Ucs5108 Install
Ucs5108 Install
Americas Headquarters
Cisco Systems, Inc.
170 West Tasman Drive
San Jose, CA 95134-1706
USA
http://www.cisco.com
Tel: 408 526-4000
800 553-NETS (6387)
Fax: 408 527-0883
THE SPECIFICATIONS AND INFORMATION REGARDING THE PRODUCTS IN THIS MANUAL ARE SUBJECT TO CHANGE WITHOUT NOTICE. ALL STATEMENTS,
INFORMATION, AND RECOMMENDATIONS IN THIS MANUAL ARE BELIEVED TO BE ACCURATE BUT ARE PRESENTED WITHOUT WARRANTY OF ANY KIND,
EXPRESS OR IMPLIED. USERS MUST TAKE FULL RESPONSIBILITY FOR THEIR APPLICATION OF ANY PRODUCTS.
THE SOFTWARE LICENSE AND LIMITED WARRANTY FOR THE ACCOMPANYING PRODUCT ARE SET FORTH IN THE INFORMATION PACKET THAT SHIPPED WITH
THE PRODUCT AND ARE INCORPORATED HEREIN BY THIS REFERENCE. IF YOU ARE UNABLE TO LOCATE THE SOFTWARE LICENSE OR LIMITED WARRANTY,
CONTACT YOUR CISCO REPRESENTATIVE FOR A COPY.
The following information is for FCC compliance of Class A devices: This equipment has been tested and found to comply with the limits for a Class A digital device, pursuant to part 15
of the FCC rules. These limits are designed to provide reasonable protection against harmful interference when the equipment is operated in a commercial environment. This equipment
generates, uses, and can radiate radio-frequency energy and, if not installed and used in accordance with the instruction manual, may cause harmful interference to radio communications.
Operation of this equipment in a residential area is likely to cause harmful interference, in which case users will be required to correct the interference at their own expense.
The following information is for FCC compliance of Class B devices: This equipment has been tested and found to comply with the limits for a Class B digital device, pursuant to part 15 of
the FCC rules. These limits are designed to provide reasonable protection against harmful interference in a residential installation. This equipment generates, uses and can radiate radio
frequency energy and, if not installed and used in accordance with the instructions, may cause harmful interference to radio communications. However, there is no guarantee that interference
will not occur in a particular installation. If the equipment causes interference to radio or television reception, which can be determined by turning the equipment off and on, users are
encouraged to try to correct the interference by using one or more of the following measures:
• Connect the equipment into an outlet on a circuit different from that to which the receiver is connected.
Modifications to this product not authorized by Cisco could void the FCC approval and negate your authority to operate the product.
The Cisco implementation of TCP header compression is an adaptation of a program developed by the University of California, Berkeley (UCB) as part of UCB’s public domain version of
the UNIX operating system. All rights reserved. Copyright © 1981, Regents of the University of California.
NOTWITHSTANDING ANY OTHER WARRANTY HEREIN, ALL DOCUMENT FILES AND SOFTWARE OF THESE SUPPLIERS ARE PROVIDED "AS IS" WITH ALL FAULTS.
CISCO AND THE ABOVE-NAMED SUPPLIERS DISCLAIM ALL WARRANTIES, EXPRESSED OR IMPLIED, INCLUDING, WITHOUT LIMITATION, THOSE OF
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT OR ARISING FROM A COURSE OF DEALING, USAGE, OR TRADE PRACTICE.
IN NO EVENT SHALL CISCO OR ITS SUPPLIERS BE LIABLE FOR ANY INDIRECT, SPECIAL, CONSEQUENTIAL, OR INCIDENTAL DAMAGES, INCLUDING, WITHOUT
LIMITATION, LOST PROFITS OR LOSS OR DAMAGE TO DATA ARISING OUT OF THE USE OR INABILITY TO USE THIS MANUAL, EVEN IF CISCO OR ITS SUPPLIERS
HAVE BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.
Any Internet Protocol (IP) addresses and phone numbers used in this document are not intended to be actual addresses and phone numbers. Any examples, command display output, network
topology diagrams, and other figures included in the document are shown for illustrative purposes only. Any use of actual IP addresses or phone numbers in illustrative content is unintentional
and coincidental.
All printed copies and duplicate soft copies of this document are considered uncontrolled. See the current online version for the latest version.
Cisco has more than 200 offices worldwide. Addresses and phone numbers are listed on the Cisco website at www.cisco.com/go/offices.
Cisco and the Cisco logo are trademarks or registered trademarks of Cisco and/or its affiliates in the U.S. and other countries. To view a list of Cisco trademarks, go to this URL:
https://www.cisco.com/c/en/us/about/legal/trademarks.html. Third-party trademarks mentioned are the property of their respective owners. The use of the word partner does not imply a
partnership relationship between Cisco and any other company. (1721R)
© 2009–2020 Cisco Systems, Inc. All rights reserved.
CHAPTER 1
Overview
This chapter contains the following sections:
• System Overview, on page 1
• Features and Benefits, on page 2
• Chassis Components, on page 4
• LEDs, on page 19
System Overview
The Cisco UCS 5108 Server Chassis and its components are part of the Cisco Unified Computing System
(UCS), which uses the Cisco UCS 5108 server system with the two I/O modules and the Cisco UCS Fabric
Interconnects to provide advanced options and capabilities in server and data management. All servers are
managed via the GUI or CLI with Cisco UCS Manager.
The Cisco UCS 5108 Server Chassis system consists of the following components:
• Chassis versions:
• Cisco UCS 5108 server chassis–AC version (UCSB-5108-AC2 or N20-C6508)
• Cisco UCS 5108 server chassis–DC version (UCSB-5108-DC2 or UCSB-5108-DC)
Note You cannot mix UCS-IOM-2304V2 and UCS-IOM-2304 in the same chassis.
UCS-IOM-2304V2 requires Cisco UCS Manager 4.0(4) or later.
Cisco UCS Mini Server Chassis, which is a smaller solution, consists of the following components:
• Cisco UCS 5108 server chassis–AC version (UCSB-5108-AC2)
• Cisco UCS 5108 server chassis–DC version (UCSB-5108-DC2)
• Cisco UCS 6324 Fabric Interconnect for the UCS Mini system (UCS-FI-M-6324)—Up to two integrated
fabric interconnect modules, each providing four SFP+ ports of 10-Gigabit Ethernet and Fibre Channel
over Ethernet (FCoE), and a QSFP+ port. This FI fits into the I/O module slot on the rear of the chassis.
• A number of SFP+ choices using copper or optical fiber
• Power supplies (UCSB-PSU-2500ACDV, UCSB-PSU-2500DC48, and UCSB-PSU-2500HVDC)—Up
to four 2500 Watt, hot-swappable power supplies
• Fan modules (N20-FAN5)—Eight hot-swappable fan modules
• UCS B-Series blade servers, including the following:
• Cisco UCS B200 M4 or M5 blade servers—Up to eight half-width blade servers, each containing
two CPUs and holding up to two hard drives capable of RAID 0 or 1
This simplicity eliminates the need for dedicated chassis management and blade switches, reduces cabling,
and enables the Cisco Unified Computing System to scale to 40 chassis without adding complexity. The Cisco
UCS 5108 server chassis is a critical component in delivering the Cisco Unified Computing System benefits
of data center simplicity and IT responsiveness.
Feature Benefit
Management by Cisco Reduces total cost of ownership by removing management modules from the
UCS Manager chassis, making the chassis stateless.
Provides a single, highly available management domain for all system chassis,
reducing administrative tasks.
Unified fabric Decreases TCO by reducing the number of network interface cards (NICs), host
bus adapters (HBAs), switches, and cables needed.
Support for one or two Eliminates switches from the chassis, including the complex configuration and
UCS I/O Modules (or management of those switches, allowing a system to scale without adding
support for one or two complexity and cost.
UCS 6324 Fabric
Allows use of two I/O modules for redundancy or aggregation of bandwidth.
Interconnects in the UCS
Mini chassis Enables bandwidth scaling based on application needs; blades can be configured
from 1.25 Gbps to 40 Gbps or more.
Auto discovery Requires no configuration; like all components in the Cisco Unified Computing
System, chassis are automatically recognized and configured by Cisco UCS
Manager.
High-performance Provides investment protection for new fabric extenders and future blade servers.
midplane
Supports up to 2x 40 Gigabit Ethernet for every blade server slot.
Provides 8 blades with 1.2 Tbps of available Ethernet throughput for future I/O
requirements. (The Cisco UCS 6324 Fabric Interconnect in the UCS Mini chassis
supports only 512 Gbps.)
Provides reconfigurable chassis to accommodate a variety of form factors and
functions.
Hot-pluggable blade Provides uninterrupted service during maintenance and server deployment.
servers, FEXes, and
fabric interconnects
Feature Benefit
Efficient front-to-back Helps reduce power consumption and increase component reliability.
airflow
Mixed blade Allows up to 8 half-width or 4 full-width blade servers, or any combination thereof,
configurations for outstanding flexibility. (When the UCS Mini chassis is configured with the
6324 Fabric Interconnect, only 8 half-width blades are supported.)
Chassis Components
This section lists an overview of chassis components.
LEDs
LEDs on the chassis indicate system connectivity and failure warnings. See LED Locations, on page 20 for
details. There is also a flashing blue Beaconing LED and button that can be triggered manually or remotely
from UCS Manager.
