multipathd: sde: path checker registered
Feb 14 01:03 multipathd: sdf: path checker registered
Feb 14 01:03 multipathd: mpath4: event checker started
Feb 14 01:03 multipathd: mpath5: event checker started
Feb 14 01:03:multipathd: mpath4: remaining active paths: 1
Feb 14 01:03 multipathd: mpath5: remaining active paths: 1
5 Repeat Step 2 through Step 4 to add paths through other HBA adapters on the Linux system
that are connected to the new device.
6 Run the multipath command to recognize the devices for DM-MP configuration. At a
terminal console prompt, enter
multipath
You can now configure the new device for multipathing.

5.11 Scanning for New Partitioned Devices

without Rebooting
Use the example in this section to detect a newly added multipathed LUN without rebooting.

1 Open a terminal console, then log in as the root user.

2 Detect the newly added multipathed LUN by entering:
echo '- - -' >>/sys/class/fc_host/host1/device/scsi_host\:host1/scan
echo '- - -' >>/sys/class/fc_host/host2/device/scsi_host\:host2/scan
3 Verify that the device is seen (link has a fresh timestamp) by entering
ls -lrt /dev/dm-*
4 Verify the new WWN of the device appears in the log by entering

70 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
tail -33 /var/log/messages
5 Use a text editor to add a new alias definition for the device in the /etc/multipath.conf file,
such as oradata3.
6 Create a partition table for the device by entering
fdisk /dev/dm-8
7 Add a /dev/dm-* link for the new partition by entering
/sbin/kpartx -a -p -part /dev/dm-8
8 Verify that the link was created by entering
ls -lrt /dev/dm-*
9 Create a filesystem and label for the new partition by entering
mke2fs -j /dev/dm-9
tune2fs -L oradata3 /dev/dm-9
10 Restart DM-Multipath to let it read the aliases by entering
/etc/init.d/multipathd restart
11 Verify that the device is recognized by multipathd by entering
multipath -ll
12 Use a text editor to add a mount entry in the /etc/fstab file.
At this point, the alias you created in Step 5 is not yet in the /dev/disk/by-label directory.
Add the mount entry the /dev/dm-9 path, then change the entry before the next time you
reboot to
LABEL=oradata3
13 Create a directory to use as the mount point, then mount the device by entering
md /oradata3
mount /oradata3

5.12 Viewing Multipath I/O Status

Querying the multipath I/O status outputs the current status of the multipath maps.

The multipath -l option displays the current path status as of the last time that the path checker
was run. It does not run the path checker.

The multipath -ll option runs the path checker, updates the path information, then displays the
current status information. This option always the displays the latest information about the path
status.

1 At a terminal console prompt, enter

multipath -ll
This displays information for each multipathed device. For example:
3600601607cf30e00184589a37a31d911
[size=127 GB][features="0"][hwhandler="1 emc"]

Managing Multipath I/O for Devices 71

 novdocx (en) 13 May 2009
\_ round-robin 0 [active][first]
\_ 1:0:1:2 sdav 66:240 [ready ][active]
\_ 0:0:1:2 sdr 65:16 [ready ][active]
\_ round-robin 0 [enabled]
\_ 1:0:0:2 sdag 66:0 [ready ][active]
\_ 0:0:0:2 sdc 8:32 [ready ][active]

For each device, it shows the device’s ID, size, features, and hardware handlers.

Paths to the device are automatically grouped into priority groups on device discovery. Only one
priority group is active at a time. For an active/active configuration, all paths are in the same group.
For an active/passive configuration, the passive paths are placed in separate priority groups.

The following information is displayed for each group:

Š Scheduling policy used to balance I/O within the group, such as round-robin
Š Whether the group is active, disabled, or enabled
Š Whether the group is the first (highest priority) group
Š Paths contained within the group

The following information is displayed for each path:

Š The physical address as host:bus:target:lun, such as 1:0:1:2

Š Device node name, such as sda
Š Major:minor numbers
Š Status of the device

5.13 Managing I/O in Error Situations

You might need to configure multipathing to queue I/O if all paths fail concurrently. In certain
scenarios, where the driver, the HBA, or the fabric experiences spurious errors, it is advisable that
DM-MP be configured to queue all I/O where those errors lead to a loss of all paths, and never
propagate errors upwards. Because this leads to I/O being queued indefinitely unless a path is
reinstated, make sure that multipathd is running and works for your scenario. Otherwise, I/O
might be stalled indefinitely on the affected multipathed device, until reboot or until you manually
return to failover instead of queuing.

To test the scenario:

1 In a terminal console, log in as the root user.

2 Activate queuing instead of failover for the device I/O by entering:
dmsetup message device_ID 0 queue_if_no_path
Replace the device_ID with the ID for your device. For example, enter:
dmsetup message 3600601607cf30e00184589a37a31d911 0 queue_if_no_path
3 Return to failover for the device I/O by entering:
dmsetup message device_ID 0 fail_if_no_path
This command immediately causes all queued I/O to fail.
Replace the device_ID with the ID for your device. For example, enter:
dmsetup message 3600601607cf30e00184589a37a31d911 0 fail_if_no_path

72 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
To set up queuing I/O for scenarios where all paths fail:

1 In a terminal console, log in as the root user.

2 Open the /etc/multipath.conf file in a text editor.
3 Uncomment the defaults section and its ending bracket, then add the default_features
setting, as follows:
defaults {
default_features "1 queue_if_no_path"
}
4 After you modify the /etc/multipath.conf file, you must run mkinitrd to re-create the
INITRD on your system, then reboot in order for the changes to take effect.
5 When you are ready to return over to failover for the device I/O, enter:
dmsetup message mapname 0 fail_if_no_path
Replace the mapname with the mapped alias name or the device ID for the device.
This command immediately causes all queued I/O to fail and propagates the error to the calling
application.

5.14 Resolving Stalled I/O

If all paths fail concurrently and I/O is queued and stalled, do the following:

1 Enter the following command at a terminal console prompt:

dmsetup message mapname 0 fail_if_no_path
Replace mapname with the correct device ID or mapped alias name for the device. This causes
all queued I/O to fail and propagates the error to the calling application.
2 Reactivate queueing by entering the following command at a terminal console prompt:
dmsetup message mapname 0 queue_if_no_path

5.15 Additional Information

For more information about configuring and using multipath I/O on SUSE Linux Enterprise Server,
see the following additional resources in the Novell Support Knowledgebase:
Š How to Setup/Use Multipathing on SLES (http://support.novell.com/techcenter/sdb/en/2005/
04/sles_multipathing.html)
Š Troubleshooting SLES Multipathing (MPIO) Problems (Technical Information Document
3231766) (http://www.novell.com/support/
search.do?cmd=displayKC&docType=kc&externalId=3231766&sliceId=SAL_Public)
Š Dynamically Adding Storage for Use with Multipath I/O (Technical Information Document
3000817) (https://secure-support.novell.com/KanisaPlatform/Publishing/911/
3000817_f.SAL_Public.html)
Š DM MPIO Device Blacklisting Not Honored in multipath.conf (Technical Information
Document 3029706) (http://www.novell.com/support/
search.do?cmd=displayKC&docType=kc&externalId=3029706&sliceId=SAL_Public&dialogI
D=57872426&stateId=0%200%2057878058)

Managing Multipath I/O for Devices 73

 novdocx (en) 13 May 2009
Š Static Load Balancing in Device-Mapper Multipathing (DM-MP) (Technical Information
Document 3858277) (http://www.novell.com/support/
search.do?cmd=displayKC&docType=kc&externalId=3858277&sliceId=SAL_Public&dialogI
D=57872426&stateId=0%200%2057878058)
Š Troubleshooting SCSI (LUN) Scanning Issues (Technical Information Document 3955167)
(http://www.novell.com/support/
search.do?cmd=displayKC&docType=kc&externalId=3955167&sliceId=SAL_Public&dialogI
D=57868704&stateId=0%200%2057878206)

5.16 What’s Next

If you want to use software RAIDs, create and configure them before you create file systems on the
devices. For information, see the following:
Š Chapter 6, “Managing Software RAIDs with EVMS,” on page 75
Š Chapter 7, “Managing Software RAIDs 6 and 10 with mdadm,” on page 97

74 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Managing Software RAIDs with
6

EVMS 6
This section describes how to create and manage software RAIDs with the Enterprise Volume
Management System (EVMS). EVMS supports only RAIDs 0, 1, 4, and 5 at this time. For RAID 6
and 10 solutions, see Chapter 7, “Managing Software RAIDs 6 and 10 with mdadm,” on page 97.
Š Section 6.1, “Understanding Software RAIDs on Linux,” on page 75
Š Section 6.2, “Creating and Configuring a Software RAID,” on page 81
Š Section 6.3, “Expanding a RAID,” on page 85
Š Section 6.4, “Adding or Removing a Spare Disk,” on page 86
Š Section 6.5, “Managing Disk Failure and RAID Recovery,” on page 87
Š Section 6.6, “Monitoring Status for a RAID,” on page 90
Š Section 6.7, “Deleting a Software RAID and Its Data,” on page 95

6.1 Understanding Software RAIDs on Linux

Š Section 6.1.1, “What Is a Software RAID?,” on page 75
Š Section 6.1.2, “Overview of RAID Levels,” on page 76
Š Section 6.1.3, “Comparison of RAID Performance,” on page 77
Š Section 6.1.4, “Comparison of Disk Fault Tolerance,” on page 77
Š Section 6.1.5, “Configuration Options for RAIDs,” on page 78
Š Section 6.1.6, “Interoperability Issues,” on page 78
Š Section 6.1.7, “Guidelines for Component Devices,” on page 79
Š Section 6.1.8, “RAID 5 Algorithms for Distributing Stripes and Parity,” on page 80
Š Section 6.1.9, “Multi-Disk Plug-In for EVMS,” on page 81
Š Section 6.1.10, “Device Mapper Plug-In for EVMS,” on page 81

6.1.1 What Is a Software RAID?

A RAID combines multiple devices into a multi-disk array to provide resiliency in the storage
device and to improve storage capacity and I/O performance. If a disk fails, some RAID levels keep
data available in a degraded mode until the failed disk can be replaced and its content reconstructed.

A software RAID provides the same high availability that you find in a hardware RAID. The key
operational differences are described in the following table:

Table 6-1 Comparison of Software RAIDs and Hardware RAIDs

Feature Linux Software RAID Hardware RAID

RAID function Multi-disk (md) driver or mdadm RAID controller on the disk array

Managing Software RAIDs with EVMS 75

 novdocx (en) 13 May 2009
Feature Linux Software RAID Hardware RAID

RAID processing In the host server’s processor RAID controller on the disk array

RAID levels 0, 1, 4, 5, and 10 plus the mdadm Varies by vendor

raid10

Component devices Disks from same or different disk Same disk array
array

6.1.2 Overview of RAID Levels

The following table describes the advantages and disadvantages of the RAID levels supported by
EVMS. The description assumes that the component devices reside on different disks and that each
disk has its own dedicated I/O capability.

IMPORTANT: For information about creating complex or nested RAID devices with mdadm, see
Chapter 7, “Managing Software RAIDs 6 and 10 with mdadm,” on page 97.

Table 6-2 RAID Levels Supported by EVMS

RAID Level Description Performance and Fault Tolerance

0 Stripes data using a round- Improves disk I/O performance for both reads and writes.
robin method to distribute Actual performance depends on the stripe size, the actual
data over the RAID’s data, and the application.
component devices.
Does not provide disk fault tolerance and data redundancy.
Any disk failure causes all data in the RAID to be lost.

1 Mirrors data by copying Improves disk reads by making multiple copies of data
blocks of one disk to another available via different I/O paths. The write performance is
and keeping them in about the same as for a single disk because a copy of the
continuous synchronization. data must be written to each of the disks in the mirror.
If disks are different sizes,
the smallest disk determines Provides 100% data redundancy. If one disk fails then the
the size of the RAID. data remains available on its mirror, and processing
continues.

4 Stripes data and records Improves disk I/O performance for both reads and writes.
parity to a dedicated disk. If Write performance is considerably slower than for RAID 0,
disks are different sizes, the because parity must be calculated and written. Write
smallest disk determines the performance is slightly slower than RAID 5. Read
size of the RAID. performance is slower than for a RAID 1 array with the same
number of component devices. The dedicated parity disk can
become a bottleneck for writing parity.

Provides disk fault tolerance. If a disk fails, performance is

degraded while the RAID uses the parity to reconstruct data
for the replacement disk.

