This document provides an overview of Cisco's Unified Computing System (UCS). It discusses how server deployment has evolved from rackmount servers to blade servers to UCS. UCS aims to unify computing, networking, and storage infrastructure and simplify management. It consists of fabric interconnects, chassis, blades, and fabric extenders. Service profiles allow logical assignment of resources for stateless server provisioning and mobility. Use cases include deployment, upgrades, provisioning, high availability, and disaster recovery.

1. Cisco Unified
Computing
Abdelkader YEDDES
Solution Architect

2. Agenda
How we got here ?
UCS introduction
UCS Components
Service Profile
Use cases

3. Server Deployment: Rackmounts
First generation
 Rack-optimized
 Top of Rack or End of Row
switches
 Cables
Benefits
 Space utilization
 Highly flexible
Weakness
 Cabling
 Serviceability
 Power efficiency
Server
Server
Server
Server
Server
Server
Server
Rack
Server
Server
Server
Server
Server
Server
Server
Rack

4. Second generation
• Blade servers
• Integrated switches
• Fixed backplane
Benefits
• Space utilization
• Cable aggregation
• Power efficient
Weakness
• I/O flexibility
• Management Aggregation
• Large chassis needed to
amortize switch/mgmt costs
Server Deployment: Blades
Rack Rack
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
ServerServer
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server

5. Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Management ManagementManagement Management
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server Deployment Management
Chassis Management
• New management layer
Benefits
• Consistency in chassis
• Shared chassis infrastructure
monitoring
Weakness
• Additional mgmt overhead
• Additional cost overhead
• Need chassis aggregation
management
• Artificial aggregation point

6. Management Management
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Server
Management Management
Server Deployment: Applications
Application Deployment
• Single application per OS
Benefits
• Isolation
• Flexibility
• Simplicity
Weakness
• Low utilization
• Power & Cooling
• Server sprawl

7. Management Management
Server Deployment
Software Switch Software Switch Software Switch

8. Infrastructure
• Servers & Switches
• Physical & Virtual
Challenges
• Many points of management
• Consistent policies
• Diagnostics
• Training
• Security
Management ManagementManagement Management
Server Deployment Virtualization
Software Switch Software Switch Software Switch

9. Management Management
Mgmt Server
Server Deployment Today
Over the past 10 years
• An evolution of size, not system
• More servers & switches than ever
• More switches per server
• Management applied, not integrated
Result
• More points of management
• More difficult to maintain policy
coherence
• More difficult to secure
• More difficult to scale02/03/20
17

10. Mgmt Server
Cisco Unified Computing System
 Embed management
 Unify fabrics
 Optimize virtualization
 Remove unnecessary
 switches,
 adapters,
 management modules
 Less than 1/3rd
infrastructure
Mgmt ServerMgmt Server

11. Mgmt Server
Cisco Unified Computing System
A single system that encompasses:
• Network: Unified fabric
• Compute: Industry standard x86
• Storage: Access options
• Virtualization optimized
Unified management model
• Dynamic resource provisioning
Efficient Scale
• Cisco network scale & services
• Fewer servers with more memory
Lower cost
• Fewer servers, switches, adapters, cables
• Lower power consumption
• Fewer points of management

12. UCS Manager
Embedded– manages entire system
UCS Fabric Interconnect
20 Port 10Gb FCoE
40 Port 10Gb FCoE
UCS Fabric Extender
Remote line card
UCS Blade Server Chassis
Flexible bay configurations
UCS Blade Server
Industry-standard architecture
UCS Virtual Adapters
Choice of multiple adapters
Building Blocks

13. Overall System (Rear)
Uplinks
Converged Network
• Uplink: 8x 10Gb
• Oversubscription 1:1
• Uplink: 2x 10Gb
• Oversubscription 1:8
• Uplink: 4x 10Gb
• Oversubscription 1:2

14. System Components
 Fabric Interconnect
(40 or 20 10GE ports) + (2 or 1 GEM slots)
 Chassis
Upto 8 half width blades or 4 full width blades
 Fabric Extender
Host to uplink traffic engineering
Up to 80Gb Flexible bandwidth allocation
 Compute Blade
 Adapter
Virtualized adapter for single OS and hypervisor
systems
Compute Chassis
x86 Computer x86 Computer
X
II
x8x8x8x8
B
MGMT
SS
B
X X X X X
CC
A
G G
G G
SAN
G
R
A
G
G G
G
R
G
PM P
SANLAN
Fabric
Interconnect
Fabric
Interconnect
Fabric
Extender
Fabric
Extender
Compute Blade
(Half slot)
Adapter
Compute Blade
(Full slot)
AdapterAdapter

15. UCS Fabric Interconnect Portfolio
40-Port Fabric Interconnect
• 40 fixed ports 10GE/FCoE, fixed
• 2 Expansion Modules
20-Port Fabric Interconnect
• 20 fixed ports 10GE/FCoE, fixed
• 1 Expansion Module
FC + Ethernet
• 4 Ports 10GbE/FCoE
• 4 Ports 1/2/4G FC
Fibre Channel
• 8 Ports 1/2/4G FC
Ethernet
• 6 Ports 10GE/FCoE
UCS 6100
Family
Expansion
Modules
Fibre Channel
• 6 Ports 1/2/4/8G FC

16. UCS 5108 Blade Chassis
 Up to 8 half slot blades
 Up to 4 full slot blades
 4x power supplies, N+N grid redundant
 8x fans included
 2x UCS 2104 Fabric Extender
 All items hot-pluggable
 Up to 40 chassis per UCS system
Chassis