Buttons
The beaconing function LED is also a button. When triggered, beaconing of the server chassis is observable
remotely from UCS Manager or by the flashing blue button/LED on the chassis.
Connectors
There are no user connectors such as RJ-45 ports on the chassis itself.
Midplane
The integral chassis midplane supports the following:
• 320 G total bandwidth to each of two I/O Modules
• Auto-discover of all components
• Redundant data and management paths
• 10 G Base-KR
Blade Servers
The Cisco UCS B-Series Blade Servers are based on industry-standard server technologies and provide the
following:
• Up to two or four Intel multi-core processors, depending on the server
• Front-accessible, hot-swappable hard drives or solid-state disk (SSD) drives
• Depending on the server, support is available for up to three adapter card connections for up to 160 Gbps
of redundant I/O throughput
• Industry-standard double-data-rate 4 (DDR4) memory
• Remote management through an integrated service processor that also executes policy established in
Cisco UCS Manager software
• Local keyboard, video, and mouse (KVM) and serial console access through a front console port on each
server
• Out-of-band access by remote KVM, Secure Shell (SSH), and virtual media (vMedia) as well as Intelligent
Platform Management Interface (IPMI)
The Cisco UCS B-Series offers multiple blade server models. The supported processor family is indicated by
M4 or M5 designations on the model.
3 Drive bay 3
Adapter Cards
Depending on the model of server in question, one to three adapter cards will reside in each blade server,
providing failover connectivity to each I/O module in the chassis. Cards are released on an ongoing basis.
The Cisco UCS 6324 Fabric Interconnect connects directly to external Cisco Nexus switches through 10-Gigabit
Ethernet ports and Fibre Channel over Ethernet (FCoE) ports.
The Cisco UCS 6324 Fabric Interconnect fits into the back of the Cisco UCS Mini chassis. Each Cisco UCS
Mini chassis can support up to two UCS 6324 Fabric Interconnects, which enables increased capacity as well
as redundancy.
It can be hot-plugged into the rear of a Cisco UCS 5108 blade server chassis. A maximum of two UCS 2408
IOMs can be installed in a chassis.
The Cisco UCS 2408 IOM joins the fourth generation of UCS products, including the following hardware:
• Cisco UCS 6400 Series Fabric Interconnects
Note The UCS 2408 IOM is supported with the UCS 6400 Series FIs only.
• Multiple Cisco UCS 1300 Series and 1400 Series Virtual Interface Cards (VICs)
• M4 and M5 generations of Cisco UCS B-Series blade servers.
• The UCS 2408 IOM requires Cisco UCS Manager 4.0(4c) or later software.
The Cisco UCS 2408 IOM provides chassis management control and blade management control, including
control of the chassis, fan trays, power supply units, and blades. It also multiplexes and forwards all traffic
from the blade servers in the chassis to the 25-Gigabit Ethernet uplink network ports that connect to the fabric
Interconnect. The IOM can also connect to a peer IOM to form a cluster interconnect (1G Ethernet connection).
3 Thumbscrews for release latches 4 HDMI combo connector: Ethernet management port
(two) and RS-232 serial via a Y-dongle cable.
LEDs
The front of the IOM has a system status LED and a port activity LED for each port.
State Definition
State Definition
Buttons
There are no buttons on the IOM.
Connectors
There are 8 25-Gigabit SFP28 uplink ports on the front of the IOM.
The HDMI connector uses a special Y-dongle cable that includes Ethernet management and serial console
connectors.
Note There is an updated, Version 2 of the UCS 2304 IOM (UCS-IOM-2304V2). You cannot mix the original
UCS-IOM-2304 and UCS-2304V2 in the same chassis. If you replace UCS-IOM-2304 with UCS-IOM-2304V2,
you must replace both IOMs in the chassis.
Note Version 2 of the UCS 2304 IOM (UCS-IOM-2304V2) requires Cisco UCS Manager 4.0(4) or later.
The Cisco UCS 2304 IOM joins the third generation of UCS products, including the following hardware:
• Cisco UCS 6332 fabric interconnect, an Ethernet or Fibre Channel over Ethernet (FCoE) chassis with
32 QSFP+ 40-Gigabit ports
• Cisco UCS 6332-16UP fabric interconnect, a Ethernet and Fibre Channel chassis with 16 1- or 10-Gigabit
SFP+ ports or 16 4-, 8-, or 16-Gigabit Fibre Channel ports, and 24 40-Gigabit QSFP+ ports
• Multiple 1300 Series VICs
The Cisco UCS 2304 IOM provides chassis management control and blade management control, including
control of the chassis, fan trays, power supply units, and blades. It also multiplexes and forwards all traffic
from the blade servers in the chassis to the 10-Gigabit Ethernet uplink network ports that connect to the fabric
Interconnect. The IOM can also connect to a peer IOM to form a cluster interconnect.
Figure 10: Cisco UCS 2304 IOM
3 HDMI console connector and 4 Captive screws for the insertion latches
Ethernet management port
LEDs
The front of the IOM has a system status LED and port activity LEDs.
Buttons
There are no buttons on the IOM.
Connectors
There are four QSFP, 40-Gigabit uplink ports on the front of the IOM. The Ethernet management port and
console connector use an HDMI connector that connects to a special Y dongle to expose the 10/100 RJ45
port and console connections.
LEDs
No LEDs are on the PDU.
Buttons
No buttons are on the PDU.
Connectors
The AC version of the PDU has four power connectors rated for 15.5 A, 200-240V @ 50-60 Hz. Only use
power cords that are certified by the relevant country safety authority or that are installed by a licensed or
certified electrician in accordance with the relevant electrical codes. All connectors, plugs, receptacles, and
cables must be rated to at least the amperage of inlet connector on the PSU or be independently fused in
accordance with the relevant electrical code. See for more information about the supported power cords. See
Supported AC Power Cords and Plugs, on page 87 for more information.
The DC version of the PDU has eight dual-post lug power connections, four positive and four negative. A
single dual-post lug grounding connection is also provided. The HDVC version of the PDU uses one Andersen
SAF-D-GRID(R) connector per power supply.
Fan Modules
The chassis can accept up to eight fan modules (N20-FAN5). A chassis must have filler plates in place if no
fan will be installed in a slot for an extended period.
LEDs
There is one LED indication of the fan module’s operational state. See Interpreting LEDs, on page 21 for
details.
Power Supplies
Different power supplies are available to work with the AC (UCSB-PSU-2500ACPL or N20-PAC5-2500W)
or DC (UCSB-PSU-2500DC48) versions of the chassis.
When configured with the Cisco UCS 6324 Fabric Interconnect, only the following power supplies are
supported: UCSB-PSU-2500ACDV dual-voltage supply and UCSB-PSU-2500DC48 -48V DC power supply.
To determine the number of power supplies needed for a given configuration, use the Cisco UCS Power
Calculator tool.
LEDs
Two LEDs indicate power connection presence, power supply operation, and fault states. See Interpreting
LEDs, on page 21 for details.
Buttons
There are no buttons on a power supply.
Connectors
The power connections are at the rear of the chassis on the PDU, with different types for AC, DC, or HVDC
input. Four hot-swappable power supplies are accessible from the front of the chassis. These power supplies
can be configured to support non-redundant, N+1 redundant, and grid-redundant configurations.
• The Desired Power Redundancy for the chassis. The supported power configurations are
non-redundant, N+1 redundancy (or any requirement greater than N+1), and grid redundancy.
To configure redundancy, see the Configuration Guide for the version of Cisco UCS Manager that you are
using. The configuration guides are available at the following URL:
http://www.cisco.com/en/US/products/ps10281/products_installation_and_configuration_guides_list.html.
Non-redundant Mode
In non-redundant mode, the system may go down with the loss of any supply or power grid associated with
any particular chassis. We do not recommend running a production system in non-redundant mode. To operate
in non-redundant mode, each chassis should have at least two power supplies installed. Supplies that are not
used by the system are placed into standby. The supplies that are placed into standby depends on the installation
order (not on the slot number). The load is balanced across active power supplies, not including any supplies
in standby.
When using Cisco UCS Release 1.3(1) or earlier releases, small configurations that use less than 25000W
may be powered up on a single power supply. When using Cisco UCS Release 1.4(1) and later releases, the
chassis requires a minimum of 2 power supplies.
Note In a non-redundant system, power supplies can be in any slot. Installing less than the required number of
power supplies results in undesired behavior such as server blade shutdown. Installing more than the required
amount of power supplies may result in lower power supply efficiency. At most, this mode will require two
power supplies.
N+1 Redundancy
The N+1 redundancy configuration implies that the chassis contains a total number of power supplies to satisfy
non-redundancy, plus one additional power supply for redundancy. All the power supplies that are participating
in N+1 redundancy are turned on and equally share the power load for the chassis. If any additional power
supplies are installed, Cisco UCS Manager recognizes these “unnecessary” power supplies and places them
on standby.
If a power supply should fail, the surviving supplies can provide power to the chassis. In addition, UCS
Manager turns on any "turned-off" power supplies to bring the system back to N+1 status.
To provide N+1 protection, the following number of power supplies is recommended:
• Three power supplies are recommended if the power configuration for that chassis requires greater than
2500 W or if using UCS Release 1.4(1) and later releases
• Two power supplies are sufficient if the power configuration for that chassis requires less than 2500 W
or the system is using UCS Release 1.3(1) or earlier releases
• Four power supplies are recommended when running the dual-voltage power supply from a 100 - 120V
source.