76 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
RAID Level Description Performance and Fault Tolerance

5 Stripes data and distributes Improves disk I/O performance for reads and writes. Write
parity in a round-robin performance is considerably less than for RAID 0, because
fashion across all disks. If parity must be calculated and written. Write performance is
disks are different sizes, the faster than RAID 4. Read performance is slower than for a
smallest disk determines the RAID 1 array with the same number of component disks.
size of the RAID. Actual performance depends on the number of component
disks, the stripe size, the actual data, and the application.

Provides disk fault tolerance. If a disk fails, performance is

degraded while the RAID uses the parity to reconstruct data
for the replacement disk. Provides slightly less data
redundancy than mirroring because it uses parity to
reconstruct the data.

6.1.3 Comparison of RAID Performance

The following table compares the read and write performance for RAID devices.

Table 6-3 Read and Write Performance for RAIDs

Raid Level Read Performance Write Performance

0 Faster than for a single disk Faster than for a single disk and other
RAIDs.

1 Faster than for a single disk, increasing as Slower than for a single disk, declining as
more mirrors are added more mirrors are added.

4 Faster than for a single disk. Slower than a Faster than for a single disk. Slower than a
RAID 0 because one disk is used for parity. RAID 0 because of writes for parity. Slower
than a RAID 5 because of possible
bottlenecks for writes of parity to the
dedicated parity disk.

5 Faster than for a single disk; comparable to a Faster than a single disk. Slower than a
RAID 0. RAID 0 because of writes for parity.

6.1.4 Comparison of Disk Fault Tolerance

The following table compares the disk fault tolerance for RAID devices.

Table 6-4 Fault Tolerance for RAIDs

Raid Level Number of Disk Failures Tolerated Data Redundancy

0 None No

1 Number of disks minus 1 100% redundancy for each mirror

4 1 Dedicated parity disk to reconstruct data. If

the parity disk fails, all parity must be
recalculated.

Managing Software RAIDs with EVMS 77

 novdocx (en) 13 May 2009
Raid Level Number of Disk Failures Tolerated Data Redundancy

5 1 Distributed parity to reconstruct data and

parity on the failed disk.

6.1.5 Configuration Options for RAIDs

In EVMS management tools, the following RAID configuration options are provided:

Table 6-5 Configuration Options in EVMS

Option Description

Spare Disk For RAIDs 1, 4, and 5, you can optionally specify a device, segment, or
region to use as the replacement for a failed disk (the member device,
segment, or region). On failure, the spare disk automatically replaces the
failed disk, then reconstructs the data.

However, if the parity disk fails on a RAID 5, parity cannot be reconstructed.

Chunk Size (KB) For RAIDs 0, 4, or 5, specify the stripe size in KB.

Consider the intended use of the RAID, such as the file system block size,
the applications used, and the actual data (file sizes and typical reads and
writes). A typical write size for large files is 128 KB.

Default: 32 KB

Range: 4 KB to 4096 KB, in powers of 2.

RAID Level If you selected MD RAID 4/5 Region Manager, specify RAID 4 or RAID 5
(default).

RAID Algorithm For RAID 5, specify one of the following algorithms to use for striping and
distributing parity on the disk.

Š Left Asymmetric
Š Left Symmetric (Default, fastest performance for large reads)
Š Right Asymmetric
Š Right Symmetric

6.1.6 Interoperability Issues

Linux software RAID cannot be used underneath clustered file systems because it does not support
concurrent activation. If you want RAID and OCFS2, you need the RAID to be handled by the
storage subsystem.

WARNING: Activating Linux software RAID devices concurrently on multiple servers can result
in data corruption or inconsistencies.

78 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
6.1.7 Guidelines for Component Devices
For efficient use of space and performance, the disks you use to create the RAID should have the
same storage capacity. Typically, if component devices are not of identical storage capacity, then
each member of the RAID uses only an amount of space equal to the capacity of the smallest
member disk.

Version 2.3 and later of mdadm supports component devices up to 4 TB in size each. Earlier versions
support component devices up to 2 TB in size.

IMPORTANT: If you have a local disk, external disk arrays, or SAN devices that are larger than the
supported device size, use a third-party disk partitioner to carve the devices into smaller logical
devices.

You can combine up to 28 component devices to create the RAID array. The md RAID device you
create can be up to the maximum device size supported by the file system you plan to use. For
information about file system limits for SUSE® Linux Enterprise Server 10, see “Large File System
Support” in the SUSE Linux Enterprise Server 10 Installation and Administration Guide. (http://
www.novell.com/documentation/sles10).

In general, each storage object included in the RAID should be from a different physical disk to
maximize I/O performance and to achieve disk fault tolerance where supported by the RAID level
you use. In addition, they should be of the same type (disks, segments, or regions).

Using component devices of differing speeds might introduce a bottleneck during periods of
demanding I/O. The best performance can be achieved by using the same brand and models of disks
and controllers in your hardware solution. If they are different, you should try to match disks and
controllers with similar technologies, performance, and capacity. Use a low number of drives on
each controller to maximize throughput.

IMPORTANT: As with any hardware solution, using the same brand and model introduces the risk
of concurrent failures over the life of the product, so plan maintenance accordingly.

The following table provides recommendations for the minimum and maximum number of storage
objects to use when creating a software RAID:

Table 6-6 Recommended Number of Storage Objects to Use in the Software RAID

Recommended
Minimum Number of
RAID Type Maximum Number of
Storage Objects
Storage Objects

RAID 0 (striping) 2 8

RAID 1 (mirroring) 2 4

RAID 4 (striping with dedicated parity) 3 8

RAID 5 (striping with distributed parity) 3 8

Connection fault tolerance can be achieved by having multiple connection paths to each storage
object in the RAID. For more information about configuring multipath I/O support before
configuring a software RAID, see Chapter 5, “Managing Multipath I/O for Devices,” on page 43.

Managing Software RAIDs with EVMS 79

 novdocx (en) 13 May 2009
6.1.8 RAID 5 Algorithms for Distributing Stripes and Parity
RAID 5 uses an algorithm to determine the layout of stripes and parity. The following table
describes the algorithms.

Table 6-7 RAID 5 Algorithms

Algorithm EVMS Type Description

Left Asymmetric 1 Stripes are written in a round-robin fashion from the first to last
member segment. The parity’s position in the striping sequence
moves in a round-robin fashion from last to first. For example:

sda1 sdb1 sdc1 sde1

0 1 2 p
3 4 p 5
6 p 7 8
p 9 10 11
12 13 14 p

Left Symmetric 2 This is the default setting and is considered the fastest method for
large reads.

Stripes wrap to follow the parity. The parity’s position in the striping
sequence moves in a round-robin fashion from last to first. For
example:

sda1 sdb1 sdc1 sde1

0 1 2 p
4 5 p 3
8 p 6 7
p 9 10 11
12 13 14 p

Right Asymmetric 3 Stripes are written in a round-robin fashion from the first to last
member segment. The parity’s position in the striping sequence
moves in a round-robin fashion from first to last. For example:

sda1 sdb1 sdc1 sde1

p 0 1 2
3 p 4 5
6 7 p 8
9 10 11 p
p 12 13 14

Right Symmetric 4 Stripes wrap to follow the parity. The parity’s position in the striping
sequence moves in a round-robin fashion from first to last. For
example:

sda1 sdb1 sdc1 sde1

p 0 1 2
5 p 3 4
7 8 p 6
9 10 11 p
p 12 13 14

80 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
For information about the layout of stripes and parity with each of these algorithms, see Linux
RAID-5 Algorithms (http://www.accs.com/p_and_p/RAID/LinuxRAID.html).

6.1.9 Multi-Disk Plug-In for EVMS

The Multi-Disk (MD) plug-in supports creating software RAIDs 0 (striping), 1 (mirror), 4 (striping
with dedicated parity), and 5 (striping with distributed parity). The MD plug-in to EVMS allows you
to manage all of these MD features as “regions” with the Regions Manager.

6.1.10 Device Mapper Plug-In for EVMS

The Device Mapper plug-in supports the following features in the EVMS MD Region Manager:

Š Multipath I/O: Connection fault tolerance and load balancing for connections between the
server and disks where multiple paths are available. If you plan to use multipathing, you should
configure MPIO for the devices that you plan to use in the RAID before configuring the RAID
itself. For information, see Chapter 5, “Managing Multipath I/O for Devices,” on page 43.

IMPORTANT: The EVMS interface manages multipathing under the MD Region Manager,
which originally supported the md multipath functions. It uses the legacy md terminology in
the interface and in naming of device nodes, but implements the storage objects with Device
Mapper.

Š Linear RAID: A linear concatenation of discontinuous areas of free space from the same or
multiple storage devices. Areas can be of different sizes.
Š Snapshots: Snapshots of a file system at a particular point in time, even while the system is
active, thereby allowing a consistent backup.

The Device Mapper driver is not started by default in the rescue system.

1 Open a terminal console, then log in as the root user or equivalent.

2 Start the Device Mapper by entering the following at the terminal console prompt:
/etc/init.d/boot.device-mapper start

6.2 Creating and Configuring a Software RAID

1 Open a terminal console, then log in as the root user or equivalent.
2 Start the EVMS GUI by entering the following at the terminal console prompt:
evmsgui
3 If the disks have not been initialized, initialize them by adding the DOS Segment Manager
now.
The following instructions assume you are initializing new disks. For information about
initializing an existing disk or a disk moved from another system, see Section 4.2, “Initializing
Disks,” on page 36.
Repeat the following steps for each disk that you want to initialize:
3a Select Actions > Add > Segment Manager to Storage Object.
3b From the list, select the DOS Segment Manager, then click Next.
3c Select the device, then click Add to initialize it.

Managing Software RAIDs with EVMS 81

 novdocx (en) 13 May 2009
4 If segments have not been created on the disks, create a segment on each disk that you plan to
use in the RAID.
For x86 platforms, this step is optional if you treat the entire disk as one segment.
For IA-64 platforms, this step is necessary to make the RAID 4/5 option available in the
Regions Manager.
For information about creating segments, see Section 4.4, “Creating Disk Segments (or
Partitions),” on page 38.
4a Select Action > Create > Segment to open the DOS Segment Manager.
4b Select the free space segment you want to use.
4c Specify the amount of space to use for the segment.
4d Specify the segment options, then click Create.
5 Create and configure a software RAID Device.
5a Select Action > Create > Region to open the Create Storage Region dialog box.

5b Specify the type of software RAID you want to create by selecting one of the following
Region Managers, then click Next.
Š MD RAID 0 Region Manager
Š MD RAID 1 Region Manager
Š MD RAID 4/5 Region Manager
5c From the Storage Objects listed, select the ones to use for the RAID device.

IMPORTANT: The order of the objects in the RAID is implied by their order in the list.

82 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
5d Specify values for Configuration Options by changing the following default settings as
desired.
Š For RAIDs 1, 4, or 5, optionally specify a device to use as the spare disk for the
RAID. The default is none.
Š For RAIDs 0, 4, or 5, specify the chunk (stripe) size in KB. The default is 32 KB.
Š For RAIDs 4/5, specify RAID 4 or RAID 5 (default).
Š For RAID 5, specify the algorithm to use for striping and parity. The default is Left
Symmetric.
For information about these settings, see “Configuration Options for RAIDs” on page 78.

Managing Software RAIDs with EVMS 83

 novdocx (en) 13 May 2009
5e Click Create to create the RAID device under the /dev/evms/md directory.
The device is given a name such as md0, so its EVMS mount location is /dev/evms/md/
md0.
6 Specify a human-readable label for the device.
6a Select Action > Create > EVMS Volume or Compatible Volume.
6b Select the device that you created in Step 5.
6c Specify a name for the device.
Use standard ASCII characters and naming conventions. Spaces are allowed.
6d Click Done.
7 Create a file system on the RAID device you created.
7a Select Action > File System > Make to view a list of file system modules.
7b Select the type of file system you want to create, such as the following:
Š ReiserFS File System Module
Š Ext2/3FS File System Module
7c Select the RAID device you created in Step 5, such as /dev/evms/md/md0.
7d Specify a name to use as the Volume Label, then click Make.
The name must not contain space or it will fail to mount later.
7e Click Save to create the file system.
8 Mount the RAID device.
8a Select Action > File System > Mount.

84 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
8b Select the RAID device you created in Step 5, such as /dev/evms/md/md0.
8c Specify the location where you want to mount the device, such as /home.
8d Click Mount.
9 Enable boot.evms to activate EVMS automatically at reboot.
9a In YaST, select System > System Services (Run Level).
9b Select Expert Mode.
9c Select Boot.evms.
9d Select Set/Reset.
9e Select Enable the Service.
10 Edit the /etc/fstab file to automount the RAID mount point created in Step 8c, or you can
mount the device manually from evmsgui.