17. B-Series Family Comparison
UCS B200 M4
General
Purpose
Blade Server
UCS B260 M4
Extended Memory
Blade Server
UCS B420 M1 High-
Performance Blade
Server
Memory-intensive
server for virtualized
and large-data-set
workloads
Compute & memory-
intensive server for
enterprise-critical
workloads
High-density server with
balanced compute
performance and I/O
flexibility
Item Size
CPU
Sock
ets
CPU Memory Disks
I/O
slots
UCS B200 M4 Half 2
Intel Xeon 5600
v3 or v4
24 DIMM
1.5 TB
2 SFF SAS 2 Mezz
UCS B260 M4 Full 2
Intel Xeon E7
v2, v3 or v4
48 DIMM
3 TB
2 SFF SAS 3 Mezz
UCS B420 M4 Full 4
Intel Xeon E5-
4600 v4 and v3
48 DIMM
3 TB
4 SFF
SAS/SATA
3 Mezz
UCS B460 M4 Full 4
Intel Xeon E7
v2, v3 or v4
96 DIMM
6 TB
2 SFF SAS 6 Mezz
UCS B460 M3
Compact -
Performance
Blade Server
Compute &
memory-intensive
server for
enterprise-critical
workloads

18. Three Pronged Adapter Strategy
 Converged network adapters (CNA)
 Ability to mix and match adapter types within a system
 Automatic discovery of component types
Virtual Machine Aware:
Virtualization and
Consolidation “Palo”
Existing Driver Stacks “Free” SAN Access for
Any Ethernet Equipped
Host
CostCompatibilityVirtualization

19. C-Series Product Details
Item CPU Size Memory Disks I/O
UCS C460 M1
Intel Xeon
7500
4RU
64 DIMM
512GB
12 SFF
SAS/SATA
10 PCIe
UCS C250 M2
Intel Xeon
5600
2RU
48 DIMM
384 GB
8 SFF
SAS/SATA
5 PCIe
UCS C210 M2
Intel Xeon
5600
2RU
12 DIMM
96 GB
16 SFF
SAS/SATA
5 PCIe
UCS C200 M2
Intel Xeon
5600
1RU
12 DIMM
96GB
4 x 3.5”
SAS/SATA
2 PCIe
UCS C200 M2
High-Density Rack-
Mount Server
UCS C210 M2
General-Purpose Rack-
Mount Server
UCS C250 M2 Extended
Memory Rack-Mount
Server
High-density server
with balanced compute
performance and I/O
flexibility
General-purpose server
for workloads requiring
economical, high-
capacity, internal storage
High-performance,
memory-intensive server
for virtualized and large-
data-set workloads
UCS C460 M1 High-
Performance Rack-
Mount Server
Compute & memory-
intensive server for
enterprise critical
workloads
Cisco’s Unified Computing System C-Series Rack-Mount servers
now address the vast majority of data center compute requirements

20. Unified Computing System Manager
 Embedded device manager for family of UCS components
 Enables stateless computing via Service Profiles
 Efficient scale: Same effort for 1 to 320 blades
 APIs for integration with new and existing data center infrastructure

21. Service Profiles
A service profile is a logical representation of a server that specifies its:
• Identity—UUID, MAC, WWN, etc.
• Configuration—Server Requirements, Boot Order, Firmware, etc.
• Connectivity—VLAN, VSAN, QoS, etc.
Service
Profile
Physical
ResourcesService
Profile
Template
Policies
Logical
Resources

22. Service Profile Components
Boot Policy
• Boot Devices
• Boot Order
vHBA
• Identity
• Fabric Connectivity
• Configuration
Operational Policy
• Scrub Policy
• Health Policy
• External Management Access
Server
Requirements
• Specific Blade
• Blade Pool
• Qualification Criteria
vNIC
• Identity
• Fabric connectivity
• High Availability
• QoS policy
• Configuration
Service Profile
• Identity (UUID, BIOS, etc)
• Local Storage Policy
• Firmware Update Group
• Stats Policies
vNIC
• Identity
• Fabric connectivity
• High Availability
• QoS policy
• Configuration
vNIC
• Identity (MAC)
• Fabric Connectivity
• High Availability
• QoS policy
• Configuration
vHBA
• Identity
• Fabric Connectivity
• Configuration
vHBA
• Identity (WWPN)
• Fabric Connectivity
• Configuration

23. 02/03/20
17
Server Mobility with Service Profiles
Compute Node A is
being upgraded.
Service Profile can be manually
migrated to Compute Node B.
Service Profile

24. Logical Server Mobility
Identity
LAN/SAN
Configuration
Service Profile:
WebServerA
Time A
Time B

25. Use Cases
• Server Deployment
• Server Upgrades:
• Within a single Cisco UCS system
• Across Cisco UCS systems
• Dynamic Server Provisioning
• High Availability
• Disaster Recovery

26. Server Availability
Service ProfilesCisco UCS enablers: Virtual Interface Card
Blade
Blade
Blade
Blade
Blade
Web
Blade
Blade
Blade
Blade
Blade
Oracle
Blade
Blade
Blade
Blade
Blade
VMware
Burst capacity
HA spare
Today’s deployment:
 Provisioned for peak
capacity
 Spare node per
workload
Blade
Blade
Blade
Blade
Blade
Web
Blade
Blade
Blade
Oracle
Blade
Blade
Blade
VMwareWith server profiles:
 Resources
provisioned as needed
 Same availability with
fewer spares

27. Questions