Adding an additional power supply to either of these configurations will provide an extra level of protection.
Cisco UCS Manager turns on the extra power supply in the event of a failure and restores N+1 protection.
Note An n+1 redundant system has either two or three power supplies, which may be in any slot.
Grid Redundancy
The grid redundant configuration is sometimes used when you have two power sources to power a chassis or
you require greater than N+1 redundancy. If one source fails (which causes a loss of power to one or two
power supplies), the surviving power supplies on the other power circuit continue to provide power to the
chassis. A common reason for using grid redundancy is if the rack power distribution is such that power is
provided by two PDUs and you want the grid redundancy protection in the case of a PDU failure.
To provide grid redundant (or greater than N+1) protection, the following number of power supplies is
recommended:
• Four power supplies are recommended if the power configuration for that chassis requires greater than
2500W or if using Cisco UCS Release 1.4(1) and later releases
• Two power supplies are recommended if the power configuration for that chassis requires less than
2500W or the system is using Cisco UCS Release 1.3(1) or earlier releases
Note Both grids in a power redundant system should have the same number of power supplies. If your system is
configured for grid redundancy, slots 1 and 2 are assigned to grid 1 and slots 3 and 4 are assigned to grid 2.
If there are only two power supplies (PS) in the a redundant mode chassis, they should be in slots 1 and 3.
Slot and cord connection numbering is shown below.
LEDs
LEDs on both the chassis and the modules installed within the chassis identify operational states, both separately
and in combination with other LEDs.
LED Locations
Figure 12: LEDs on a Cisco UCS 5108 Server Chassis—Front View
Figure 13: LEDs on the Cisco UCS 5108 Server Chassis—Rear View
Figure 14: Cisco UCS 5108 Server Chassis—Rear View with the Cisco UCS 6324 Fabric Interconnect
Interpreting LEDs
Table 4: Chassis, Fan, and Power Supply LEDs
Fan Module Off No power to the chassis or the fan module was
removed from the chassis.
Power Supply
Amber Standby.
Installation Notes and Warnings for the Cisco UCS 5108 Server
Chassis
The following notes and warnings apply to all installation tasks:
Note Before you install, operate, or service the system, see the Regulatory Compliance and Safety Information for
Cisco UCS for important safety information.
Warning This unit is intended for installation in restricted access areas. A restricted access area can be accessed only
through the use of a special tool, lock and key, or other means of security. Statement 1017
Warning Only trained and qualified personnel must be allowed to install, replace, or service this equipment.
Statement 1030
Rack Requirements
This section provides the requirements for installing in a standard open rack, assuming an external ambient
air temperature range of 50 to 95°F (10 to 35°C):
Note Do not use racks that have obstructions. These obstructions could impair access to field-replaceable units
(FRUs).
The Cisco R Series Racks are an ideal choice. If other racks will be used, the rack must be of the following
type:
• Standard 19-inch (48.3 cm) four-post EIA rack, a minimum of 39.4 inches (100 cm) deep, with mounting
rails that conform to English universal hole spacing per section 1 of ANSI/EIA-310-D-1992.
• The mounting holes of the rails in the rack must be square (unless the optional round hole adapter kit is
used).
• The tool-less rack-mount kit shipped with the chassis is required. The adjustable rack rails shipped with
each enclosure extend from 29 inches (73.66 cm) to 35 inches (88.9 cm)
• Front and rear doors—If your server rack includes closing front and rear doors, the doors must have 65
percent open perforated area evenly distributed from top to bottom to permit adequate airflow.
Caution Always use blanking panels to fill all remaining empty front panel U-spaces in
the rack. This arrangement ensures proper airflow. Using a rack without blanking
panels results in improper cooling that can lead to thermal damage.
Cable Management
To help with cable management, allow additional space in the rack above and below the chassis to make it
easier to route copper cables (plus up to eight copper cables per Cisco UCS 5108 server chassis) through the
rack.
Cable management can be an important factor in preventing overheating issues. In the following figure, the
“before“ illustration shows cables blocking the rear of the chassis, and preventing the fans from exhausting
warm air from the chassis. This situation causes failed DIMMs in the blade servers, and seemingly random
server shutdowns when internal temperatures exceed specification. Use cable ties and other wiring practices
to keep the rear of the chassis unobstructed as shown in the “after“ illustration.
Airflow Considerations
Airflow through the chassis is from front to back. Air enters the chassis through the blade servers and power
supply grills at the front of the chassis and exits through the fan modules on the back of the chassis. To ensure
proper airflow, follow these guidelines:
• Maintain ambient airflow throughout the data center to ensure normal operation.
• Consider the heat dissipation of all equipment when determining air-conditioning requirements. Do not
allow the exhaust of one system to be the intake for another system.
• When evaluating airflow requirements, take into consideration that the hot air generated by equipment
at the bottom of the rack can be drawn in the intake of the equipment above.
• Make sure that the exhaust at the rear of the chassis is unobstructed for at least 24 in. (61 cm). This
includes obstruction due to messy cabling practices.
• Some blade servers ship with internal air baffles that are placed over the DIMMs and CPUs. They are
used to channel airflow to where it is needed the most. The blades are designed to operate with air baffles
installed and the system will not cool correctly if they are not installed.
• If an enclosed rack is used, the front door must be 65 percent perforated to ensure adequate airflow to
the servers.
Caution Do not try to lift the chassis using the handles on the side. These handles are intended only for moving and
adjusting the chassis position.
• Never lift the chassis alone—Always use two people to lift the chassis. If available, use a scissor jack or
other lifting device designed for installing heavy equipment into data center racks.
• Disconnect all power and external cables before lifting the chassis.
• Remove all FEXes, power supplies, fans, and servers from the chassis before lifting.
• Ensure that your footing is solid and the weight of the system is evenly distributed between your feet.
• Lift the system slowly, keeping your back straight. Lift with your legs, not with your back. Bend at the
knees, not at the waist.
Caution Do not remove the Power Distribution Unit (PDU) located at the back of the chassis.
Installation Guidelines
When installing the chassis, follow these guidelines:
• Plan your site configuration and prepare the site before installing the chassis. See Site Planning and
Maintenance Records, on page 93 for the recommended site planning tasks. For details, see the Cisco
UCS Site Preparation Guide.
• Record the information listed in Site Planning and Maintenance Records, on page 93 as you install and
configure the chassis.
• Ensure that there is adequate space around the chassis to allow for servicing the chassis and for airflow.
• Ensure that the air-conditioning meets the heat dissipation requirements listed in Technical Specifications,
on page 81
• Ensure that the cabinet or rack meets the requirements listed in Rack Requirements, on page 26.
Note Jumper power cords are available for use in a rack. See Specifications for the
Cisco UCS 5108 Blade Server Chassis Power Supply Units, on page 84.
• Ensure that the site power meets the power requirements listed in Technical Specifications, on page 81.
We recommend that you use a UPS to protect the UCS system. Using an unprotected supply exposes
you to a risk of system failure due to input supply voltage variations or failures.
Avoid UPS types that use ferroresonant technology. These UPS types can become unstable with systems
such as the Cisco UCS, which can have substantial current draw fluctuations due to fluctuating data
traffic patterns.
• Ensure that circuits are sized according to local and national codes. For North America, the power supply
requires a 20 A circuit.
To prevent loss of input power, ensure that the total maximum loads on the circuits supplying power to
the chassis are within the current ratings for the wiring and breakers.
• Use the following torque values when installing the chassis:
• 10-32 screws: 20 in-lb
Required Equipment
Before you begin the installation, ensure that you have the following items:
• Number 1 and number 2 Phillips-head screwdrivers with torque measuring capabilities
• Tape measure and level
• ESD wrist strap or other grounding device
• Antistatic mat or antistatic foam
Caution When handling chassis components, wear an ESD strap and handle modules by the carrier edges only.
Tip Keep the shipping container in case the chassis requires shipping in the future.
Note The chassis is thoroughly inspected before shipment. If any damage occurred during transportation or any
items are missing, contact your customer service representative immediately.
Procedure
Step 1 Remove the chassis from its cardboard container. Save all packaging material.
Step 2 Compare the shipment to the equipment list provided by your customer service representative and verify that
you have received the following items:
• Any printed documentation
• Tool-less rack-mount kit (N20-CRMK2=)—mounting rails can be installed in a rack without the use of
tools. The optional round hole adapter kit (N20-CRMK2-RHA=) does require tools.
• ESD wrist strap
• Cables with connectors (including the N20-BKVM=, which is the KVM/local I/O console dongle)
• Any optional items ordered
Step 3 Verify that all unused blade slots and power supply bays have blank covers.
Note The chassis tool-less rails are designed for racks that have square mounting holes. You must use the round
hole adapters (N20-CRMK2-RHA=) to install the chassis in racks that have round mounting holes.
This round hole adapter kit allows you to adapt the rail kit (N20-CRMK2=) to install into rack (front and/or
rear) posts that use either threaded or non-threaded round holes. Four adapters in the kit are for adapting the
rail kit to install into rack posts with threaded round holes, and the other four adapters in the kit are for adapting
the rail kit to install into rack posts with non-threaded round holes. You can use a combination of adapters
based on the type of holes in the rack posts. Various sizes and lengths of screws are also included in the kit.