6.3 Expanding a RAID

This section explains how to expand a RAID by adding segments to it.

IMPORTANT: Before you can expand the size of a RAID device, you must deactivate it.

Š Section 6.3.1, “Adding Mirrors to a RAID 1 Device,” on page 85

Š Section 6.3.2, “Adding Segments to a RAID 4 or 5,” on page 86

6.3.1 Adding Mirrors to a RAID 1 Device

In a RAID 1 device, each member segment contains its own copy of all of the data stored in the
RAID. You can add a mirror to the RAID to increase redundancy. The segment must be at least the
same size as the smallest member segment in the existing RAID 1 device. Any excess space in the
segment is not used. Ideally, all member segments of a RAID 1 device are the same size.

Adding an Available Segment as the New Mirror

1 Deactivate the RAID 1 device.

2 Use the Add Active (addactive plug-in) function.
3 From the list of available segments, select one that is the same size or larger than the smallest
existing member of the RAID device.
4 Reactivate the RAID device.

Activating A Spare Disk as the New Mirror

1 If you have not set up a spare disk, do it now.

For information, see Section 6.4, “Adding or Removing a Spare Disk,” on page 86.
2 Use the Activate Spare (activatespare plug-in) function to add it to the RAID 1 device as a
new mirror.

Managing Software RAIDs with EVMS 85

 novdocx (en) 13 May 2009
6.3.2 Adding Segments to a RAID 4 or 5
If the RAID region is clean and operating normally, the kernel driver adds the new object as a
regular spare, and it acts as a hot standby for future failures. If the RAID region is currently
degraded, the kernel driver immediately activates the new spare object and begins synchronizing the
data and parity information.

6.4 Adding or Removing a Spare Disk

The MD driver allows you to optionally designate a spare disk (device, segment, or region) for
RAID 1, 4, and 5 devices. You can assign a spare disk when you create the RAID or at any time
thereafter. The RAID can be active and in use when you add or remove the spare. The spare is
activated for the RAID only on disk failure.
Š Section 6.4.1, “Do You Need a Spare Disk?,” on page 86
Š Section 6.4.2, “Adding a Spare Disk When You Create the RAID,” on page 87
Š Section 6.4.3, “Adding a Spare Disk to an Existing RAID,” on page 87
Š Section 6.4.4, “Removing a Spare Disk from a RAID,” on page 87

6.4.1 Do You Need a Spare Disk?

The advantage of specifying a spare disk for a RAID is that the system monitors the failure and
begins recovery without human interaction. The disadvantage is that the space on the spare disk is
not available until it is activated by a failed RAID.

As noted in “Overview of RAID Levels” on page 76, RAIDs 1, 4, and 5 can tolerate at least one disk
failure. Any given RAID can have one spare disk designated for it, but the spare itself can serve as
the designated spare for one RAID, for multiple RAIDs, or for all arrays. The spare disk is a hot
standby until it is needed. It is not an active member of any RAIDs where it is assigned as the spare
disk until it is activated for that purpose.

If a spare disk is defined for the RAID, the RAID automatically deactivates the failed disk and
activates the spare disk on disk failure. The MD driver then begins synchronizing mirrored data for a
RAID 1 or reconstructing the missing data and parity information for RAIDs 4 and 5. The I/O
performance remains in a degraded state until the failed disk’s data is fully remirrored or
reconstructed.

Creating a spare-group name allows a single hot spare to service multiple RAID arrays. The spare-
group name can be any character string, but must be uniquely named for the server. For mdadm to
move spares from one array to another, the different arrays must be labelled with the same spare-
group name in the configuration file.

For example, when mdadm detects that an array is missing a component device, it first checks to see
if the array has a spare device. If no spare is available, mdadm looks in the array’s assigned spare-
group for another array that has a full complement of working drives and a spare. It attempts to
remove the spare from the working array and add it to the degraded array. If the removal succeeds
but the adding fails, then the spare is added back to its source array.

86 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
6.4.2 Adding a Spare Disk When You Create the RAID
When you create a RAID 1, 4, or 5 in EVMS, specify the Spare Disk in the Configuration Options
dialog box. You can browse to select the available device, segment, or region that you want to be the
RAID’s spare disk. For information, see Step 5d in Section 6.2, “Creating and Configuring a
Software RAID,” on page 81.

6.4.3 Adding a Spare Disk to an Existing RAID

The RAID 1, 4, or 5 device can be active and in use when you add a spare disk to it. If the RAID is
operating normally, the specified disk is added as a spare and it acts as a hot standby for future
failures. If the RAID is currently degraded because of a failed disk, the specified disk is added as a
spare disk, then it is automatically activated as a replacement disk for the failed disk, and it begins
synchronizing the data and parity information.

1 Prepare a disk, segment, or region to use as the replacement disk, just as you did for the
component devices of the RAID device.
2 In EVMS, select the Actions > Add > Spare Disk to a Region (the addspare plug-in for the
EVMS GUI).
3 Select the RAID device you want to manage from the list of Regions, then click Next.
4 Select the device to use as the spare disk.
5 Click Add.

6.4.4 Removing a Spare Disk from a RAID

The RAID 1, 4, or 5 device can be active and in use when you remove its spare disk.

1 In EVMS, select the Actions > Remove > Spare Disk from a Region (the remspare plug-in for
the EVMS GUI).
2 Select the RAID device you want to manage from the list of Regions, then click Next.
3 Select the spare disk.
4 Click Remove.

6.5 Managing Disk Failure and RAID Recovery

Š Section 6.5.1, “Understanding the Disk Failure and RAID Recovery,” on page 87
Š Section 6.5.2, “Identifying the Failed Drive,” on page 88
Š Section 6.5.3, “Replacing a Failed Device with a Spare,” on page 89
Š Section 6.5.4, “Removing the Failed Disk,” on page 90

6.5.1 Understanding the Disk Failure and RAID Recovery

RAIDs 1, 4, and 5 can survive a disk failure. A RAID 1 device survives if all but one mirrored array
fails. Its read performance is degraded without the multiple data sources available, but its write
performance might actually improve when it does not write to the failed mirrors. During the

Managing Software RAIDs with EVMS 87

 novdocx (en) 13 May 2009
synchronization of the replacement disk, write and read performance are both degraded. A RAID 5
can survive a single disk failure at a time. A RAID 4 can survive a single disk failure at a time if the
disk is not the parity disk.

Disks can fail for many reasons such as the following:

Š Disk crash
Š Disk pulled from the system
Š Drive cable removed or loose
Š I/O errors

When a disk fails, the RAID removes the failed disk from membership in the RAID, and operates in
a degraded mode until the failed disk is replaced by a spare. Degraded mode is resolved for a single
disk failure in one of the following ways:
Š Spare Exists: If the RAID has been assigned a spare disk, the MD driver automatically
activates the spare disk as a member of the RAID, then the RAID begins synchronizing (RAID
1) or reconstructing (RAID 4 or 5) the missing data.
Š No Spare Exists: If the RAID does not have a spare disk, the RAID operates in degraded mode
until you configure and add a spare. When you add the spare, the MD driver detects the RAID’s
degraded mode, automatically activates the spare as a member of the RAID, then begins
synchronizing (RAID 1) or reconstructing (RAID 4 or 5) the missing data.

6.5.2 Identifying the Failed Drive

On failure, md automatically removes the failed drive as a component device in the RAID array. To
determine which device is a problem, use mdadm and look for the device that has been reported as
“removed”.

1 Enter the following a a terminal console prompt

mdadm -D /dev/md1
Replace /dev/md1 with the actual path for your RAID.

For example, an mdadm report for a RAID 1 device consisting of /dev/sda2 and /dev/sdb2 might
look like this:
blue6:~ # mdadm -D /dev/md1

/dev/md1:

Version : 00.90.03

Creation Time : Sun Jul 2 01:14:07 2006

Raid Level : raid1

Array Size : 180201024 (171.85 GiB 184.53 GB)

Device Size : 180201024 (171.85 GiB 184.53 GB)

Raid Devices : 2

Total Devices : 1

Preferred Minor : 1

88 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Persistence : Superblock is persistent

Update Time : Tue Aug 15 18:31:09 2006

State : clean, degraded

Active Devices : 1

Working Devices : 1
Failed Devices : 0

Spare Devices : 0

UUID : 8a9f3d46:3ec09d23:86e1ffbc:ee2d0dd8

Events : 0.174164

Number Major Minor RaidDevice State

0 0 0 0 removed

1 8 18 1 active sync /dev/sdb2

The “Total Devices : 1”, “Active Devices : 1”, and “Working Devices : 1” indicate that only one of
the two devices is currently active. The RAID is operating in a “degraded” state.

The “Failed Devices : 0” might be confusing. This setting has a non-zero number only for that brief
period where the md driver finds a problem on the drive and prepares to remove it from the RAID.
When the failed drive is removed, it reads “0” again.

In the devices list at the end of the report, the device with the “removed” state for Device 0 indicates
that the device has been removed from the software RAID definition, not that the device has been
physically removed from the system. It does not specifically identify the failed device. However, the
working device (or devices) are listed. Hopefully, you have a record of which devices were members
of the RAID. By the process of elimination, the failed device is /dev/sda2.
The “Spare Devices : 0” indicates that you do not have a spare assigned to the RAID. You must
assign a spare device to the RAID so that it can be automatically added to the array and replace the
failed device.

6.5.3 Replacing a Failed Device with a Spare

When a component device fails, the md driver replaces the failed device with a spare device assigned
to the RAID. You can either keep a spare device assigned to the RAID as a hot standby to use as an
automatic replacement, or assign a spare device to the RAID as needed.

IMPORTANT: Even if you correct the problem that caused the problem disk to fail, the RAID does
not automatically accept it back into the array because it is a “faulty object” in the RAID and is no
longer synchronized with the RAID.

If a spare is available, md automatically removes the failed disk, replaces it with the spare disk, then
begins to synchronize the data (for RAID 1) or reconstruct the data from parity (for RAIDs 4 or 5).

If a spare is not available, the RAID operates in degraded mode until you assign spare device to the
RAID.

Managing Software RAIDs with EVMS 89

 novdocx (en) 13 May 2009
To assign a spare device to the RAID:

1 Prepare the disk as needed to match the other members of the RAID.
2 In EVMS, select the Actions > Add > Spare Disk to a Region (the addspare plug-in for the
EVMS GUI).
3 Select the RAID device you want to manage from the list of Regions, then click Next.
4 Select the device to use as the spare disk.
5 Click Add.
The md driver automatically begins the replacement and reconstruction or synchronization
process.
6 Monitor the status of the RAID to verify the process has begun.
For information about how monitor RAID status, see Section 6.6, “Monitoring Status for a
RAID,” on page 90.
7 Continue with Section 6.5.4, “Removing the Failed Disk,” on page 90.

6.5.4 Removing the Failed Disk

You can remove the failed disk at any time after it has been replaced with the spare disk. EVMS
does not make the device available for other use until you remove it from the RAID. After you
remove it, the disk appears in the Available-Objects list in the EVMS GUI, where it can be used for
any purpose.

NOTE: If you pull a disk or if it is totally unusable, EVMS no longer recognizes the failed disk as
part of the RAID.

The RAID device can be active and in use when you remove its faulty object.

1 In EVMS, select the Actions > Remove > Faulty Object from a Region (the remfaulty plug-in
the EVMS GUI).
2 Select the RAID device you want to manage from the list of Regions, then click Next.
3 Select the failed disk.
4 Click Remove.

6.6 Monitoring Status for a RAID

Š Section 6.6.1, “Monitoring Status with EVMSGUI,” on page 90
Š Section 6.6.2, “Monitoring Status with /proc/mdstat,” on page 91
Š Section 6.6.3, “Monitoring Status with mdadm,” on page 91
Š Section 6.6.4, “Monitoring a Remirror or Reconstruction,” on page 93
Š Section 6.6.5, “Configuring mdadm to Send an E-Mail Alert for RAID Events,” on page 93

6.6.1 Monitoring Status with EVMSGUI

The Regions tab in EVMS GUI (evmsgui) reports any software RAID devices that are defined and
whether they are currently active.

90 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
6.6.2 Monitoring Status with /proc/mdstat
A summary of RAID and status information (active/not active) is available in the /proc/mdstat
file.

1 Open a terminal console, then log in as the root user or equivalent.

2 View the /proc/mdstat file by entering the following at the console prompt:
cat /proc/mdstat
3 Evaluate the information.
The following table shows an example output and how to interpret the information.

