SAP SUM DMO
Table of Contents
1. Introduction to SUM and SUM DMO
o Overview of SUM
o Overview of SUM DMO
o Differences Between SUM and SUM DMO
2. SUM DMO with System Move
o Database Migration
o Application Upgrade
o System Move
3. Key Concepts in SUM DMO
o Pipe Mode vs. File Mode in R3load
o Parallel Export/Import
o Custom Code Adjustments
o Handling Large Databases
4. Detailed Scenario: Migrating a 16TB Oracle Database to HANA
o System Resource Assessment
o Calculating Parallel Export/Import Jobs
o Using R3ta for Large Table Splitting
5. SUM vs. SUM DMO: When to Use Each
o Application Upgrade with No Database Migration
o Integrated Database Migration and Application Upgrade
6. Step-by-Step Procedure for SUM DMO with System Move
o Initial Preparation
o Execution Phase
o Post-Migration Activities
7. SUM DMO Variants
o Standard DMO
o DMO with System Move
o DMO without System Update
o DMO with Downtime Optimization
8. Common Issues and Troubleshooting
o Log Analysis
o Performance Bottlenecks
9. Conclusion
o Best Practices
o Final Thoughts
1. Introduction to SUM and SUM DMO
Overview of SUM (Software Update Manager)
The Software Update Manager (SUM) is SAP’s standard tool for performing system
maintenance tasks such as upgrades, updates, and patching of SAP systems. SUM is used to
manage changes to the application layer, including upgrading the SAP system to a new
enhancement package, applying support package stacks, or performing a release upgrade.
Typical Scenarios:
o Upgrading from SAP ECC 6.0 EHP6 to EHP7.
o Applying Support Package Stacks.
o Patching the system for bug fixes or enhancements.
Overview of SUM DMO (Database Migration Option)
SUM DMO extends the capabilities of SUM by incorporating database migration into the
upgrade process. This is particularly useful when moving an SAP system from a legacy
database platform (e.g., Oracle, SQL Server) to SAP HANA. With SUM DMO, both the
database migration and the application upgrade can be performed in a single, integrated
process.
Typical Scenarios:
Types of SUM DMO and Their Use Cases
o SUM DMO offers several migration paths tailored to different scenarios:
o Homogeneous System Copy (HSC): This involves migrating an existing SAP
system to a new environment while retaining the same database platform.
HSC is typically used for hardware refreshes, system consolidations, or disaster
recovery setups.
o Heterogeneous System Copy (HSC): Here, the migration includes a change in
the underlying database platform. A prime example is migrating from Oracle
or DB2 to SAP HANA, unlocking the benefits of in-memory computing and
advanced analytics.
o System Conversion: This scenario combines a system upgrade or update with
a database migration. It streamlines the process of modernizing your SAP
landscape by performing both tasks simultaneously.
Differences between SUM and SUM DMO
SUM: Focuses solely on application upgrades and maintenance tasks without
changing the underlying database.
SUM DMO: Combines database migration with application upgrades, enabling a
streamlined transition to SAP HANA or other supported databases.
2. SUM DMO with System Move
SUM DMO with System Move is an advanced option that allows for the migration of both the
database and the system itself (e.g., moving to a new data center, cloud, or hardware) during
the upgrade process. This option is particularly useful when upgrading the application to a
newer version while also changing the physical environment.
Database Migration
Purpose: Migrate the existing database to SAP HANA (or another target database).
Process: The data is exported from the source database (e.g., Oracle) and imported
into the target database (HANA) on the new system.
Application Upgrade
Purpose: Upgrade the SAP application layer (e.g., from ECC 6.0 EHP7 to EHP8 or to
S/4HANA).
Process: The SUM tool handles the application upgrade as part of the overall
migration, ensuring that the system is updated to the latest version during the
database migration.
System Move
Purpose: Physically move the SAP system from one environment to another (e.g., on-
premise to cloud, or old hardware to new).
Process: SUM DMO with System Move facilitates this transition, ensuring that both
the database migration and application upgrade are executed seamlessly on the new
system.
3. Key Concepts in SUM DMO
Pipe Mode vs. File Mode in R3load
R3load is a utility used for exporting and importing data during SAP migrations. Traditionally,
R3load operates in two modes:
Pipe Mode: Data is transferred directly between the export and import processes
using pipes (inter-process communication channels). This eliminates the need to write
intermediate data files to disk, resulting in faster and more efficient migrations.
File Mode: Data is written to disk as export files and then read from disk during the
import phase. While still supported, File Mode is generally slower than Pipe Mode
due to the additional disk I/O overhead.
In the latest SUM DMO versions, the choice between these modes is managed internally by
SUM based on system configuration and the migration scenario. The preferred R3load type
for most scenarios is Pipe Mode
Parallel Export/Import
Parallel processing is critical for optimizing migration time, especially for large databases.
SUM DMO manages multiple parallel export and import jobs, leveraging system resources
such as CPU, memory, and disk I/O to minimize downtime.
Handling Large Databases
For large databases (e.g., 16TB), special considerations such as table splitting (using R3ta) and
parallel processing are required to efficiently handle the migration within acceptable
downtime limits.
4. Detailed Scenario: Migrating a 16TB Oracle Database to HANA
System Resource Assessment
CPU: Determine the number of available cores on both the source (Oracle) and target
(HANA) systems.
Memory: Assess total RAM and ensure sufficient memory for parallel processing.
Disk I/O: Evaluate disk throughput to handle the export/import processes.