Procedure
Step 1 Insert the adapter tab into the mounting rail as shown in callout 1.
Step 2 Slide the adapter up to lock it into position as shown in callout 2.
Step 3 Secure the adapter into place using the provided pan-head screw as shown in callout 3.
Caution Never attempt to lift the chassis by using an installed module’s handle as a grip point. only use the handles
on the sides of the chassis.
Caution If the rack has wheels, ensure that the brakes are engaged, the stabilizing pads are extended, or that the rack
is otherwise stabilized.
Table 9: Contents of the Cisco UCS 5108 Server Chassis Rack-Mount Kit (N20-CRMK2=)
Table 10: Contents of the Cisco UCS 5108 Server Round Hole Adapter Kit (N20-CRMK2-RHA=)
Warning The plug-socket combination must be accessible at all times, because it serves as the main disconnecting
device. Statement 1019
Step 1 Remove the mounting template (Cisco 78-19093-01) from the accessory box. The template is designed to
show you the proper holes within which the rails and cage nuts should be placed. Once the rack holes line up
with the template, you should mark the holes so that their position is known after removing the template.
Step 2 Adjust the length of the rail by sliding the ends of the rail back and forth until they match the depth of the
rack.
Step 3 Place the two hooks at each end of the rail into the first two holes at a rack unit boundary.
Figure 18: Hole Spacing for the Tool-less Rails in Relationship to a Rack Unit
The following figure shows a rail mounted into a rack in the proper position with respect to a rack-unit
boundary. Measurements are in inches between the centers of the holes.
Figure 19: Installing Tool-less Chassis Support Rails into the Rack
Step 4 Press down firmly on the rail until the hooks seat firmly and securely into the holes, and the spring clip latches
into place.
Step 5 Follow the same procedure to install the other rack rail.
Step 6 Use a tape measure and level to verify that the rack rails are horizontal and at the same height.
Step 7 Insert the cage nuts on to the rack in the needed square holes as shown below. When the rails are installed on
a rack unit boundary, the first two cage nuts are installed into the seventh holes above the rails’ horizontal
plates. The next two cage nuts are installed into the fifth holes above the first cage nut. Finally, the two cage
nuts are installed into the fourth holes above the second cage nuts.
Figure 20: Placement of Rails and Cage Nuts with Respect to the Rack Unit Boundary
Figure 21: Proper Placement for the Rails and Cage Nuts
Step 8 Remove all power supplies, fan assemblies, server blades, and fabric extenders to lighten the chassis. Even
with devices removed, the chassis weighs 90 lbs (40.83 kg).
Procedure
Step 1 Remove the mounting template (Cisco 78-19093-01) from the accessory box. The template is designed to
show you the proper holes within which the rails and cage nuts should be placed. Once the rack holes line up
with the template, you should mark the holes so that their position is known after removing the template.
Step 2 Adjust the length of the rail by sliding the ends of the rail back and forth until they match the depth of the
rack.
Step 3 Place the adapters and rails even with a rack boundary at each end of the rail.
Step 4 Secure the rail to the rack with the provided pan head screws as shown below.
Figure 22: Attaching the Mounting Brackets to a Round Hole Rack
Step 5 Follow the same procedure to install the other rack rail as shown below.
Step 6 Use a tape measure and level to verify that both rack rails are horizontal and at the same height.
Step 7 Remove all power supplies, fan assemblies, server blades, and I/O modules to lighten the chassis. Even with
devices removed, the chassis weighs 90 lbs (40.83 kg).
Step 1 With the help of another person (or special lifting equipment), lift the chassis and place it on the mounting
rail as shown.
Figure 24: Mounting Rail Weight Distribution (Square Hole Mount Shown)
Caution The mounting rails may come loose and cause the chassis to fall if the weight is resting on the
wrong surface. Make sure that the bottom of the chassis is resting on the correct rail surface.
Step 2 Slide the chassis into the rack until the front flange is flat against the cage nuts. (Cage nuts are not needed in
round hole racks.)
Step 3 Using the six Phillips round washer head screws and the cage nuts (used in square hole installations), secure
the chassis by its flanges to the rack as shown.
Step 4 Replace all servers, fans, and power supplies back into their respective chassis slots.
Step 5 To power up the chassis, connect the appropriate AC power cables to the inlet connector corresponding to
each installed power supply, and then connect the other end of the cables to the power source. For a DC
installation, see Connecting a DC Power Supply, on page 43.To determine the number of power supplies
needed for a given configuration, use the Cisco UCS Power Calculator tool.
Note Both grids in a power redundant system should have the same number of power supplies. If your
system is configured for grid redundancy, slots 1 and 2 are assigned to grid 1 and slots 3 and 4
are assigned to grid 2. If only two power supplies (PS) are in a redundant- mode chassis, they
should be in slots 1 and 3. Slot and cord connection numbering is shown below.
Figure 26: Power Supply Bay and Connector Numbering
Step 6 Connect the server chassis to the fabric interconnect as described in Proper FEX and Fabric Interconnect Port
Connectivity, on page 48.
Use cage nuts and mounting screws to secure the brackets to the rack rail.
From the rear of the rack, the rear brackets should be installed as follows:
Procedure
Step 1 Install the UCS 5108 or UCS Mini chassis. See callout 1 in the following figure.
Step 2 Attach the supplied cage nuts to the rack rails for the left and right brackets. See callout 2. The tightening
torque is 45 lb-in (5.0 N-m).
Step 3 Secure the brackets using the supplied screws. See callout 3.
Figure 28: Installing the Rear Brackets to the Chassis
Required Tools
You must have the following tools to perform this procedure:
• A Phillips screwdriver
• A 10-mm wrench or socket
• Connectors and wire for the DC circuit or circuits
Warning When stranded wiring is required, use approved wiring terminations, such as closed-loop or spade-type with
upturned lugs. These terminations should be the appropriate size for the wires and should clamp both the
insulation and conductor. Statement 1002
Warning Before performing any of the following procedures, ensure that power is removed from the DC circuit.
Statement 1003
Warning A readily accessible two-poled disconnect device must be incorporated in the fixed wiring. Statement 1022
Warning This product requires short-circuit (overcurrent) protection, to be provided as part of the building installation.
Install only in accordance with national and local wiring regulations. Statement 1045
Warning When installing or replacing the unit, the ground connection must always be made first and disconnected last.
Statement 1046
Warning Installation of the equipment must comply with local and national electrical codes. Statement 1074
Warning Hazardous voltage or energy may be present on DC power terminals. Always replace cover when terminals
are not in service. Be sure uninsulated conductors are not accessible when cover is in place. Statement 1075
Procedure
Step 1 Install the DC power supply in the chassis, making note of the bay number, so you are sure to connect the
wiring to the appropriate terminals on the DC PDU at the chassis rear.
Step 2 Verify that power is off to the DC circuit or circuits on the power supply that you are installing.
Step 3 Ensure that all site power and grounding requirements have been met.
Step 4 Remove the plastic cover from the DC terminals by squeezing the flanges at the top and bottom of the cover.
Step 5 Connect the ground wires to the power supply terminal block, shown as a green wire below. Only one ground
connection is required, though there may be up to four DC connections.
Step 6 Connect the DC-input wires to the power supply terminal block. The proper wiring sequence is positive to
positive (red wire), and negative to negative (black wire). The figure below shows a connection to terminal
1.
Note
The positive and negative wires can be installed pointing either to the right or to the left as long
as the terminal cover is used. The figure below shows them pointed to the right. Panduit
LCD4-14A-L connectors may be used for the supply and return wires, and Panduit LCD4-14AF-L
or equivalent connectors may be used for the 90-degree ground lug wire. Both connections have
double lugs with .25 inch holes measuring .625 inches from center to center.
Figure 29: Connecting DC Power to the Chassis (shows DC PDU only, Chassis is Omitted)
Step 7 Replace the terminal cover as shown. This cover should always be in place when power is applied to the
terminals.
Step 8 Connect the other end of the power wires to a DC-power input source.
Step 9 Set the DC disconnect switch in the circuit to ON.
Caution In a system with multiple power supplies, connect each power supply to a separate DC power
source. In the event of a power source failure, if the second source is still available, it can maintain
system operation.
Step 10 Verify power supply operation by checking the power supply's front-panel LEDs. You should see the following:
• The LED labeled INPUT OK is green.
• The LED labeled OUTPUT FAIL is not lit.
Step 11 Check the power supply and system status from the UCS console by entering the show system command or
the show power command, do using the GUI. For more information on these commands, refer to the command
reference for your software.
Caution Adding a second link to a fabric-port-channel port group is disruptive and will automatically increase the
available amount of VIF namespace from 63 to 118. Adding further links is not disruptive and the VIF
namespace stays at 118.
Caution Linking a chassis to two fabric-port-channel port groups does not affect the VIF namespace unless it is manually
acknowledged. The VIF namespace is then automatically set to the smaller size fabric port-channel port group
usage (either 63 or 118 VIFs) of the two groups.
For high availability cluster-mode applications, we strongly recommend symmetric cabling configurations.
If the cabling is asymmetric, the maximum number of VIFs available is the smaller of the two cabling
configurations.
For more information on the maximum number of VIFs for your Cisco UCS environment, see the Configuration
Limits document for your hardware and software configuration.