Status Information Description Interpretation

Personalities : [raid5] List of the RAIDs on the server You have two RAIDs defined
[raid4] by RAID label. with labels of raid5 and
raid4.

md0 : active <device> : <active | not active> The RAID is active and
raid5 <RAID label you specified> mounted at /dev/evms/
sdg1[0] sdk1[4] sdj1[3] < storage object> [RAID order] md/md0.
sdi1[2] The RAID label is raid5.
The active segments are sdg1,
sdi1, sdj1, and sdk1, as
ordered in the RAID.
The RAID numbering of 0 to 4
indicates that the RAID has 5
segments, and the second
segment [1] is missing from
the list. Based on the
segment names, the missing
segment is sdh1.

35535360 blocks level <number of blocks> blocks If the block size on the server is
5, 128k chunk, level < 0 | 1 | 4 | 5 > 4 KB, the total size of the
algorithm 2 [5/4] <stripe size in KB> chunk RAID (including parity) is
[U_UUU] 142 GB, with a data capacity
algorithm <1 | 2 | 3 | 4 >
of 113.7 GB.
[number of devices/number of
The stripe size is 128 KB.
working devices]
The RAID is using left
[U-UUU]
symmetric.
algorithm <1 | 2 | 3 | 4 >
[number of devices/number of
working devices]
[U-UUU]

unused devices: <none> All segments in the RAID are in There are no spare devices
use. available on the server.

6.6.3 Monitoring Status with mdadm

To view the RAID status with the mdadm command, enter the following at a terminal prompt:
mdadm -D /dev/mdx

Managing Software RAIDs with EVMS 91

 novdocx (en) 13 May 2009
Replace mdx with the RAID device number.

Example 1: A Disk Fails

In the following example, only four of the five devices in the RAID are active (Raid Devices : 5,
Total Devices : 4). When it was created, the component devices in the device were numbered 0
to 5 and are ordered according to their alphabetic appearance in the list where they were chosen,
such as /dev/sdg1, /dev/sdh1, /dev/sdi1, /dev/sdj1, and /dev/sdk1. From the pattern of
filenames of the other devices, you determine that the device that was removed was named /dev/
sdh1.

/dev/md0:
Version : 00.90.03
Creation Time : Sun Apr 16 11:37:05 2006
Raid Level : raid5
Array Size : 35535360 (33.89 GiB 36.39 GB)
Device Size : 8883840 (8.47 GiB 9.10 GB)
Raid Devices : 5
Total Devices : 4
Preferred Minor : 0
Persistence : Superblock is persistent
Update Time : Mon Apr 17 05:50:44 2006
State : clean, degraded
Active Devices : 4
Working Devices : 4
Failed Devices : 0
Spare Devices : 0
Layout : left-symmetric
Chunk Size : 128K
UUID : 2e686e87:1eb36d02:d3914df8:db197afe
Events : 0.189
Number Major Minor RaidDevice State
0 8 97 0 active sync /dev/sdg1
1 8 0 1 removed
2 8 129 2 active sync /dev/sdi1
3 8 45 3 active sync /dev/sdj1
4 8 161 4 active sync /dev/sdk1

Example 2: Spare Disk Replaces the Failed Disk

In the following mdadm report, only 4 of the 5 disks are active and in good condition (Active
Devices : 4, Working Devices : 5). The failed disk was automatically detected and removed
from the RAID (Failed Devices: 0). The spare was activated as the replacement disk, and has
assumed the diskname of the failed disk (/dev/sdh1). The faulty object (the failed disk that was
removed from the RAID) is not identified in the report. The RAID is running in degraded mode
(State : clean, degraded, recovering). The data is being rebuilt (spare rebuilding /
dev/sdh1), and the process is 3% complete (Rebuild Status : 3% complete).

92 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
mdadm -D /dev/md0
/dev/md0:
Version : 00.90.03
Creation Time : Sun Apr 16 11:37:05 2006
Raid Level : raid5
Array Size : 35535360 (33.89 GiB 36.39 GB)
Device Size : 8883840 (8.47 GiB 9.10 GB)
Raid Devices : 5
Total Devices : 5
Preferred Minor : 0
Persistence : Superblock is persistent
Update Time : Mon Apr 17 05:50:44 2006
State : clean, degraded, recovering
Active Devices : 4
Working Devices : 5
Failed Devices : 0
Spare Devices : 1
Layout : left-symmetric
Chunk Size : 128K
Rebuild Status : 3% complete
UUID : 2e686e87:1eb36d02:d3914df8:db197afe
Events : 0.189
Number Major Minor RaidDevice State
0 8 97 0 active sync /dev/sdg1
1 8 113 1 spare rebuilding /dev/sdh1
2 8 129 2 active sync /dev/sdi1
3 8 145 3 active sync /dev/sdj1
4 8 161 4 active sync /dev/sdk1

6.6.4 Monitoring a Remirror or Reconstruction

You can follow the progress of the synchronization or reconstruction process by examining the /
proc/mdstat file.

You can control the speed of synchronization by setting parameters in the /proc/sys/dev/raid/
speed_limit_min and /proc/sys/dev/raid/speed_limit_max files. To speed up the process,
echo a larger number into the speed_limit_min file.

6.6.5 Configuring mdadm to Send an E-Mail Alert for RAID

Events
You might want to configure the mdadm service to send an e-mail alert for software RAID events.
Monitoring is only meaningful for RAIDs 1, 4, 5, 6, 10 or multipath arrays because only these have
missing, spare, or failed drives to monitor. RAID 0 and Linear RAIDs do not provide fault tolerance
so they have no interesting states to monitor.

Managing Software RAIDs with EVMS 93

 novdocx (en) 13 May 2009
The following table identifies RAID events and indicates which events trigger e-mail alerts. All
events cause the program to run. The program is run with two or three arguments: the event name,
the array device (such as /dev/md1), and possibly a second device. For Fail, Fail Spare, and Spare
Active, the second device is the relevant component device. For MoveSpare, the second device is
the array that the spare was moved from.

Table 6-8 RAID Events in mdadm

Trigger
RAID Event E-Mail Description
Alert

Device Disappeared No An md array that was previously configured appears to no longer be

configured. (syslog priority: Critical)

If mdadm was told to monitor an array which is RAID0 or Linear, then

it reports DeviceDisappeared with the extra information Wrong-Level.
This is because RAID0 and Linear do not support the device-failed,
hot-spare, and resynchronize operations that are monitored.

Rebuild Started No An md array started reconstruction. (syslog priority: Warning)

Rebuild NN No Where NN is 20, 40, 60, or 80. This indicates the percent completed
for the rebuild. (syslog priority: Warning)

Rebuild Finished No An md array that was rebuilding is no longer rebuilding, either

because it finished normally or was aborted. (syslog priority:
Warning)

Fail Yes An active component device of an array has been marked as faulty.
(syslog priority: Critical)

Fail Spare Yes A spare component device that was being rebuilt to replace a faulty
device has failed. (syslog priority: Critical)

Spare Active No A spare component device that was being rebuilt to replace a faulty
device has been successfully rebuilt and has been made active.
(syslog priority: Info)

New Array No A new md array has been detected in the /proc/mdstat file.
(syslog priority: Info)

Degraded Array Yes A newly noticed array appears to be degraded. This message is not
generated when mdadm notices a drive failure that causes
degradation. It is generated only when mdadm notices that an array is
degraded when it first sees the array. (syslog priority: Critical)

Move Spare No A spare drive has been moved from one array in a spare group to
another to allow a failed drive to be replaced. (syslog priority: Info)

Spares Missing Yes The mdadm.conf file indicates that an array should have a certain
number of spare devices, but mdadm detects that the array has fewer
than this number when it first sees the array. (syslog priority:
Warning)

Test Message Yes An array was found at startup, and the --test flag was given.
(syslog priority: Info)

94 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
To configure an e-mail alert:

1 At a terminal console, log in as the root user.

2 Edit the /etc/mdadm/mdadm.conf file to add your e-mail address for receiving alerts. For
example, specify the MAILADDR value (using your own e-mail address, of course):
DEVICE partitions
ARRAY /dev/md0 level=raid1 num-devices=2
UUID=1c661ae4:818165c3:3f7a4661:af475fda
devices=/dev/sdb3,/dev/sdc3
MAILADDR [email protected]
The MAILADDR line gives an e-mail address that alerts should be sent to when mdadm is
running in --monitor mode with the --scan option. There should be only one MAILADDR
line in mdadm.conf, and it should have only one address.
3 Start mdadm monitoring by entering the following at the terminal console prompt:
mdadm --monitor [email protected] --delay=1800 /dev/md0
The --monitor option causes mdadm to periodically poll a number of md arrays and to report
on any events noticed. mdadm never exits once it decides that there are arrays to be checked, so
it should normally be run in the background.
In addition to reporting events in this mode, mdadm might move a spare drive from one array to
another if they are in the same spare-group and if the destination array has a failed drive but no
spares.
Listing the devices to monitor is optional. If any devices are listed on the command line, mdadm
monitors only those devices. Otherwise, all arrays listed in the configuration file are monitored.
Further, if --scan option is added in the command, then any other md devices that appear in /
proc/mdstat are also monitored.
For more information about using mdadm, see the mdadm(8) and mdadm.conf(5) man pages.
4 To configure the /etc/init.d/mdadmd service as a script:
suse:~ # egrep 'MAIL|RAIDDEVICE' /etc/sysconfig/mdadm
MDADM_MAIL="[email protected]"
MDADM_RAIDDEVICES="/dev/md0"
MDADM_SEND_MAIL_ON_START=no
suse:~ # chkconfig mdadmd --list
mdadmd 0:off 1:off 2:off 3:on 4:off 5:on 6:off

6.7 Deleting a Software RAID and Its Data

If you want to remove the prior multipath settings, deactivate the RAID, delete the data on the
RAID, and release all resources used by the RAID, do the following:

1 If you want to keep the data stored on the software RAID device, make sure to back up the data
to alternate media, using your normal backup procedures. Make sure the backup is good before
proceeding.
2 Open a terminal console prompt as the root user or equivalent. Use this console to enter the
commands described in the remaining steps.
3 Dismount the software RAID device by entering

Managing Software RAIDs with EVMS 95

 novdocx (en) 13 May 2009
umount <raid-device>
4 Stop the RAID device and its component devices by entering
mdadm --stop <raid-device>
mdadm --stop <member-devices>
For more information about using mdadm, please see the mdadm(8) man page.
5 Delete all data on the disk by literally overwriting the entire device with zeroes. Enter
mdadm --misc --zero-superblock <member-devices>
6 You must now reinitialize the disks for other uses, just as you would when adding a new disk to
your system.

96 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Managing Software RAIDs 6 and
7

10 with mdadm 7
This section describes how to create software RAID 6 and 10 devices, using the Multiple Devices
Administration (mdadm(8)) tool. You can also use mdadm to create RAIDs 0, 1, 4, and 5. The mdadm
tool provides the functionality of legacy programs mdtools and raidtools.
Š Section 7.1, “Creating a RAID 6,” on page 97
Š Section 7.2, “Creating Nested RAID 10 Devices with mdadm,” on page 98
Š Section 7.3, “Creating a Complex RAID 10 with mdadm,” on page 101
Š Section 7.4, “Creating a Degraded RAID Array,” on page 104

7.1 Creating a RAID 6

Š Section 7.1.1, “Understanding RAID 6,” on page 97
Š Section 7.1.2, “Creating a RAID 6,” on page 98

7.1.1 Understanding RAID 6

RAID 6 is essentially an extension of RAID 5 that allows for additional fault tolerance by using a
second independent distributed parity scheme (dual parity). Even if one of the hard disk drives fails
during the data recovery process, the system continues to be operational, with no data loss.

RAID6 provides for extremely high data fault tolerance by sustaining multiple simultaneous drive
failures. It handles the loss of any two devices without data loss. Accordingly, it requires N+2 drives
to store N drives worth of data. It requires a minimum of 4 devices.

The performance for RAID 6 is slightly lower but comparable to RAID 5 in normal mode and single
disk failure mode. It is very slow in dual disk failure mode.

