Principle of Cloud Computing

(314261)
Week
10

Dr.Hassan Al-Sukhni
 PRINCIPLES OF
CLOUD COMPUTING
Module Two Part 3

Src: www.favpng.com
Copyright of the Course Materials

The course’s materials are adapted from several online and

offline sources.

3
Outline

 Learning Objectives

 Datacenter Overview

 Datacenter Architecture

 Datacenter Challenges

 Cloud Computing Datacenters

4
Learning Objectives

 Understanding the concept of datacenters and their role in modern computing.

 Identifying the components and infrastructure of a typical datacenter and the
different types of datacenters.
 Understanding the design principles and best practices for building a datacenter
 Understanding the architecture and challenges of cloud datacenters.

5
DATACENTER OVERVIEW

6
Data Centers

 Data center (DC) is a physical facility that enterprises use to house computing and storage
infrastructure in a variety of networked formats.

 Main function is to deliver utilities needed

by the equipment and personnel:
 Power
 Cooling
 Shelter
 Security

 Size of typical data centers:

 500 – 5000 sqm buildings
 1 MW to 10-20 MW power (avg 5 MW)

7
Example data centers

8
 Datacenters around the globe

9
https://docs.microsoft.com/en-us/learn/modules/explore-azure-infrastructure/2-azure-datacenter-locations
Modern DC for the Cloud Architecture
 Geography:
- Two or more regions
- Meets data residency requirements
- Fault-tolerant from complete region failures

 Region:
- Set of datacenters within a metropolitan area
- Network latency perimeter < 2ms

 Availability Zones:
- Unique physical locations within a region
- Each zone made up of one or more DCs
- Independent power, cooling, networking
- Inter-AZ network latency < 2ms
- Fault tolerance from DC failure 10
Src: Inside Azure Datacenter Architecture with Mark Russinovich.
Data Centers
 Traditional data centers
 Host a large number of relatively small- or medium-sized applications, each running on a
dedicated hardware infrastructure that is decoupled and protected from other systems
in the same facility.
 Usually for multiple organizational units or companies.

 Modern data centers (a.k.a., Warehouse-scale computers)

 Usually belong to a single company to run a small number of large-scale applications
 Google, Facebook, Microsoft, Amazon, Alibaba, etc.
 Use relatively homogeneous hardware and system software.
 Share a common systems management layer.
 Sizes can vary depending on needs

11
DATACENTER ARCHITECTURE

12
Scale-up vs. scale-out
 Scale-up: high-cost powerful CPUs, more cores, more memory.
 Scale-out: adding more low cost, commodity servers.

 Supercomputer vs. data center

 Scale
 Blue waters = 40K 8-core “servers”
 Microsoft Chicago Data centers = 50 containers = 100K 8-core servers
 Network architecture
 Supercomputers: InfiniBand, low-latency, high bandwidth protocols
 Data Centers: (mostly) Ethernet-based networks
 Storage
 Supercomputers: separate data farm
 Data Centers: use the disk on node + memory cache
13
Main components of a datacenter

14
src: The Datacenter as a Computer – Barroso, Clidaras, Holzle
Traditional Data Center Architecture
Servers mounted on 19’’
rack cabinets

Racks are placed in single rows forming

corridors between them.

15
 Src: the datacenter as a computer – an introduction to the design of warehouse-scale machines
 A Row of Servers in a Google Data Center

 Src: the datacenter as a computer – an introduction to the design of warehouse-

scale machines
16
Inside a modern data center

 Today’s DC use shipping containers packed

with 1000s servers each.

 For repairs, whole containers are replaced.

17
Costs for operating a data center
 DCs consume 3% of global electricity Monthly cost = $3’530’920
supply (416.2 TWh > UK’s 300 TWh).

 DCs produce 2% of total greenhouse gas Servers

emissions. 1 4%
3
Networking Equipment
 DCs produce as much CO2 as The %
Netherlands 1 Power Distribution & Cooling
or Argenti. 8
% 57%
31% power Power
8
%
Other Infrastructure

45,978 servers, 3yr server & 10 yr infrastructure amortization

18
Power Usage Effectiveness (PUE)

 PUE is the ratio of

 The total amount of energy used by a DC facility
 To the energy delivered to the computing equipment

 PUE is the inverse of data center infrastructure efficiency

 Total facility power = covers IT systems (servers, network, storage) +

other equipment (cooling, UPS, switchgear, generators, lights, fans, etc.)

19
Achieving PUE
 Location of the DC – cooling and power load factor

 Raise temperature of aisles

 Usually 18-20 C; Google at 27 C
 Possibly up to 35 C (trade off failures vs. cooling costs)

 Reduce conversion of energy

 E.g., Google motherboards work at 12V
rather than 3.3/5V

 Go to extreme environments
 Arctic circle (Facebook)
 Floating boats (Google)
 Underwater DC (Microsoft)

 Reuse dissipated heat 20

Evolution of data center design
 Case study: Microsoft

21
https://www.nextplatform.com/2016/09/26/rare-tour-microsofts-hyperscale-datacenters/
Evolution of datacenter design
 Gen 6: scalable form factor (2017)
- Reduced infrastructure, scale to demand
- 1.17-1.19 PUE

 Gen 7: Ballard (2018)

- Design execution efficiency
- Flex capacity enabled
- 1.15-1.18 PUE

 Gen 8: Rapid deploy datacenter (2020)

- Modular construction and delivery
- Equipment skidding and preassembly
- Faster speed to market
 Project Natick (future) – 1.07 PUE or less
- Project Natick: (future) rapid deployment, close
to population centers, high energy density,
resistant to hurricanes, solar storms,
Src: Inside Azure Datacenter Architecture with Mark Russinovich 22
earthquake, 1.07 or less PUE