Note The following illustrations are for example only; you do not need to skip available ports to provide future
expansion room. See the UCS Manager configuration guide for FI port configuration considerations and
limitations.
Figure 31: Invalid Connection for the Server Chassis and two Cisco UCS 6120XP Fabric Interconnects
• Both fabric interconnects should be wired identically: if port 1 on FEX 1 for a chassis goes to FI-A
port 5, then port 1 on FEX 2 goes to FI-B port 5.
The following figure shows valid connections from FEXes in two chassis to two separate fabric
interconnects. When you connect the server chassis to the fabric interconnect do not connect the
FEXes to the fabric interconnect's expansion modules. While similar in appearance to the other
ports on the fabric interconnect, the expansion modules are never used for direct chassis connections,
they are used for uplink or SAN connections.
Figure 32: Proper Connection for the Server Chassis and two Cisco UCS 6120XP Fabric Interconnects
Step 2 Disconnect the power cords and networking cables from the chassis.
Step 3 Remove all modules and blades from the chassis to lighten its weight.
Step 4 Remove the screws holding the front rack-mount flange to the rack.
Step 5 With two people holding the chassis, make sure that its weight is fully supported.
Step 6 Gently slide the chassis off the rails, and out of the rack.
Step 7 Replace the modules and blades in the server chassis.
If you are returning the chassis to Cisco, contact your Cisco customer service representative to arrange for
return shipment to Cisco.
SFP+ Transceivers
Each FEX within the chassis supports Small Form-Factor Pluggable (SFP+) copper or optical transceivers.
Each transceiver runs at 10 Gb.
Model Description
The figure below shows the SFP-H10GB-CU5M transceiver. The rubber loop is used for removing the SFP+
from its port on the I/O module.
Model Description
SFP and SFP+ Transceivers for the UCS 6324 Fabric Interconnect
Each UCS 6324 Fabric Interconnect supports up to four SFP optical or copper transceivers.
Model Description
Model Description
Model Description
The UCS 6324 Fabric Interconnect also supports Twinax copper transceivers. The enhanced SFP+ 10-Gb
Ethernet transceiver is a bidirectional device with a transmitter and receiver in the same physical package. It
has a 20-pin connector on the electrical interface.
Model Description
Model Description
Model Description
QSFP+ Copper Optical Transceivers for the UCS 6324 Fabric Interconnect
Each UCS 6324 Fabric Interconnect supports up one copper/optical transceiver.
Model Description
Step 1 Remove the copper Twinax SFP+ from the FEX port by pulling gently on the rubber loop. The cable and
SFP+ transceiver come out as a single unit, leaving the FEX port empty.
Step 2 Insert the optical SFP+ transceiver into the FEX port. Make sure that it clicks firmly into place.
Step 3 Plug the fiber-optic cable into the optical SFP+ transceiver.
Figure 35: Replacing a Copper SFP+ Transceiver with an Optical SFP+ Transceiver
Components
The following figure shows an empty Cisco UCS 5108 server chassis and identifies the front, back, and module
slots.
Note Whenever you remove a module from the chassis for an extended period of time, always replace the module
with the appropriate blank panel. Failing to do so can result in heating and EMI issues. Blank panels can be
ordered from Cisco Systems.
Note Before you install, operate, or service the system, see the Regulatory Compliance and Safety Information for
Cisco UCS for important safety information.
Warning This unit is intended for installation in restricted access areas. A restricted access area can be accessed only
through the use of a special tool, lock and key, or other means of security. Statement 1017
Warning Only trained and qualified personnel must be allowed to install, replace, or service this equipment.
Statement 1030
Caution To prevent ESD damage, wear grounding wrist straps during these procedures and handle modules by the
carrier edges only.
Procedure
Step 1 Place the hard drive lever into the open position by pushing the release button.
Step 2 Gently slide the hard drive into the opening in the blade server until it seats into place.
Step 3 Push the hard drive lever into the closed position.
The following figure shows the positioning of a hard drive within a blade server.
Figure 37: Positioning a Hard Drive in a Blade Server (Cisco UCS B200 M1 shown)
192524
1
2
Procedure
Step 1 Push the button to release the ejector, and then pull the hard drive from its slot.
Figure 38: Pressing the Button on the Front of the Hard Drive to Deploy the Lever (Cisco UCS B200 M1 shown)
Step 2 Place the hard drive on an antistatic mat or antistatic foam if you are not immediately reinstalling it in another
blade server.
Step 3 Install a blank faceplate (N20-BBLKD) to keep dust out of the blade server if the slot will remain empty.
Note The AC and DC power supplies are keyed to work only with the power distribution system in their respective
chassis version. Mixing AC and DC in a given chassis is not supported.
Procedure
Step 6 Connect the other end of the power cable to the AC–power source.
Step 7 Verify the power supply is operating by checking the power supply LEDs. See LED Locations, on page 20
and Interpreting LEDs, on page 21.
Figure 39: Positioning a Power Supply in the Cisco UCS Server Chassis
Note Both grids in a power redundant system should have the same number of power supplies. If your
system is configured for grid redundancy, slots 1 and 2 are assigned to grid 1 and slots 3 and 4
are assigned to grid 2. If there are only two power supplies in a redundant- mode chassis, they
should be in slots 1 and 3. This would be a very unusual configuration, with a single B200 blade
server in the chassis. A larger configuration would require two power supplies per grid. Slot and
cord connection numbering is shown below.
Figure 40: Power Supply Bay and Connector Numbering (AC Version Shown)
Caution If you are using the Cisco UCS 5108 server chassis with one power supply (this is only supported in earlier
versions of the system software), removing the power supply will cause the servers and chassis to shut down.
If you are using more than two power supplies, and you remove one of them, the servers continue to operate
as long as the other power supplies are sufficient to meet the power requirements of the number of servers in
the chassis.
Procedure
Step 1 Loosen the captive screw on the front of the power supply.
Step 2 Lift up the power supply’s handle to unseat the power supply.
Step 3 Using the lever, pull the power supply from its slot. Place your other hand under the power supply to support
it while you slide it out of the chassis.
Step 4 Install a blank power supply filler panel (N20-CBLKP) if the power supply bay is to remain empty.
Caution The Power Distribution Unit should only be replaced by a Cisco certified technician. This section is for
reference only.
Caution You can not hot swap a PDU (N01-UAC1). The entire chassis will need to be shut down and all power cords
should be unplugged before attempting this procedure.
Installing a PDU
Procedure
Step 1 Hold the PDU module from below, with the captive screws at the top of the module.
Figure 41: Positioning a PDU (N01-UAC1) in the Cisco UCS Server Chassis, AC Version
Step 2 Push the PDU module into the chassis until it seats properly. If power supplies are installed, you may need
to partially remove them.
Step 3 Tighten the captive screws.
Step 4 If necessary, reseat the power supplies and restart the system to observe LED behavior.
Removing a PDU
Procedure
Step 1 Partially remove all installed power supplies. If a power supply is seated into the PDU, removal is difficult
or impossible.
Step 2 Loosen the captive screws.
Step 3 Pull the PDU clear of the chassis by pulling on the captive screws. Support its weight from below.
Note The Cisco UCS 2304 IOM cannot be installed in the same chassis with a UCS 2208XP or UCS 2204XP IOM
except during migration.
Note There is a version 2 of the UCS 2304 IOM. You cannot mix UCS-IOM-2304V2 and UCS-IOM-2304 in the
same chassis. If you replace one UCS-IOM-2304 with UCS-IOM-2304V2, you must replace both IOMs with
V2.
UCS-IOM-2304V2 requires Cisco UCS Manager 4.0(4) or later.
Step 1 Remove all cables from the IOM or UCS 6324 Fabric Interconnect.
Step 2 Loosen the captive screws on the IOM or UCS 6324 Fabric Interconnect.
Step 3 Pull the levers outward to unseat the IOM or UCS 6324 Fabric Interconnect.
Step 4 Using the levers, pull the IOM or UCS 6324 Fabric Interconnect from its slot.
Step 5 Only if the slot is to remain empty, install a blank IOM filler panel (N20-CBLKI).
Note If you plan to remove an IOM or UCS 6324 fabric interconnect for an extended period of time,
you must replace it with the N20-CBLKI blank to avoid EMI issues. Blanks can be ordered from
Cisco.
Step 1 Make sure that the two levers at the front of the FEX or the UCS 6324 Fabric Interconnect are pulled open.
Step 2 Slide the FEX or UCS 6324 Fabric Interconnect into the chassis slot, ensuring that the module is fully seated.
Figure 42: Positioning an IOM or UCS 6324 Fabric Interconnect into the Server Chassis
Step 3 Close the levers and tighten the captive screw on each lever.
Step 4 Connect the networking cables (see SFP+ Transceivers, on page 51 and Proper FEX and Fabric Interconnect
Port Connectivity, on page 48 ).
Step 5 Verify that the module is functioning correctly. See LED Locations, on page 20 and Interpreting LEDs, on
page 21.
Note If you plan to migrate to UCS 2408 IOMs, you must first migrate from UCS 6200 Series FIs to UCS 6400
Series FIs. See Supported Migration Paths, on page 67.