Table 7-1 Comparison of RAID 5 and RAID 6

Feature RAID 5 RAID 6

Number of devices N+1, minimum of 3 N+2, minimum of 4

Parity Distributed, single Distributed, dual

Performance Medium impact on write and More impact on sequential write

rebuild than RAID 5

Fault-tolerance Failure of one component device Failure of two component devices

Managing Software RAIDs 6 and 10 with mdadm 97

 novdocx (en) 13 May 2009
7.1.2 Creating a RAID 6
The procedure in this section creates a RAID 6 device /dev/md0 with four devices: /dev/sda1, /
dev/sdb1, /dev/sdc1, and /dev/sdd1. Make sure to modify the procedure to use your actual
device nodes.
1 Open a terminal console, then log in as the root user or equivalent.
2 Create a RAID 6 device. At the command prompt, enter
mdadm --create /dev/md0 --run --level=raid6 --chunk=128 --raid-devices=4 /
dev/sdb1 /dev/sdc1 /dev/sdc1 /dev/sdd1
The default chunk size is 64 (KB).
3 Create a file system on the RAID 6 device /dev/md0, such as a Reiser file system (reiserfs).
For example, at the command prompt, enter
mkfs.reiserfs /dev/md0
Modify the command if you want to use a different file system.
4 Edit the /etc/mdadm.conf file to add entries for the component devices and the RAID device
/dev/md0.
5 Edit the /etc/fstab file to add an entry for the RAID 6 device /dev/md0.
6 Reboot the server.
The RAID 6 device is mounted to /local.
7 (Optional) Add a hot spare to service the RAID array. For example, at the command prompt
enter:
mdadm /dev/md0 -a /dev/sde1

7.2 Creating Nested RAID 10 Devices with

mdadm
Š Section 7.2.1, “Understanding Nested RAID Devices,” on page 98
Š Section 7.2.2, “Creating Nested RAID 10 (1+0) with mdadm,” on page 99
Š Section 7.2.3, “Creating Nested RAID 10 (0+1) with mdadm,” on page 100

7.2.1 Understanding Nested RAID Devices

A nested RAID device consists of a RAID array that uses another RAID array as its basic element,
instead of using physical disks. The goal of this configuration is to improve the performance and
fault tolerance of the RAID.

Linux supports nesting of RAID 1 (mirroring) and RAID 0 (striping) arrays. Generally, this
combination is referred to as RAID 10. To distinguish the order of the nesting, this document uses
the following terminology:
Š RAID 1+0: RAID 1 (mirror) arrays are built first, then combined to form a RAID 0 (stripe)
array.
Š RAID 0+1: RAID 0 (stripe) arrays are built first, then combined to form a RAID 1 (mirror)
array.

98 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
The following table describes the advantages and disadvantages of RAID 10 nesting as 1+0 versus
0+1. It assumes that the storage objects you use reside on different disks, each with a dedicated I/O
capability.

Table 7-2 RAID Levels Supported in EVMS

RAID Level Description Performance and Fault Tolerance

10 (1+0) RAID 0 (stripe) RAID 1+0 provides high levels of I/O performance, data redundancy,
built with RAID 1 and disk fault tolerance. Because each member device in the RAID 0 is
(mirror) arrays mirrored individually, multiple disk failures can be tolerated and data
remains available as long as the disks that fail are in different mirrors.

You can optionally configure a spare for each underlying mirrored array,
or configure a spare to serve a spare group that serves all mirrors.

10 (0+1) RAID 1 (mirror) RAID 0+1 provides high levels of I/O performance and data
built with RAID 0 redundancy, but slightly less fault tolerance than a 1+0. If multiple disks
(stripe) arrays fail on one side of the mirror, then the other mirror is available. However,
if disks are lost concurrently on both sides of the mirror, all data is lost.

This solution offers less disk fault tolerance than a 1+0 solution, but if
you need to perform maintenance or maintain the mirror on a different
site, you can take an entire side of the mirror offline and still have a fully
functional storage device. Also, if you lose the connection between the
two sites, either site operates independently of the other. That is not
true if you stripe the mirrored segments, because the mirrors are
managed at a lower level.

If a device fails, the mirror on that side fails because RAID 1 is not fault-
tolerant. Create a new RAID 0 to replace the failed side, then
resynchronize the mirrors.

7.2.2 Creating Nested RAID 10 (1+0) with mdadm

A nested RAID 1+0 is built by creating two or more RAID 1 (mirror) devices, then using them as
component devices in a RAID 0.

IMPORTANT: If you need to manage multiple connections to the devices, you must configure
multipath I/O before configuring the RAID devices. For information, see Chapter 5, “Managing
Multipath I/O for Devices,” on page 43.

The procedure in this section uses the device names shown in the following table. Make sure to
modify the device names with the names of your own devices.

Managing Software RAIDs 6 and 10 with mdadm 99

 novdocx (en) 13 May 2009
Table 7-3 Scenario for Creating a RAID 10 (1+0) by Nesting

Raw Devices RAID 1 (mirror) RAID 1+0 (striped mirrors)

/dev/sdb1 /dev/md0 /dev/md2

/dev/sdc1

/dev/sdd1 /dev/md1

/dev/sde1

1 Open a terminal console, then log in as the root user or equivalent.

2 Create 2 software RAID 1 devices, using two different devices for each RAID 1 device. At the
command prompt, enter these two commands:
mdadm --create /dev/md0 --run --level=1 --raid-devices=2 /dev/sdb1 /dev/
sdc1
mdadm --create /dev/md1 --run --level=1 --raid-devices=2 /dev/sdd1 /dev/
sde1
3 Create the nested RAID 1+0 device. At the command prompt, enter the following command
using the software RAID 1 devices you created in Step 2:
mdadm --create /dev/md2 --run --level=0 --chunk=64 --raid-devices=2 /dev/
md0 /dev/md1
The default chunk size is 64 KB.
4 Create a file system on the RAID 1+0 device /dev/md2, such as a Reiser file system (reiserfs).
For example, at the command prompt, enter
mkfs.reiserfs /dev/md2
Modify the command if you want to use a different file system.
5 Edit the /etc/mdadm.conf file to add entries for the component devices and the RAID device
/dev/md2.
6 Edit the /etc/fstab file to add an entry for the RAID 1+0 device /dev/md2.
7 Reboot the server.
The RAID 1+0 device is mounted to /local.
8 (Optional) Add hot spares to service the underlying RAID 1 mirrors.
For information, see Section 6.4, “Adding or Removing a Spare Disk,” on page 86.

7.2.3 Creating Nested RAID 10 (0+1) with mdadm

A nested RAID 0+1 is built by creating two to four RAID 0 (striping) devices, then mirroring them
as component devices in a RAID 1.

IMPORTANT: If you need to manage multiple connections to the devices, you must configure
multipath I/O before configuring the RAID devices. For information, see Chapter 5, “Managing
Multipath I/O for Devices,” on page 43.

In this configuration, spare devices cannot be specified for the underlying RAID 0 devices because
RAID 0 cannot tolerate a device loss. If a device fails on one side of the mirror, you must create a
replacement RAID 0 device, than add it into the mirror.

100 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
The procedure in this section uses the device names shown in the following table. Make sure to
modify the device names with the names of your own devices.

Table 7-4 Scenario for Creating a RAID 10 (0+1) by Nesting

Raw Devices RAID 0 (stripe) RAID 0+1 (mirrored stripes)

/dev/sdb1 /dev/md0 /dev/md2

/dev/sdc1

/dev/sdd1 /dev/md1

/dev/sde1

1 Open a terminal console, then log in as the root user or equivalent.

2 Create 2 software RAID 0 devices, using two different devices for each RAID 0 device. At the
command prompt, enter these two commands:
mdadm --create /dev/md0 --run --level=0 --chunk=64 --raid-devices=2 /dev/
sdb1 /dev/sdc1
mdadm --create /dev/md1 --run --level=0 --chunk=64 --raid-devices=2 /dev/
sdd1 /dev/sde1
The default chunk size is 64 KB.
3 Create the nested RAID 0+1 device. At the command prompt, enter the following command
using the software RAID 0 devices you created in Step 2:
mdadm --create /dev/md2 --run --level=1 --raid-devices=2 /dev/md0 /dev/md1
4 Create a file system on the RAID 0+1 device /dev/md2, such as a Reiser file system (reiserfs).
For example, at the command prompt, enter
mkfs.reiserfs /dev/md2
Modify the command if you want to use a different file system.
5 Edit the /etc/mdadm.conf file to add entries for the component devices and the RAID device
/dev/md2.
6 Edit the /etc/fstab file to add an entry for the RAID 0+1 device /dev/md2.
7 Reboot the server.
The RAID 0+1 device is mounted to /local.

7.3 Creating a Complex RAID 10 with mdadm

Š Section 7.3.1, “Understanding the mdadm RAID10,” on page 101
Š Section 7.3.2, “Creating a RAID10 with mdadm,” on page 104

7.3.1 Understanding the mdadm RAID10

In mdadm, the RAID10 level creates a single complex software RAID that combines features of both
RAID 0 (striping) and RAID 1 (mirroring). Multiple copies of all data blocks are arranged on
multiple drives following a striping discipline. Component devices should be the same size.
Š “Comparison of RAID10 Option and Nested RAID 10 (1+0)” on page 102

Managing Software RAIDs 6 and 10 with mdadm 101

 novdocx (en) 13 May 2009
Š “Number of Replicas in the mdadm RAID10” on page 102
Š “Number of Devices in the mdadm RAID10” on page 102
Š “Near Layout” on page 103
Š “Far Layout” on page 103

Comparison of RAID10 Option and Nested RAID 10 (1+0)

The complex RAID 10 is similar in purpose to a nested RAID 10 (1+0), but differs in the following
ways:

Table 7-5 Complex vs. Nested RAID 10

Feature mdadm RAID10 Option Nested RAID 10 (1+0)

Number of devices Allows an even or odd number of Requires an even number of

component devices component devices

Component devices Managed as a single RAID Manage as a nested RAID device

device

Striping Striping occurs in the near or far Striping occurs consecutively

layout on component devices. across component devices

The far layout provides sequential

read throughput that scales by
number of drives, rather than
number of RAID 1 pairs.

Multiple copies of data Two or more copies, up to the Copies on each mirrored
number of devices in the array segment

Hot spare devices A single spare can service all Configure a spare for each
component devices underlying mirrored array, or
configure a spare to serve a
spare group that serves all
mirrors.

Number of Replicas in the mdadm RAID10

When configuring a RAID10-level array, you must specify the number of replicas of each data block
that are required. The default number of replicas is 2, but the value can be 2 to the number of devices
in the array.

Number of Devices in the mdadm RAID10

You must use at least as many component devices as the number of replicas you specify. However,
number of component devices in a RAID10-level array does not need to be a multiple of the number
of replicas of each data block. The effective storage size is the number of devices divided by the
number of replicas.

For example, if you specify 2 replicas for an array created with 5 component devices, a copy of each
block is stored on two different devices. The effective storage size for one copy of all data is 5/2 or
2.5 times the size of a component device.

102 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Near Layout

With the near layout, copies of a block of data are striped near each other on different component
devices. That is, multiple copies of one data block are at similar offsets in different devices. Near is
the default layout for RAID10. For example, if you use an odd number of component devices and
two copies of data, some copies are perhaps one chunk further into the device.

The near layout for the mdadm RAID10 yields read and write performance similar to RAID 0 over
half the number of drives.
Near layout with an even number of disks and two replicas:
sda1 sdb1 sdc1 sde1
0 0 1 1
2 2 3 3
4 4 5 5
6 6 7 7
8 8 9 9

Near layout with an odd number of disks and two replicas:

sda1 sdb1 sdc1 sde1 sdf1

0 0 1 1 2
2 3 3 4 4
5 5 6 6 7
7 8 8 9 9
10 10 11 11 12

Far Layout

The far layout stripes data over the early part of all drives, then stripes a second copy of the data
over the later part of all drives, making sure that all copies of a block are on different drives. The
second set of values start halfway through the component drives.

With a far layout, the read performance of the mdadm RAID10 is similar to a RAID 0 over the full
number of drives, but write performance is substantially slower than a RAID 0 because there is more
seeking of the drive heads. It is best used for read-intensive operations such as for read-only file
servers.

Far layout with an even number of disks and two replicas:

sda1 sdb1 sdc1 sde1
0 1 2 3
3 5 6 7
. . .
3 1 2 3
7 4 5 6

Far layout with an odd number of disks and two replicas:

sda1 sdb1 sdc1 sde1 sdf1

0 1 2 3 4

Managing Software RAIDs 6 and 10 with mdadm 103

 novdocx (en) 13 May 2009
5 6 7 8 9
. . .
4 0 1 2 3
9 5 6 7 8

7.3.2 Creating a RAID10 with mdadm

The RAID10-level option for mdadm creates a RAID 10 device without nesting. For information
about the RAID10-level, see Section 7.3, “Creating a Complex RAID 10 with mdadm,” on
page 101.

The procedure in this section uses the device names shown in the following table. Make sure to
modify the device names with the names of your own devices.