Network Bandwidth: Ensure the network can support the data transfer rates required
for the migration.
Calculating Parallel Export/Import Jobs
Based on the system resources, calculate the number of parallel jobs that can be run without
overloading the system. This involves balancing CPU, memory, and disk I/O capacity.
Example Calculation:
o Available CPU Cores: 64
o Memory Available for R3load: 384 GB
o Disk Throughput: 3000 MB/s
o Set MaxLoadJobs=36 and MaxImportJobs=36 to optimize performance.
Using R3ta for Large Table Splitting
For large tables like BSEG, use R3ta to split the table into smaller segments. This allows
R3load to process these segments in parallel, reducing the overall migration time.
For very large tables (e.g., BSEG, COEP), use the R3ta tool to split these tables into smaller
segments. This allows R3load to handle these tables in parallel, reducing the overall export
time.
Command: R3ta -t BSEG -f -p 8 -s "/path/to/control/file/BSEG_split" -c BELNR
This command splits the BSEG table based on the BELNR field into segments that can be
processed by 8 parallel jobs.
5. SUM vs. SUM DMO: When to Use Each
Application Upgrade with No Database Migration
If the database remains unchanged (e.g., same HANA version), use SUM for application
upgrades. SUM will handle the necessary application updates without altering the database.
Example: Upgrading from ECC 6.0 EHP7 to EHP8 on the same HANA database.
Integrated Database Migration and Application Upgrade
When migrating the database and upgrading the application simultaneously, use SUM DMO.
This tool manages both processes in an integrated manner, ensuring a smooth transition to
the new environment.
Example: Migrating from Oracle to HANA while converting ECC to S/4HANA.
6. Step-by-Step Procedure for SUM DMO with System Move
Initial Preparation
Data Archiving: Archive old data to reduce the size of the database before migration.
System Sizing: Perform sizing exercises for both the source and target systems to
ensure they can handle the migration load.
Execution Phase
Start SUM DMO: Launch SUM DMO on the source system to begin the migration
process.
System Move: Transfer the SUM directory to the target system, extract the
appropriate SUM version, and continue the migration.
Parallel Processing: Configure and monitor parallel export/import jobs to optimize
performance.
Post-Migration Activities
Testing: Conduct thorough testing of the migrated system to ensure all functionalities
are working correctly.
Performance Tuning: Optimize the new HANA system for performance, adjusting
parameters as needed.
Go-Live: After validation, move the system to production.
7. SUM DMO Variants
Standard DMO
Purpose: Perform a database migration and application upgrade in one step.
Use Case: Migrating from Oracle to HANA while upgrading from ECC 6.0 EHP7 to
EHP8.
DMO with System Move
Purpose: Migrate the database and move the system to new hardware or a new
environment.
Use Case: Moving an ECC system from on-premise Oracle to cloud-based HANA while
upgrading to S/4HANA.
DMO without System Update
Purpose: Migrate the database without upgrading the application.
Use Case: Moving from Oracle to HANA while keeping the application on ECC 6.0
EHP7.
DMO with Downtime Optimization
Purpose: Minimize system downtime during migration.
Use Case: Critical systems where downtime must be kept to an absolute minimum,
such as production systems in industries with high availability requirements.
Key Features:
o Near-Zero Downtime: SUM DMO with Downtime Optimization enables the
migration of data during the uptime of the system, leaving only the final
cutover phase to be performed during a brief downtime window.
o Parallel Data Transfer: Like standard DMO, this variant also leverages parallel
processing to transfer large volumes of data efficiently.
o Shadow Operations: A shadow system may be used to prepare much of the
migration while the main system remains operational, further reducing the
required downtime.
8. Common Issues and Troubleshooting
Even with meticulous planning, issues can arise during the SUM DMO process. Understanding
common problems and knowing how to troubleshoot them is essential for a successful
migration.
Log Analysis
SUM Logs: SUM generates detailed logs during the migration process. These logs can
be found in the abap/log/ directory and should be reviewed regularly to monitor
progress and identify issues.
o Key Logs: R3load.log, SLOG, and SUMSTART.LOG provide insights into
export/import activities, errors, and general SUM operations.
o Error Messages: Identify and resolve any errors or warnings promptly to avoid
disruptions in the migration process.
Performance Bottlenecks
Parallel Processing Issues: If parallel jobs are not performing as expected, check
system resources (CPU, memory, I/O) and adjust the number of parallel jobs.
Overcommitting resources can lead to performance degradation.
Disk I/O Bottlenecks: Ensure that the disk subsystem can handle the data throughput
required by multiple parallel R3load jobs. Consider increasing disk I/O capacity or
redistributing jobs to balance the load.
9. Conclusion
Best Practices
Thorough Planning: Begin with a detailed project plan that includes system sizing, data
archiving, and custom code assessment. Ensure that all stakeholders are aware of the
timeline and potential risks.
Pre-Migration Testing: Conduct extensive testing in a sandbox environment to identify
and resolve issues before they impact the production environment.
Regular Monitoring: During the SUM DMO process, continuously monitor system
performance, log files, and the progress of the migration. Be prepared to adjust
parameters or take corrective action as needed.
Post-Migration Validation: After the migration, thoroughly test the system in the new
environment to ensure that all functionalities are working correctly and that
performance is optimal.
Final Thoughts
The SUM DMO process, especially when combined with a system move and an application
upgrade, is a powerful tool for modernizing SAP landscapes. By carefully planning and
executing each step, you can minimize downtime, reduce risks, and ensure a smooth
transition to SAP HANA and S/4HANA.