UCS 6200 FI with 2200 IOMs UCS 6400 Series FI with 2200 IOMs
UCS 6400 Series FI with 2200 IOMs UCS 6400 Series FI with 2408 IOMs
Caution Cisco UCS Manager Release 4.0(1a) is the minimum version that provides support for Cisco UCS 6454 Fabric
Interconnects. To migrate from Cisco UCS 6200 Series to Cisco UCS 6454 Fabric Interconnects:
• Cisco UCS 6200 Series Fabric Interconnects must be on Cisco UCS Manager Release 4.0(1) or a later
release.
• Cisco UCS 6454 Fabric Interconnects must be loaded with the same build version that is on the Cisco
UCS 6200 Series Fabric Interconnect that it will replace.
Note Cisco UCS 6454 FI requires Cisco UCS Manager 4.0(1a) or later. Cisco UCS 2408 IOM requires Cisco UCS
Manager 4.0(4c) or later. If you will also be migrating to UCS 2408 IOM, upgrade to 4.0(4c) or later.
• Licenses from Cisco UCS 6200 Series Fabric Interconnects are not transferable to Cisco UCS 6454
Fabric Interconnects. You must obtain licenses for the Cisco UCS 6454 Fabric Interconnects before you
upgrade.
• During migration, the Cisco UCS 6200 Series Fabric Interconnect and the Cisco UCS 6454 Fabric
Interconnect must use the same allowed SSL protocol, either default or Only TLSv1.2, to successfully
complete compatibility checks.
• Cisco UCS 6454 Fabric Interconnects use the IDLE fill pattern for FC uplink ports and FC storage ports
when using 8 Gbps speed.
When migrating to Cisco UCS 6454 Fabric Interconnects and configuring FC Uplink Ports or FC Storage
Ports at 8Gbps speed, ensure that the fill pattern is set as IDLE on the corresponding FC switch ports
and the direct-attached FC storage array ports. If the fill pattern is not set as IDLE, FC uplink ports and
FC storage ports operating at 8 Gbps might go to an errDisabled state, lose SYNC intermittently, or
receive errors or bad packets.
Cisco UCS 6454 Fabric Interconnects do not support 8 Gbps direct-attached FC connectivity (FC uplink
ports or FC storage ports) without fill-pattern set to IDLE. When migrating to Cisco UCS 6454 Fabric
Interconnects from Cisco UCS 6200 Series Fabric Interconnects, do one of the following:
• Use a SAN switch between the Cisco UCS 6454 Fabric Interconnect and the storage array with 8
GB FC connectivity.
• Upgrade the storage array to 16 GB or 32 GB FC connectivity.
• Before migrating from Cisco UCS 6200 Series Fabric Interconnects to Cisco UCS 6454 Fabric
Interconnects, ensure that you unconfigure the unified ports on the Cisco UCS 6200 Series Fabric
Interconnects.
After migrating to Cisco UCS 6454 Fabric Interconnects, reconfigure the unified ports based on their
location on the Cisco UCS 6454 Fabric Interconnects, and reacknowledge the newly configured ports.
For example, a unified port on a UCS 6248 Fabric Interconnect should be reconfigured on any port
between 1 and 16 on a Cisco UCS 6454 Fabric Interconnect.
• Upgrading the fabric interconnect should be done before upgrading to a new FEX/IOM or virtual interface
card.
• During fabric interconnect migration, image synchronization between fabric interconnects is not allowed.
This is done to prevent incompatible images from getting synchronized. We recommend that you download
B-Series and C-Series server software bundles again after migration is complete.
• Do not attempt to implement new software features from the new Cisco UCS software version until all
required hardware is installed.
• Changes to the topology, such as the number of servers or uplink connections, should be performed after
the fabric interconnect migration is complete.
• Make a detailed record of the cabling between FEXes/IOMs and fabric interconnects. You must preserve
the physical port mapping to maintain the server pinning already configured and minimize down time.
• For a cluster configuration, both fabric interconnects must have symmetrical connection topologies
between fabric interconnect and FEXes/IOMs.
• Cisco UCS VIC 1455 and 1457 adapters support cables of 10G and 25G speed. However, the cables
connecting Cisco UCS VIC 1455 or 1457 adapter ports to each 6454 fabric interconnect must be of
uniform speed-either all 10G or all 25G cables. If you connect these adapter ports to a 6454 fabric
interconnect through a mix of 10G and 25G cables, UCS rack-mount server discovery fails and ports
may go to a suspended state.
• Standalone installations should expect down time. Migrating or upgrading a fabric interconnect is
inherently traffic disruptive.
• A best practice would be to perform a full configuration and software backup before performing this
hardware upgrade.
• A WWN pool can include only WWNNs or WWPNs in the ranges from 20:00:00:00:00:00:00:00 to
20:FF:00:FF:FF:FF:FF:FF or from 50:00:00:00:00:00:00:00 to 5F:FF:00:FF:FF:FF:FF:FF. All other
WWN ranges are reserved. When fibre channel traffic is sent through the UCS infrastructure the source
WWPN is converted to a MAC address. You cannot use WWPN pool which can translate to source
multicast MAC addresses. To ensure the uniqueness of the Cisco UCS WWNNs and WWPNs in the
SAN fabric, Cisco recommends using the following WWN prefix for all blocks in a pool:
20:00:00:25:B5:XX:XX:XX
Note If you are using appliance ports for direct attached storage, you must add VLANs to the ethernet uplinks. This
will ensure that vNICS can properly pin on boot.
On the UCS 6248 fabric interconnect, you can separate the 32 physical ports in slot one into two contiguous
pools, low numbered ports being Ethernet ports and high numbered ports being Fibre Channel ports. On the
UCS 6454 fabric interconnect, you can configure the first 16 ports as Fibre Channel ports (when using Cisco
UCS Manager releases earlier than 4.0(4), only ports 1-8 are unified ports).
Because a UCS 6248 has 32 ports in slot 1 and a UCS 6454 has all ports in slot 1, any ports on GEM slots
will be removed during the hardware upgrade process.
Caution If you ever need to change the pool sizes for slot 1, you must reboot the fabric interconnect, which can lead
to a service disruption. If you ever need to change the pool sizes for slot 2, you must reset the expansion
module in slot 2. To minimize disruption, plan to have at least a few Ethernet uplink and Fibre Channel uplink
ports configured on slot 1. Implement this fail-safe after the upgrade is complete and the system restabilizes.
Note Migration from UCS 6300 Series FIs to UCS 6400 Series is not supported at this time.
You can migrate from a UCS 6200 Series Fabric Interconnect to a UCS 6454 Fabric Interconnect. However,
you cannot migrate back to a UCS 6200 Series Fabric Interconnect after you have migrated to a UCS 6454
Fabric Interconnect.
The UCS 6454 Fabric Interconnect is intended as a replacement for the UCS 6200 Series Fabric Interconnect,
but not as a replacement for the higher speed (or 40Gb) UCS 6332/6332-16UP Fabric Interconnect. Therefore,
Cisco has not tested or published a plan to migrate from UCS 6332/6332-16UP Fabric Interconnects to UCS
6454 Fabric Interconnects.
Unless otherwise noted, for more information about how to perform configuration procedures in Cisco UCS
Manager for a particular step, see the appropriate Cisco UCS Manager configuration guide for Cisco UCS
Manager Release 4.0.
Procedure
Step 1 Download Cisco UCS Manager, Release 4.0 or later versions to the UCS 6200 Series Fabric Interconnects
and upgrade to this version.
Step 2 Evacuate traffic from the subordinate fabric interconnect to ensure there is no data traffic impact during
migration.
See Firmware Management Guide, FI Traffic Evacuation
Step 3 Unconfigure all the server ports on the subordinate fabric interconnect.
Step 4 Power down the subordinate fabric interconnect by unplugging it from the power source.
If you are monitoring the migration using a KVM session, you may need to reconnect the KVM session when
you power down the fabric interconnect.
Step 5 Mount the replacement UCS 6454 fabric interconnect into either the same rack or an adjacent rack.
See the Cisco UCS 6400 Series Fabric Interconnect Hardware Installation Guide.
Step 6 Disconnect the cables from the chassis FEXes/IOMs to the subordinate fabric interconnect ports in slot 1 on
the UCS 6200 Series Fabric Interconnect.
Step 7 Connect these cables into the corresponding ports on slot 1 of one of the new Cisco UCS 6454 fabric
interconnects, using the connection records to preserve the port mapping and the configured server pinning.
To change the port mapping, especially while reconfiguring FC ports, you must reacknowledge the newly
configured ports.
Step 8 Connect the L1/L2 cables that you disconnected onto the new Cisco UCS 6454 fabric interconnect. L1 connects
to L1; L2 connects to L2.
Step 9 Connect the server and uplink cables.
See the Cisco UCS 6400 Series Fabric Interconnect Hardware Installation Guide.
Step 10 Connect the power to the new Cisco UCS 6454. It will automatically boot and run POST tests. If it reboots
itself, this is a normal behavior.
Important Directly connect the console port to a terminal and observe the boot sequence. You should at
some point see the Basic System Configuration Dialog, where you will configure the switch as
a subordinate interconnect. If you do not see this dialog, you either have different builds of
software on your old primary and new subordinate, or the new subordinate has previously been
part of a cluster and will need to have all configuration information wiped before it can be added
to a cluster as a subordinate. In either case, immediately disconnect the L1 and L2 connections
and complete the bringup as a standalone fabric interconnect, then correct the issue before
proceeding further.
Step 11 Configure the server and uplink ports on the new Cisco UCS 6454 fabric interconnect.