Table 7-6 Scenario for Creating a RAID 10 Using the mdadm RAID10 Option

Raw Devices RAID10 (near or far striping scheme)

/dev/sdf1 /dev/md3

/dev/sdg1
/dev/sdh1

/dev/sdi1

1 In YaST, create a 0xFD Linux RAID partition on the devices you want to use in the RAID, such
as /dev/sdf1, /dev/sdg1, /dev/sdh1, and /dev/sdi1.
2 Open a terminal console, then log in as the root user or equivalent.
3 Create a RAID 10 command. At the command prompt, enter (all on the same line):
mdadm --create /dev/md3 --run --level=10 --chunk=4 --raid-devices=4 /dev/
sdf1 /dev/sdg1 /dev/sdh1 /dev/sdi1
4 Create a Reiser file system on the RAID 10 device /dev/md3. At the command prompt, enter
mkfs.reiserfs /dev/md3
5 Edit the /etc/mdadm.conf file to add entries for the component devices and the RAID device
/dev/md3. For example:
DEVICE /dev/md3
6 Edit the /etc/fstab file to add an entry for the RAID 10 device /dev/md3.
7 Reboot the server.
The RAID10 device is mounted to /raid10.

7.4 Creating a Degraded RAID Array

A degraded array is one in which some devices are missing. Degraded arrays are supported only for
RAID 1, RAID 4, RAID 5, and RAID 6. These RAID types are designed to withstand some missing
devices as part of their fault-tolerance features. Typically, degraded arrays occur when a device fails.
It is possible to create a degraded array on purpose.

104 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
RAID Type Allowable Number of Slots Missing

RAID 1 All but one device

RAID 4 One slot

RAID 5 One slot

RAID 6 One or two slots

To create a degraded array in which some devices are missing, simply give the word missing in
place of a device name. This causes mdadm to leave the corresponding slot in the array empty.
When creating a RAID 5 array, mdadm automatically creates a degraded array with an extra spare
drive. This is because building the spare into a degraded array is generally faster than
resynchronizing the parity on a non-degraded, but not clean, array. You can override this feature
with the --force option.
Creating a degraded array might be useful if you want create a RAID, but one of the devices you
want to use already has data on it. In that case, you create a degraded array with other devices, copy
data from the in-use device to the RAID that is running in degraded mode, add the device into the
RAID, then wait while the RAID is rebuilt so that the data is now across all devices. An example of
this process is given in the following procedure:

1 Create a degraded RAID 1 device /dev/md0, using one single drive /dev/sd1, enter the
following at the command prompt:
mdadm --create /dev/md0 -l 1 -n 2 /dev/sda1 missing
The device should be the same size or larger than the device you plan to add to it.
2 If the device you want to add to the mirror contains data that you want to move to the RAID
array, copy it now to the RAID array while it is running in degraded mode.
3 Add a device to the mirror. For example, to add /dev/sdb1 to the RAID, enter the following at
the command prompt:
mdadm /dev/md0 -a /dev/sdb1
You can add only one device at a time. You must wait for the kernel to build the mirror and
bring it fully online before you add another mirror.
4 Monitor the build progress by entering the following at the command prompt:
cat /proc/mdstat

Managing Software RAIDs 6 and 10 with mdadm 105

 novdocx (en) 13 May 2009

106 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Resizing Software RAID Arrays
8

with mdadm 8
This section describes how to increase or reduce the size of a software RAID 1, 4, 5, or 6 device
with the Multiple Device Administration (mdadm(8)) tool.

WARNING: Before starting any of the tasks described in this chapter, make sure that you have a
valid backup of all of the data.

Š Section 8.1, “Understanding the Resizing Process,” on page 107

Š Section 8.2, “Increasing the Size of a Software RAID,” on page 108
Š Section 8.3, “Decreasing the Size of a Software RAID,” on page 112

8.1 Understanding the Resizing Process

Resizing an existing software RAID device involves growing or shrinking the space contributed by
each component partition.
Š Section 8.1.1, “Guidelines for Resizing a Software RAID,” on page 107
Š Section 8.1.2, “Overview of Tasks,” on page 108

8.1.1 Guidelines for Resizing a Software RAID

The mdadm(8) tool supports resizing only for software RAID levels 1, 4, 5, and 6. These RAID
levels provide disk fault tolerance so that one component partition can be removed at a time for
resizing. In principle, it is possible to perform a hot resize for RAID partitions, but you must take
extra care for your data when doing so.

The file system that resides on the RAID must also be able to be resized in order to take advantage
of the changes in available space on the device. In SUSE® Linux Enterprise Server 10 SP1, file
system resizing utilities are available for file systems Ext2, Ext3, JFS, and ReiserFS. The utilities
support growing and shrinking the size as follows:

Table 8-1 File System Support for Resizing

File System Utility Increase Size Decrease Size

Ext2 or Ext3 resize2fs Yes, offline only Yes, offline only

JFS mount -o remount,resize Yes, online only No

ReiserFS resize_reiserfs Yes, online or offline Yes, offline only

Resizing any partition or file system involves some risks that can potentially result in losing data.

WARNING: To avoid data loss, make sure to back up your data before you begin any resizing task.

Resizing Software RAID Arrays with mdadm 107

 novdocx (en) 13 May 2009
8.1.2 Overview of Tasks
Resizing the RAID involves the following tasks. The order in which these tasks is performed
depends on whether you are increasing or decreasing its size.

Table 8-2 Tasks Involved in Resizing a RAID

Order If Order If
Tasks Description Increasing Decreasing
Size Size

Resize each of the Increase or decrease the active size of each 1 2

component partitions. component partition. You remove only one
component partition at a time, modify its size, then
return it to the RAID.

Resize the software The RAID does not automatically know about the 2 3
RAID itself. increases or decreases you make to the underlying
component partitions. You must inform it about the
new size.

Resize the file system. You must resize the file system that resides on the 3 1
RAID. This is possible only for file systems that
provide tools for resizing, such as Ext2, Ext3, JFS,
and ReiserFS.

8.2 Increasing the Size of a Software RAID

Before you begin, review the guidelines in Section 8.1, “Understanding the Resizing Process,” on
page 107.
Š Section 8.2.1, “Increasing the Size of Component Partitions,” on page 108
Š Section 8.2.2, “Increasing the Size of the RAID Array,” on page 109
Š Section 8.2.3, “Increasing the Size of the File System,” on page 110

8.2.1 Increasing the Size of Component Partitions

Apply the procedure in this section to increase the size of a RAID 1, 4, 5, or 6. For each component
partition in the RAID, remove the partition from the RAID, modify its size, return it to the RAID,
then wait until the RAID stabilizes to continue. While a partition is removed, the RAID operates in
degraded mode and has no or reduced disk fault tolerance. Even for RAIDs that can tolerate multiple
concurrent disk failures, do not remove more than one component partition at a time.

WARNING: If a RAID does not have disk fault tolerance, or it is simply not consistent, data loss
results if you remove any of its partitions. Be very careful when removing partitions, and make sure
that you have a backup of your data available.

The procedure in this section uses the device names shown in the following table. Make sure to
modify the names to use the names of your own devices.

108 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Table 8-3 Scenario for Increasing the Size of Component Partitions

RAID Device Component Partitions

/dev/md0 /dev/sda1

/dev/sdb1

/dev/sdc1

To increase the size of the component partitions for the RAID:

1 Open a terminal console, then log in as the root user or equivalent.

2 Make sure that the RAID array is consistent and synchronized by entering
cat /proc/mdstat
If your RAID array is still synchronizing according to the output of this command, you must
wait until synchronization is complete before continuing.
3 Remove one of the component partitions from the RAID array. For example, to remove /dev/
sda1, enter
mdadm /dev/md0 --fail /dev/sda1 --remove /dev/sda1
In order to succeed, both the fail and remove actions must be done.
4 Increase the size of the partition that you removed in Step 3 by doing one of the following:
Š Increase the size of the partition, using a disk partitioner such as fdisk(8), cfdisk(8),
or parted(8). This is the usual choice.
Š Replace the disk on which the partition resides with a higher-capacity device.
This option is possible only if no other file systems on the original disk are accessed by the
system. When the replacement device is added back into the RAID, it takes much longer
to synchronize the data because all of the data that was on the original device must be
rebuilt.
5 Re-add the partition to the RAID array. For example, to add /dev/sda1, enter
mdadm -a /dev/md0 /dev/sda1
Wait until the RAID is synchronized and consistent before continuing with the next partition.
6 Repeat Step 2 through Step 5 for each of the remaining component devices in the array. Make
sure to modify the commands for the correct component partition.
7 If you get a message that tells you that the kernel could not re-read the partition table for the
RAID, you must reboot the computer after all partitions have been resized to force an update of
the partition table.
8 Continue with Section 8.2.2, “Increasing the Size of the RAID Array,” on page 109.

8.2.2 Increasing the Size of the RAID Array

After you have resized each of the component partitions in the RAID (see Section 8.2.1, “Increasing
the Size of Component Partitions,” on page 108), the RAID array configuration continues to use the
original array size until you force it to be aware of the newly available space. You can specify a size
for the RAID or use the maximum available space.

Resizing Software RAID Arrays with mdadm 109

 novdocx (en) 13 May 2009
The procedure in this section uses the device name /dev/md0 for the RAID device. Make sure to
modify the name to use the name of your own device.

1 Open a terminal console, then log in as the root user or equivalent.

2 Check the size of the array and the device size known to the array by entering
mdadm -D /dev/md0 | grep -e "Array Size" -e "Device
Size"
3 Do one of the following:
Š Increase the size of the array to the maximum available size by entering
mdadm --grow /dev/md0 -z max
Š Increase the size of the array to a specified value by entering
mdadm --grow /dev/md0 -z size
Replace size with an integer value in kilobytes (a kilobyte is 1024 bytes) for the desired
size.
4 Recheck the size of your array and the device size known to the array by entering
mdadm -D /dev/md0 | grep -e "Array Size" -e "Device
Size"
5 Do one of the following:
Š If your array was successfully resized, continue with Section 8.2.3, “Increasing the Size of
the File System,” on page 110.
Š If your array was not resized as you expected, you must reboot, then try this procedure
again.

8.2.3 Increasing the Size of the File System

After you increase the size of the array (see Section 8.2.2, “Increasing the Size of the RAID Array,”
on page 109), you are ready to resize the file system.

You can increase the size of the file system to the maximum space available or specify an exact size.
When specifying an exact size for the file system, make sure the new size satisfies the following
conditions:
Š The new size must be greater than the size of the existing data; otherwise, data loss occurs.
Š The new size must be equal to or less than the current RAID size because the file system size
cannot extend beyond the space available.

Ext2 or Ext3

Ext2 and Ext3 file systems can be resized when mounted or unmounted with the command
resize2fs.

1 Open a terminal console, then log in as the root user or equivalent.

2 Increase the size of the file system using one of the following methods:
Š To extend the file system size to the maximum available size of the software RAID device
called /dev/md0, enter
resize2fs /dev/md0
If a size parameter is not specified, the size defaults to the size of the partition.

110 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Š To extend the file system to a specific size, enter
resize2fs /dev/md0 size
The size parameter specifies the requested new size of the file system. If no units are
specified, the unit of the size parameter is the block size of the file system. Optionally, the
size parameter may be suffixed by one of the following the unit designators: s for 512 byte
sectors; K for kilobytes (1 kilobyte is 1024 bytes); M for megabytes; or G for gigabytes.
Wait until the resizing is completed before continuing.
3 If the file system is not mounted, mount it now.
For example, to mount an Ext2 file system for a RAID named /dev/md0 at mount point /raid,
enter
mount -t ext2 /dev/md0 /raid
4 Check the effect of the resize on the mounted file system by entering
df -h
The Disk Free (df) command shows the total size of the disk, the number of blocks used, and
the number of blocks available on the file system. The -h option print sizes in human-readable
format, such as 1K, 234M, or 2G.

JFS

Resizing a JFS partition is done with a special option to the mount command that is specific to the
JFS file system:
mount -o remount,resize /mnt

Using the resize option is valid only during a remount when the volume is already mounted read-
write. The mount point is specified rather than the device name.

1 Open a terminal console, then log in as the root user or equivalent.

2 Increase the size of the file system on the software RAID device mounted at /mnt/point,
using one of the following methods:
Š To extend the file system size to the maximum available size of the device, enter
mount -o remount,resize /mnt/point
The resize option with no value specified grows the volume to the full size of the partition.
Š To extend the file system to a specific size, enter
mount -o remount,resize=size /mnt/point
Replace size with the desired size in blocks based on the block size used for the file
system.
For example, if you have a 4 GB device with a block size of 4KB, enter
mount -o remount,resize=1048576 /mnt/point
Wait until the resizing is completed before continuing.
3 Check the effect of the resize on the mounted file system by entering
df -h
The Disk Free (df) command shows the total size of the disk, the number of blocks used, and
the number of blocks available on the file system. The -h option print sizes in human-readable
format, such as 1K, 234M, or 2G.