Step 12 Wait for the new subordinate Cisco UCS 6454 fabric interconnect to automatically synchronize the configuration
and database/state information from the primary UCS 6200 Series Fabric Interconnect.
Synchronization between primary and subordinate fabric interconnects can take several minutes. You may
see an error message that will persist until the server ports are enabled.
The port configuration is copied from the subordinate switch to the new hardware.
Cable the second new fabric interconnect identically to the first, and then allow the reconfiguration to be
applied to the second new fabric interconnect.
Note The Cisco UCS component compatibility with the UCS 2408 IOM is as follows:
• Requires Cisco UCS Manager 4.0(4e) or later software
• Supported with the UCS 6400 Series FIs only
• Supported with Cisco UCS 1300 and 1400 Series Virtual Interface Cards (VICs) only
• Supported with Cisco UCS B-Series M4 and M5 server generations only
Note If you plan to migrate to UCS 2408 IOMs, you must first migrate from UCS 6200 Series FIs to UCS 6400
Series FIs. See Migrating from UCS 6200 to UCS 6454 Fabric Interconnects, on page 67.
Note Migration from UCS 2300 Series IOMs to UCS 2400 Series IOMs is not supported at this time.
Note If the primary and the subordinate IOMs are different models, the overall status of the IOMs is shown as Peer
Comm Problem. However, the IOMs can still be replaced.
IOM 2408 connects only to the 6454 Fabric Interconnects. Please follow the steps described in 6454
migration-guide to ensure the UCS domains with IOM 2200 series have 6454 Fabric Interconnects.
Procedure
Step 1 UCS IOM 2408 connects only to the UCS 6454 Fabric Interconnects. Follow the steps to ensure the UCS
domains with IOM 2200 series have 6454 Fabric Interconnects. See Migrating from UCS 6200 to UCS 6454
Fabric Interconnects, on page 67.
Step 2 Ensure the supported 25-GbE transceivers and cables are used to connect the IOM 2408 with FI 6454. See
the Cisco UCS 2408 Fabrix Extender Data Sheet.
Step 3 See the list of UCS components compatible with 2408 IOM above this procedure.
Use the hardware replacement procedures in Replacing an I/O Module (or UCS 6324 Fabric Interconnect For
UCS Mini), on page 66.
Step 9 On the subordinate FI 6454, unconfigure server ports that connect to the migrating IOM 220x. When you
unconfigure the server ports, the primary FI takes over traffic forwarding.
Step 10 Disconnect cables connecting the subordinate FI 6454 and the corresponding IOM 220x.
Step 11 Remove and replace the migrating IOM 220x with IOM 2408. Connect the IOM 2408 to the subordinate FI
6454 with 25-gigabit ethernet cables. At this point, the migrating UCS chassis has a mix of IOM 220x and
IOM 2408.
Step 12 Configure server ports from UCS Manager on the subordinate FI 6454 ports connected to the IOM 2408,
one-by-one to each chassis. Verify the 25GbE links come up between the IOM and FI.
Step 13 Wait for auto-upgrade of IOM 2408 to complete. Due to a known limitation, the migration of IOM 2208 to
IOM 2408 requires a reboot of the FI after replacing the IOM. After the reboot of the subordinate FI, wait for
the subordinate FI and migrating IOM 2408 to be fully operational with traffic resuming over the subordinate
FI fabric.
Note At this time, when migrating from UCS 2208 IOM to 2408 IOM, Vethernet interfaces are not
pinned to server interfaces (HIFPC) dynamically on the FI. You must reboot the FI to bring up
the Vethernet interfaces. Refer to the Release Notes for 4.0(4e) for more details. The issue is
being tracked in CDETS CSCvr46327.
RNE for 4.0(4e) UCSM
Step 14 Ensure the IOM 2408 has loaded the latest firmware and verify there are no faults on UCS chassis with the
new IOM 2408. Cluster HA redundancy will not be present due to mismatch of IOM's on the migrating UCS
chassis but traffic forwarding over FI's is not affected.
Step 15 After completing IOM migration on the subordinate FI, repeat steps 9 to 14 to replace the other IOM 220x
connected to the primary FI 6454 and complete the IOM 2408 migration for the UCS domain.
Note After reload of theprimary FI, fabric cluster failover will happen. However, this does not affect
data traffic forwarding in the UCS domain.
Step 16 With the completion of IOM 2408 migration on all the UCS 5108 chassis in the domain, verify that cluster
HA redundancy is back on the Fabric Interconnects, that there are no faults on the UCS Manager, and that
traffic is forwarding over the new IOM 2408s on both Fabric A and B.
Note There is a version 2 of the UCS 2304 IOM. UCS-IOM-2304V2 requires Cisco
UCS Manager 4.0(4) or later. You cannot mix UCS-IOM-2304V2 and
UCS-IOM-2304 in the same chassis.
• If you plan to migrate to a UCS 6332-16UP Fabric Interconnect, determine the number of VSANs that
are configured in your system. If more than 15 VSAN IDs are configured, reduce this number to 15 or
less to avoid possible configuration issues with Fibre Channel or FCoE ports on a UCS 6332-16UP Fabric
Interconnect. It is not necessary to reduce the number of configured VSANs before migrating to a UCS
6332 Fabric Interconnect.
You can migrate from a UCS 6200 Series Fabric Interconnect to a UCS 6300 Series Fabric Interconnect. You
can also revert back to UCS 6200 Series Fabric Interconnect after you have migrated to a UCS 6300 Series
Fabric Interconnect.
You cannot migrate from a UCS 6100 Series Fabric Interconnect to a UCS 6300 Series Fabric Interconnect.
Note Following a migration, remove port-related configurations in the fabric interconnect and then reconfigure the
ports. For example, you must reconfigure the SPAN destination ports after migrating from a UCS 6200 Series
Fabric Interconnect to a UCS 6300 Series Fabric Interconnect.
Note As stated in the Prerequisites, make sure that you reduce the number of configured VSANs in your UCS
system to 15 or less before you begin the migration, if you plan to migrate to a UCS 6332-16UP Fabric
Interconnect. It is not necessary to reduce the number of configured VSANs before migrating to a UCS 6332
Fabric Interconnect.
Procedure
Step 1 If you are upgrading to the Cisco UCS 6332-16UP, verify that the environment does not exceed 15 VSANs.
Step 2 Download Cisco UCS Release 3.1(1) or later from the Cisco Support site at
https://software.cisco.com/download/navigator.html and upgrade the existing Cisco UCS 6200 Series domain
to Release 3.1(1) or later.
Step 3 Fail over traffic from the subordinate server to the primary server. You can do this in several ways, as listed
in the following table:
Disable uplink Individually disable the uplink ports for Ethernet, Fibre Channel, and FCoE
connections.
Ethernet only: A virtual network interface card (vNIC) with network control
policy with a warning on “Action on Uplink Fail” will not be disabled. Thus,
the NIC needs to be disabled manually for failover.
Disable all uplink interfaces With a single click, this method disables all uplinks that are not in a port
channel. This method needs to be implemented on Ethernet, Fibre Channel,
and FCoE connections.
Ethernet only: As with the disable uplink method, vNICs with network
control policy with a warning on “Action on Uplink Fail” are not disabled.
Manual intervention is required.
Disable all port channels With a single click, this method disables all uplinks that are in a port channel,
links that are not in a port channel are excluded. This method needs to be
implemented on Ethernet, Fibre Channel, and FCoE connections.
Ethernet only: As with the disable uplink method, vNICs with network
control policy with a warning on “Action on Uplink Fail” are not disabled.
Manual intervention is required.
Fabric evacuation This method disables all the server ports, which disables both the vNIC and
virtual host bus adapter (vHBA) for the host. This method does not disable
the virtual interface (VIF) on Cisco UCS C-Series servers that are directly
connected to the fabric interconnect. For those connections, you need to
disable the ports manually.
Note For Cisco UCS domains that use blades only, Cisco recommends the fabric evacuation method.
This approach provides the fastest way to validate proper traffic failover. It also lets you back
out the failover simply by turning the fabric evacuation off.
If Cisco UCS C-Series servers are directly attached to the fabric interconnect, then manually
disable either the port or the uplink for Ethernet, Fibre Channel, and FCoE.
Step 4 Verify that traffic is flowing over the primary fabric interconnect.
Step 5 If the Cisco Nexus 2232PP 10GE is being retired and replaced by Cisco Nexus 2348UPQ 10GE, then
decommission and remove old fabric extender from Cisco UCS Manager.
Step 6 Port schemes between UCS 6200 and 6300 Series FIs are vastly different, so unconfigure all the ports on the
subordinate interconnect and reconfigure the ports on the new fabric interconnect after it joins the cluster.
Note You do not need to delete the port-channel groups because they can be reused, but you should
verify that the individual ports are removed from LAN and SAN port channels, VLAN groups,
etc..
Step 7 Remove the cable from the downed subordinate fabric interconnect.
If you are replacing the IOM and Cisco Nexus 2232 with the Cisco UCS 2304 and Cisco Nexus 2348-UPQ,
then make the replacement now.
Step 8 Install the new Cisco UCS 6332 fabric interconnect and connect the L1 to L2 connections between the Cisco
UCS 6200 Series primary interconnect to the Cisco UCS 6332 platform subordinate interconnect.
Step 9 Reconnect the components (IOM, Cisco Nexus 2200 and 2300 Series, and Cisco UCS C-Series servers) to
the Cisco UCS 6332 platform subordinate interconnect according to your port planning table.
Step 10 Power up the Cisco UCS 6332 platform subordinate interconnect. If it is correctly cabled with the correct
software version, then the subordinate interconnect will recognize that it is connecting to an existing cluster.
Step 11 Enter show cluster extended-state to verify the cluster state.
Step 12 Configure the ports (Fibre Channel ports, server ports, appliance ports, breakout ports, uplink ports, etc.) on
the Cisco UCS 6332 platform subordinate fabric interconnect.
Step 13 Verify server discovery. The IOM discovery process on the Cisco UCS 6332 platform subordinate interconnect
may take a few minutes to complete and become operational. In certain situations, you may need to acknowledge
the IOM (choose Equipment > Chassis > Chassis X > IO Modules > IO Module X).
a) Cisco UCS C-Series servers that are directly connected with dual adapters or are connected to a different
port require server acknowledgment to properly update the port mapping. To reduce the number of server
reboots to one, you can perform this step after the other fabric interconnect has been replaced.
b) For servers connected to the Cisco Nexus 2200 platform fabric extenders, in most cases a fabric extender
reacknowledgment will reestablish a connection. Otherwise, a server reacknowledgment is needed.
Step 14 After all port configuration processes are complete, including the process of adding ports back into LAN and
SAN port channels and VLAN groups, you need to validate LAN and SAN connectivity. Use the appropriate
show commands, such as the following:
Note Note that a Cisco UCS M4 server with a Cisco UCS VIC 1340 and a port expander that is
connected to a Cisco UCS 2304 IOM will be connected as a native 40-Gbps port. You can verify
this connection by entering the show interface status command in the Cisco NX-OS Software
shell.
Step 15 Reestablish flow on the subordinate fabric interconnect, the Cisco UCS 6332 platform.
If a method other than fabric evacuation is used, then enable uplinks for Ethernet, Fibre Channel, and FCoE.
Note Some backplane ports may report link down because no service profile is associated with that
server or the host is powered off.
Step 16 Verify that traffic is flowing normally on the Cisco UCS 6332 platform subordinate interconnect.
Step 17 After verifying traffic flow, promote the Cisco UCS 6332 platform subordinate interconnect to primary status
by entering the cluster lead b command in the local-mgmt shell on the primary fabric interconnect.
Step 18 Enter the show cluster extended-state command to verify that the primary role has switched to the Cisco
UCS 6332 platform and that high availability (HA) is in the ready state.
Step 19 Repeat the steps 3 through 14 to replace the other Cisco UCS 6200 Series Fabric Interconnect.
Step 20 (Optional) Promote fabric interconnect A to the primary role with the command cluster lead a in the local-mgmt
shell of the primary fabric interconnect (B).
Step 21 Check for faults from old configurations, policies, software packages, etc. Remove any noncompliant
configurations, policies, and software packages to clear faults.
Note When a fan is removed, louvers inside the chassis prevent recirculation of air into the system or loss of cooling.
However, you should not leave the chassis for extended periods of time without all fan modules in place.
Step 1 Hold the fan module with the spring latch at the top of the module.
Step 2 Push the fan module into the chassis until it seats properly and the spring latch snaps into place.
Step 3 Listen for the fans if the chassis is powered on. You should immediately hear it operating. If you do not hear
it, ensure that the fan module is inserted completely in the chassis and the faceplate is flush with the outside
surface of the chassis.
Step 4 Verify that the LED behavior is as expected. See LED Locations, on page 20 and Interpreting LEDs, on page
21.
Figure 43: Positioning a Fan Module (N20-FAN5) in the Cisco UCS Server Chassis
KVM Cable
The KVM cable (N20-BKVM) provides a connection into a Cisco UCS blade server, providing a DB9 serial
connector, a VGA connector for a monitor, and dual USB ports for a keyboard and mouse. With this cable
you can create a direct connection to the operating system and the BIOS running on a blade server.
Figure 44: KVM Cable for Blade Servers
4
2 3
1
192621
2 DB9 serial connector 4 2-port USB connector for a mouse and keyboard
Chassis Specifications
Table 17: Cisco UCS Server Chassis
Description Specification
Height x Width x Depth 10.5 in (26.7 cm) x 17.5 in (44.5 cm) x 32 in (81.2 cm)
FEX slots 2
Description Specification
Fully Populated UCS 5108 Server Chassis Approximately 255 lbs (115.66 kg), depending on
models and options selected
2
The system weight listed here is an estimate for a fully configured system and will vary depending on
the devices installed.
Environmental Specifications
Table 19: Environmental Specifications for the Chassis
Description Specification
Case temperature TC 0 50 °C
Description Specification
AC-input voltage Voltage Range 100-120 VAC, 200-240 VAC nominal
(range: 90-132 VAC, 180-264 VAC)
Description Specification
Minimum Software requirement UCS Software Release 2.0(2)
AC-input voltage 200 to 240 VAC nominal (Range: 180 to 264 VAC)
Maximum output power per power supply 2500 W (up to four power supplies)
Description Specification
Maximum inrush current 35 A (sub cycle duration)
Item Specification
Minimum software requirement Cisco UCS Software Release 2.0(1)
Capability Catalog Version 42
Item Specification
Maximum holdup time 8 ms at 50% load
4 ms at 100% load
Item Specification
DC-input voltage 200 to 380 VDC nominal (Range: 180 to 400 VDC)
Table 25: AC-input Dual Voltage Platinum Power Supply (UCSB-PSU-2500ACDV) Specifications
Description Specification
AC-input voltage 200 to 240 VAC nominal (Range: 180 to 264 VAC)
100 to 120 VAC nominal (Range: 90 to 127 VAC)
supported only with the UCS 6324 Fabric Interconnect
Maximum output power per power supply 2500 W @ 200 to 240 VAC
1300 W @ 100 to 120 VAC
Description Specification
Power supply standby voltage 3.3 VDC @ 5A
DC wiring must meet your local codes and regulations, we recommend using a licensed local electrician to
install the DC wiring needed.
To determine the number of power supply units needed for the blade server, remember that each single slot
server is budgeted a max 550 W and each full width server is budgeted a max 1100 W. For a more detailed
estimate, contact Cisco Sales.
For information about supported power cords, see the Cisco UCS 5108 Server Chassis Installation Guide.
Note Only the regular power cords or jumper power cords provided with the chassis are supported.
Figure 45: CAB-AC-16A-AUS Power Cord for the Cisco UCS 5108 Blade Server Chassis
Continental Europe
Power Cord Part Number—CAB-AC-2500W-EU
Cord Set Rating—16A, 250 VAC
Figure 46: CAB-AC-2500W-EU Power Cord for the UCS 5108 Blade Server Chassis
International
Power Cord Part Number—CAB-AC-2500W-INT
Cord Set Rating—16A, 250 VAC
Figure 47: CAB-AC-2500W-INT Power Cord for the UCS 5108 Blade Server Chassis
Israel
Power Cord Part Number—CAB-AC-2500W-ISRL
Cord Set Rating—16A, 250 VAC
Figure 48: CAB-AC-2500W-ISRL Power Cord for the UCS 5108 Blade Server Chassis
Figure 49: CAB-AC-2500W-US1 Power Cord for the UCS 5108 Blade Server Chassis
Taiwan
Power Cord—CAB-AC-C19-TW
Plug—250 VAC 16 A, C19
Length—7.5 feet / 2.3 meters
Switzerland
Power Cord Part Number—CAB-ACS-16
Cord Set Rating—16A, 250 VAC
Figure 52: CAB-ACS-16 Power Cord for the UCS 5108 Blade Server Chassis
Note For information about how to query the chassis for configuration information, see the Cisco UCS Configuration
Guide.
This appendix includes the following records to use when installing the Cisco UCS server chassis:
• Site Preparation Checklist, on page 93
• Contact and Site Information, on page 95
• Chassis and Module Information, on page 95
• FEX Port Connection Record, on page 96
• UCS 6324 Fabric Interconnect Port Connection Record, on page 97
1 Space evaluation:
• Space and layout
• Floor covering
• Impact and vibration
• Lighting
• Maintenance access
2 Environmental evaluation:
• Ambient temperature
• Humidity
• Altitude
• Atmospheric contamination
• Air flow
3 Power evaluation:
• Input power type
• Power receptacles
• Receptacle proximity to the
equipment
• Dedicated circuit for power supply
• Dedicated (separate) circuits for
redundant power supplies
• UPS for power failures
4 Grounding evaluation:
• Circuit breaker size
• CO ground (AC- powered systems)
6 EMI evaluation:
• Distance limitations for signaling
• Site wiring
• RFI levels
3
Verify that the power supply installed in the chassis has a dedicated AC source circuit.
4
UPS: uninterruptable power supply.
5
EMI: electromagnetic interference.
6
RFI: radio frequency interference.
Contact person
Contact phone
Contact e-mail
Building/site name
Floor location
Address (line 1)
Address (line 2)
City
State
Zip code
Country
Blade Server–2
Blade Server–3
Blade Server–4
Blade Server–5
Blade Server–7
Blade Server–8
I/O Module-1
I/O Module-2
Note The serial numbers of all server chassis modules can be obtained using Cisco UCS Manager.
FEX Connected to
1 1
FEX Connected to
2 1
1 1
2 1