Resizing Software RAID Arrays with mdadm 111

 novdocx (en) 13 May 2009
ReiserFS

As with Ext2 and Ext3, a ReiserFS file system can be increased in size while mounted or
unmounted. The resize is done on the block device of your RAID array.

1 Open a terminal console, then log in as the root user or equivalent.

2 Increase the size of the file system on the software RAID device called /dev/md0, using one of
the following methods:
Š To extend the file system size to the maximum available size of the device, enter
resize_reiserfs /dev/md0
When no size is specified, this grows the volume to the full size of the partition.
Š To extend the file system to a specific size, enter
resize_reiserfs -s size /dev/md0
Replace size with the desired size in bytes. You can also specify units on the value, such as
50000K (kilobytes), 250M (megabytes), or 2G (gigabytes). Alternatively, you can specify
an increase to the current size by prefixing the value with a plus (+) sign. For example, the
following command increases the size of the file system on /dev/md0 by 500 MB:
resize_reiserfs -s +500M /dev/md0
Wait until the resizing is completed before continuing.
3 If the file system is not mounted, mount it now.
For example, to mount an ReiserFS file system for a RAID named /dev/md0 at mount point /
raid, enter
mount -t reiserfs /dev/md0 /raid
4 Check the effect of the resize on the mounted file system by entering
df -h
The Disk Free (df) command shows the total size of the disk, the number of blocks used, and
the number of blocks available on the file system. The -h option print sizes in human-readable
format, such as 1K, 234M, or 2G.

8.3 Decreasing the Size of a Software RAID

Before you begin, review the guidelines in Section 8.1, “Understanding the Resizing Process,” on
page 107.
Š Section 8.3.1, “Decreasing the Size of the File System,” on page 112
Š Section 8.3.2, “Decreasing the Size of Component Partitions,” on page 114
Š Section 8.3.3, “Decreasing the Size of the RAID Array,” on page 115

8.3.1 Decreasing the Size of the File System

When decreasing the size of the file system on a RAID device, make sure the new size satisfies the
following conditions:
Š The new size must be greater than the size of the existing data; otherwise, data loss occurs.
Š The new size must be equal to or less than the current RAID size because the file system size
cannot extend beyond the space available.

112 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
In SUSE® Linux Enterprise Server SP1, only Ext2, Ext3, and ReiserFS provide utilities for
shrinking the size of the file system. Use the appropriate procedure below for decreasing the size of
your file system.

The procedures in this section use the device name /dev/md0 for the RAID device. Make sure to
modify commands to use the name of your own device.

Ext2 or Ext3

The Ext2 and Ext3 file systems can be resized when mounted or unmounted.

1 Open a terminal console, then log in as the root user or equivalent.

2 Decrease the size of the file system on the RAID by entering
resize2fs /dev/md0 <size>
Replace size with an integer value in kilobytes for the desired size. (A kilobyte is 1024 bytes.)
Wait until the resizing is completed before continuing.
3 If the file system is not mounted, mount it now. For example, to mount an Ext2 file system for
a RAID named /dev/md0 at mount point /raid, enter
mount -t ext2 /dev/md0 /raid
4 Check the effect of the resize on the mounted file system by entering
df -h
The Disk Free (df) command shows the total size of the disk, the number of blocks used, and
the number of blocks available on the file system. The -h option print sizes in human-readable
format, such as 1K, 234M, or 2G.

JFS

JFS does not support shrinking a volume.

ReiserFS

ReiserFS file systems can be shrunk only if the volume is unmounted.

1 Open a terminal console, then log in as the root user or equivalent.

2 Unmount the device by entering
umount /mnt/point
If the partition you are attempting to shrink contains system files (such as the root (/) volume),
unmounting is possible only when booting from a bootable CD or floppy.
3 Decrease the size of the file system on the software RAID device called /dev/md0 by entering
resize_reiserfs -s size /dev/md0
Replace size with the desired size in bytes. You can also specify units on the value, such as
50000K (kilobytes), 250M (megabytes), or 2G (gigabytes). Alternatively, you can specify a
decrease to the current size by prefixing the value with a minus (-) sign. For example, the
following command reduces the size of the file system on /dev/md0 by 500 MB:
resize_reiserfs -s -500M /dev/md0
Wait until the resizing is completed before continuing.
4 Mount the file system by entering

Resizing Software RAID Arrays with mdadm 113

 novdocx (en) 13 May 2009
mount -t reiserfs /dev/md0 /mnt/point
5 Check the effect of the resize on the mounted file system by entering
df -h
The Disk Free (df) command shows the total size of the disk, the number of blocks used, and
the number of blocks available on the file system. The -h option print sizes in human-readable
format, such as 1K, 234M, or 2G.

8.3.2 Decreasing the Size of Component Partitions

Resize the RAID’s component partitions one at a time. For each component partition, you remove it
from the RAID, modify its partition size, return the partition to the RAID, then wait until the RAID
stabilizes. While a partition is removed, the RAID operates in degraded mode and has no or reduced
disk fault tolerance. Even for RAIDs that can tolerate multiple concurrent disk failures, you should
never remove more than one component partition at a time.

WARNING: If a RAID does not have disk fault tolerance, or it is simply not consistent, data loss
results if you remove any of its partitions. Be very careful when removing partitions, and make sure
that you have a backup of your data available.

The procedure in this section uses the device names shown in the following table. Make sure to
modify the commands to use the names of your own devices.

Table 8-4 Scenario for Increasing the Size of Component Partitions

RAID Device Component Partitions

/dev/md0 /dev/sda1

/dev/sdb1
/dev/sdc1

To resize the component partitions for the RAID:

1 Open a terminal console, then log in as the root user or equivalent.

2 Make sure that the RAID array is consistent and synchronized by entering
cat /proc/mdstat
If your RAID array is still synchronizing according to the output of this command, you must
wait until synchronization is complete before continuing.
3 Remove one of the component partitions from the RAID array. For example, to remove /dev/
sda1, enter
mdadm /dev/md0 --fail /dev/sda1 --remove /dev/sda1
In order to succeed, both the fail and remove actions must be done.
4 Increase the size of the partition that you removed in Step 3 by doing one of the following:
Š Use a disk partitioner such as fdisk, cfdisk, or parted to increase the size of the
partition.

114 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Š Replace the disk on which the partition resides with a different device.
This option is possible only if no other file systems on the original disk are accessed by the
system. When the replacement device is added back into the RAID, it takes much longer
to synchronize the data.
5 Re-add the partition to the RAID array. For example, to add /dev/sda1, enter
mdadm -a /dev/md0 /dev/sda1
Wait until the RAID is synchronized and consistent before continuing with the next partition.
6 Repeat Step 2 through Step 5 for each of the remaining component devices in the array. Make
sure to modify the commands for the correct component partition.
7 If you get a message that tells you that the kernel could not re-read the partition table for the
RAID, you must reboot the computer after resizing all of its component partitions.
8 Continue with Section 8.3.3, “Decreasing the Size of the RAID Array,” on page 115.

8.3.3 Decreasing the Size of the RAID Array

After you have resized each of the component partitions in the RAID, the RAID array configuration
continues to use the original array size until you force it to be aware of the newly available space.
You can specify a size for the RAID or use the maximum available space.

The procedure in this section uses the device name /dev/md0 for the RAID device. Make sure to
modify commands to use the name of your own device.

1 Open a terminal console, then log in as the root user or equivalent.

2 Check the size of the array and the device size known to the array by entering
mdadm -D /dev/md0 | grep -e "Array Size" -e "Device
Size"
3 Do one of the following:
Š Increase the size of the array to the maximum available size by entering
mdadm --grow /dev/md0 -z max
Š Increase the size of the array to a specified value by entering
mdadm --grow /dev/md0 -z size
Replace size with an integer value in kilobytes for the desired size. (A kilobyte is 1024
bytes.)
4 Recheck the size of your array and the device size known to the array by entering
mdadm -D /dev/md0 | grep -e "Array Size" -e "Device
Size"
5 Do one of the following:
Š If your array was successfully resized, you are done.
Š If your array was not resized as you expected, you must reboot, then try this procedure
again.

Resizing Software RAID Arrays with mdadm 115

 novdocx (en) 13 May 2009

116 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Installing and Managing DRBD
9

Services 9
This section describes how to install, configure, and manage a device-level software RAID 1 across
a network using DRBD* (Distributed Replicated Block Device) for Linux.
Š Section 9.1, “Understanding DRBD,” on page 117
Š Section 9.2, “Installing DRBD Services,” on page 117
Š Section 9.3, “Configuring the DRBD Service,” on page 118
Š Section 9.4, “Testing the DRBD Service,” on page 119
Š Section 9.5, “Troubleshooting DRBD,” on page 120
Š Section 9.6, “Additional Information,” on page 121

9.1 Understanding DRBD

DRBD allows you to create a mirror of two block devices that are located at two different sites
across an IP network. When used with HeartBeat 2 (HB2), DRBD supports distributed high-
availability Linux clusters.

IMPORTANT: The data traffic between mirrors is not encrypted. For secure data exchange, you
should deploy a virtual private network (VPN) solution for the connection.

Data on the primary device is replicated to the secondary device in a way that ensures that both
copies of the data are always identical.

By default, DRBD uses the TCP port 7788 for communications between DRBD nodes.

The open source version of DRBD supports up to 4 TB as maximal overall size of all devices. If you
need a device larger than 4 TB, you can use DRBD+, which is commercially available from Linbit*.

You must set up the DRBD devices before creating file systems on them. Everything to do with user
data should be done solely via the /dev/drbd<n> device, not on the raw device, because DRBD
uses the last 128 MB of the raw device for metadata.

IMPORTANT: Make sure to create file systems only on the /dev/drbd<n> device, not on the raw
device.

For example, if the raw device is 1024 MB in size, the DRBD device has only 896 MB available for
data, with 128 MB hidden and reserved for the metadata. Any attempt to access the space between
896 MB and 1024 MB fails because it is not available for user data.

9.2 Installing DRBD Services

1 Install the High Availability (HA) pattern on both SUSE Linux Enterprise Servers in your
networked cluster. Installing HA also installs the drbd program files.
1a Log in as the root user or equivalent, then open YaST.

Installing and Managing DRBD Services 117

 novdocx (en) 13 May 2009
1b Choose Software > Software Management.
1c Change the filter to Patterns.
1d Under Base Technologies, select High Availability.
1e Click Accept.
2 Install the drbd kernel modules on both servers.
2a Log in as the root user or equivalent, then open YaST.
2b Choose Software > Software Management.
2c Change the filter to Search.
2d Type drbd, then click Search.
2e Select all of the drbd-kmp-* packages.
2f Click Accept.

9.3 Configuring the DRBD Service

NOTE: The following procedure uses the server names node 1 and node 2, and the cluster resource
name r0. It sets up node 1 as the primary node. Make sure to modify the instructions to use your own
node and file names.

1 Log in as the root user or equivalent on each node.

2 Open the /etc/drbd.conf file on the primary node (node1) in a text editor, modify the
following parameters in the on hostname {} sections, then save the file.
Š hostname
Š device
Š disk
Š address
Š meta-disk
All of these options are explained in the examples in the /usr/share/doc/packages/drbd/
drbd.conf file and in the man page of drbd.conf(5).
3 Copy the /etc/drbd.conf file to the /etc/drbd.conf location on the secondary server
(node 2).
scp /etc/drbd.conf <node 2> /etc
4 Initialize and start the drbd service on both systems by entering the following on each node:
rcdrbd start
5 Configure node1 as the primary node by entering the following on node1:
drbdsetup /dev/drbd0 primary --do-what-I-say

NOTE: The --do-what-i-say option has been renamed to --overwrite-data-of-peer in the recent
versions of DRBD.

6 Check the DRBD service status by entering the following on each node:
rcdrbd status
Before proceeding, wait until the block devices on both nodes are fully synchronized. Repeat
the rcdrbd status command to follow the synchronization progress.

118 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
7 After the block devices on both nodes are fully synchronized, format the DRBD device on the
primary with a file system such as reiserfs. Any Linux file system can be used. For example,
enter
mkfs.reiserfs -f /dev/drbd0

IMPORTANT: Always use the /dev/drbd<n> name in the command, not the actual /dev/
disk device name.

9.4 Testing the DRBD Service

If the install and configuration procedures worked as expected, you are ready to run a basic test of
the drbd functionality. This test also helps with understanding how the software works.

1 Test the DRBD service on node 1.

1a Open a terminal console, then log in as the root user or equivalent.
1b Create a mount point on node 1, such as /r0mount, by entering
mkdir /r0mount
1c Mount the drbd device by entering
mount -o rw /dev/drbd0 /r0mount
1d Create a file from the primary node by entering
touch /r0mount/from_node1
2 Test the DRBD service on node 2.
2a Open a terminal console, then log in as the root user or equivalent.
2b Dismount the disk on node 1 by typing the following command on node 1:
umount /r0mount
2c Downgrade the DRBD service on node 1 by typing the following command on node 1:
drbdadm secondary r0
2d On node 2, promote the DRBD service to primary by entering
drbdadm primary r0
2e On node 2, check to see if node 2 is primary by entering
rcdrbd status
2f On node 2, create a mount point such as /r0mount, by entering
mkdir /r0mount
2g On node 2, mount the DRBD device by entering
mount -o rw /dev/drbd0 /r0mount
2h Verify that the file you created on node 1 in Step 1d is viewable by entering
ls /r0mount
The /r0mount/from_node1 file should be listed.
3 If the service is working on both nodes, the DRBD setup is complete.
4 Set up node 1 as the primary again.
4a Dismount the disk on node 2 by typing the following command on node 2:
umount /r0mount

Installing and Managing DRBD Services 119

 novdocx (en) 13 May 2009
4b Downgrade the DRBD service on node 2 by typing the following command on node 2:
drbdadm secondary r0
4c On node 1, promote the DRBD service to primary by entering
drbdadm primary r0
4d On node 1, check to see if node 1 is primary by entering
service drbd status
5 To get the service to automatically start and fail over if the server has a problem, you can set up
DRBD as a high availability service with HeartBeat 2.
For information about installing and configuring HeartBeat 2 for SUSE® Linux Enterprise
Server 10, see the HeartBeat 2 Installation and Setup Guide (http://www.novell.com/
documentation/sles10/hb2/data/hb2_config.html) on the Novell Documentation Web site for
SUSE Linux Enterprise Server 10 (http://www.novell.com/documentation/sles10).

9.5 Troubleshooting DRBD

Use the following to troubleshoot problems with your DRBD setup:

Š Section 9.5.1, “Configuration,” on page 120

Š Section 9.5.2, “Host Names,” on page 120
Š Section 9.5.3, “TCP Port 7788,” on page 121
Š Section 9.5.4, “The --do-what-i-say Option,” on page 121

9.5.1 Configuration
If the initial drbd setup does not work as expected, there is probably something wrong with your
configuration.

To get information about the configuration:

1 Open a terminal console, then log in as the root user or equivalent.

2 Test the configuration file by running drbdadm with the -d option. Enter
drbdadm -d adjust r0
In a dry run of the adjust option, drbdadm compares the actual configuration of the DRBD
resource with your DRBD configuration file, but it does not execute the calls. Review the
output to make sure you know the source and cause of any errors.
3 If there are errors in the drbd.conf file, correct them before continuing.
4 If the partitions and settings are correct, run drbdadm again without the -d option. Enter
drbdadm adjust r0
This applies the configuration file to the DRBD resource.

9.5.2 Host Names

Please note that for DRBD, hostnames are case sensitive and therefore Node0 would be a different
host than node0.

120 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
9.5.3 TCP Port 7788
If your system is unable to connect to the peer, this also may be a problem of a local firewall. By
default, DRBD uses the TCP port 7788 to access the other node. Make sure that this port is
accessible on both nodes.

9.5.4 The --do-what-i-say Option

The --do-what-i-say option has been renamed to --overwrite-data-of-peer in the recent versions of
DRBD.

9.6 Additional Information

The following open-source resources are available for DRBD:

Š The following man pages for DRBD are available in the distribution:

drbd(8)
drbddisk(8)
drbdsetup(8)
drbdadm(8)
drbd.conf(5)
Š DRBD.org (http://www.drbd.org)
Š DRBD references at the Linux High-Availability Project (http://linux-ha.org/DRBD) by the
Linux High-Availability Project

For information about installing and configuring HeartBeat 2 for SUSE® Linux Enterprise Server
10, see the HeartBeat 2 Installation and Setup Guide (http://www.novell.com/documentation/
sles10/hb2/data/hb2_config.html) on the Novell Documentation Web site for SUSE Linux
Enterprise Server 10 (http://www.novell.com/documentation/sles10).

Installing and Managing DRBD Services 121

 novdocx (en) 13 May 2009

122 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Troubleshooting Storage Issues
10

This section describes how to work around known issues for EVMS devices, software RAIDs,
10
multipath I/O, and volumes.
Š Section 10.1, “Is DM-MP Available for the Boot Partition?,” on page 123
Š Section 10.2, “Rescue System Cannot Find Devices That Are Managed by EVMS,” on
page 123
Š Section 10.3, “Volumes on EVMS Devices Do Not Appear After Reboot,” on page 123
Š Section 10.4, “Volumes on EVMS Devices Do Not Appear When Using iSCSI,” on page 124
Š Section 10.5, “Device Nodes Are Not Automatically Re-Created on Restart,” on page 124

10.1 Is DM-MP Available for the Boot Partition?

In the initial release of SUSE® Linux Enterprise Server 10, DM-MP is not available for the boot
partition, because the boot loader cannot handle multipath I/O. Therefore, we recommend you set up
a separate boot (/boot) partition when using multipathing. This issue has been resolved in Support
Pack 1 and later.

10.2 Rescue System Cannot Find Devices That

Are Managed by EVMS
The Linux rescue system does not automatically activate volume manager support for LVM or
EVMS. For example, if you are using EVMS as the volume manager for your system device, you
might not be able to see the device in order to mount it under the rescue system.

You must manually activate EVMS in order for the Linux rescue system to detect system devices
that are managed by EVMS.

1 At the terminal console prompt, enter the following as the root user:
evms_activate

10.3 Volumes on EVMS Devices Do Not Appear

After Reboot
If you create volumes on EVMS devices and they cannot be found after you reboot the server, run
chkconfig to make sure that evms_activate runs before FSTAB.

1 At a terminal console prompt, enter either

chkconfig evms on
or
chkconfig boot.evms on
This ensures that evms_activate runs before FSTAB each time your servers reboot.

Troubleshooting Storage Issues 123

 novdocx (en) 13 May 2009
10.4 Volumes on EVMS Devices Do Not Appear
When Using iSCSI
If you have installed and configured an iSCSI SAN, and have created and configured EVMS disks
or volumes on that iSCSI SAN, your EVMS volumes might not be visible or accessible after reboot.
This problem is caused by EVMS starting before the iSCSI service. iSCSI must be started and
running before any disks or volumes on the iSCSI SAN can be accessed.

To resolve this problem, use the chkconfig command at the Linux server console of every server
that is part of your iSCSI SAN to correct the order that iSCSI and EVMS are started.
1 At a terminal console prompt, enter
chkconfig boot.evms on
This ensures that EVMS and iSCSI start in the proper order each time your servers reboot.

10.5 Device Nodes Are Not Automatically Re-

Created on Restart
Effective in SUSE® Linux Enterprise 10, the /dev directory is on tmpfs and the device nodes are
automatically re-created on boot. It is no longer necessary to modify the /etc/init.d/boot.evms
script to delete the device nodes on system reboot as was required for previous versions of SUSE
Linux.

The following procedure is provided for users who might encounter this issue for SUSE Linux
Enterprise Server 9 and earlier:

1 Open the /etc/init.d/boot.evms script in a text editor.

2 Add the following lines to the Stop section:
mount -n -o remount,rw /
echo -en "\nDeleting devices nodes"
rm -rf /dev/evms
mount -n -o remount,ro /
For example, the Stop section looks like this after the edit:
stop)
echo -n "Stopping EVMS"
mount -n -o remount,rw /
echo -en "\nDeleting devices nodes"
rm -rf /dev/evms
mount -n -o remount,ro /
rc_status -v
;;
3 Save the file.
4 Continue with “Restart the Server” on page 22.

124 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
Documentation Updates
A

This section contains information about documentation content changes made to the SUSE Linux
A
Enterprise Server Storage Administration Guide since the initial release of SUSE® Linux Enterprise
Server 10. If you are an existing user, review the change entries to readily identify modified content.
If you are a new user, simply read the guide in its current state.

Refer to the publication date, which appears on the title page, to determine the release date of this
guide. For the most recent version of the SUSE Linux Enterprise Server Storage Administration
Guide, see the Novell documentation Web site for SUSE Linux Enterprise Server 10 (http://
www.novell.com/documentation/sles10/#additional).

In this section, content changes appear in reverse chronological order, according to the publication
date. Within a dated entry, changes are grouped and sequenced alphabetically. Each change entry
provides a link to the related topic and a brief description of the change.

This document was updated on the following dates:

Š Section A.1, “May 15, 2009,” on page 125
Š Section A.2, “November 24, 2008,” on page 126
Š Section A.3, “June 10, 2008,” on page 126
Š Section A.4, “March 20, 2008 (SLES 10 SP2),” on page 127

A.1 May 15, 2009

Updates were made to the following section. The changes are explained below.

Š Section A.1.1, “Managing Multipath I/O,” on page 125

A.1.1 Managing Multipath I/O

The following changes were made to this section:

Location Change

Section 5.8, “Configuring Multipath I/O DM-MP is available but is not supported for /boot and /
for the Root Device,” on page 67 root in SUSE Linux Enterprise Server 10 SP1 and later if all
online updates have been applied. More specifically, you
need mkinitrd 1.2-106.61 and multipath-tools 0.4.7-
34.23 or later. However, if you install the packages and set up
the configuration, you might run into update issues later.

Full multipath support is available in SUSE Linux Enterprise

Server 11.

Documentation Updates 125

 novdocx (en) 13 May 2009
A.2 November 24, 2008
Updates were made to the following sections. The changes are explained below.

Š Section A.2.1, “Managing Multipath I/O,” on page 126

Š Section A.2.2, “Using UUIDs to Mount Devices,” on page 126

A.2.1 Managing Multipath I/O

The following changes were made to this section:

Location Change

Section 5.2.7, “Supported Storage Arrays After you modify the /etc/multipath.conf file, you must
for Multipathing,” on page 47 run mkinitrd to re-create the INITRD on your system, then
reboot in order for the changes to take effect.
“Applying Changes Made to the /etc/
multipath.conf File” on page 58

A.2.2 Using UUIDs to Mount Devices

The following change was made to this section:

Location Change

Section 3.2.1, “Using UUIDs to Assemble The device ID assigned by the manufactuer for a drive never
or Activate File System Devices,” on changes, no matter where the device is mounted, so it can
page 32 always be found at boot. The UUID is a property of the
filesystem and can change if you reformat the drive.

A.3 June 10, 2008

Updates were made to the following section. The changes are explained below.

Š Section A.3.1, “Managing Multipath I/O,” on page 126

A.3.1 Managing Multipath I/O

The following changes were made to this section:

Location Change

Section 5.3.3, “Using mdadm for For information about modifying the /etc/lvm/lvm.conf
Multipathed Devices,” on page 51 file, see Section 5.2.3, “Using LVM2 on Multipath Devices,” on
page 46.

“Verifying the Setup in the etc/ Use the -v3 option to show the full path list.
multipath.conf File” on page 56

126 SLES 10 SP2: Storage Administration Guide

 novdocx (en) 13 May 2009
A.4 March 20, 2008 (SLES 10 SP2)
Updates were made to the following section. The changes are explained below.

Š Section A.4.1, “Managing Multipath I/O,” on page 127

A.4.1 Managing Multipath I/O

The following changes were made to this section:

Location Change

Section 5.2.6, “Supported Architectures This section is new.

for Multipath I/O,” on page 47

Section 5.2.7, “Supported Storage Arrays This section was reorganized for clarity.
for Multipathing,” on page 47

Section 5.4.5, “Creating and Configuring The /etc/multipath.conf file does not exist unless you
the /etc/multipath.conf File,” on page 55 create it.

Section 5.6, “Configuring Path Failover This section is new.

Policies and Priorities,” on page 59

Section 5.8, “Configuring Multipath I/O DM-MP is now available and supported for /boot and /root
for the Root Device,” on page 67 in SUSE Linux Enterprise Server 10 SP1 and later if all online
updates have been applied. More specifically, you need
mkinitrd 1.2-106.61 and multipath-tools 0.4.7-34.23
or later.

Documentation Updates 127

 novdocx (en) 13 May 2009

128 SLES 10 SP2: Storage Administration Guide