Cisco UCS B22 M3 Blade Server
The UCS B22 M3 blade server is a half-width blade with 12 DIMM slots; it supports one dedicated slot for Cisco's Virtual Interface Card (VIC) 1240, and one open adapter slot. You may install up to eight UCS B22 Blade Servers to a UCS chassis, or mix with other UCS blade servers. The Cisco UCS B22 M3 harnesses the power of the latest Intel Xeon processor E5-2400 product family with expandability to 192 GB of RAM (using 12 x 16 GB DIMMs), 2 hot-plug drives, and 2 PCIe adapter slots for up to 80 Gigabit Ethernet throughput.
Each server has a blank plastic asset tag that pulls out of the front panel, provided so you can add your own asset tracking label without interfering with the intended air flow.
Server LEDs indicate whether the blade server is in active or standby mode, the status of the network link, the over all health of the blade server, and whether the server is set to give a flashing blue beaconing indication.
Table 1 Blade Server LEDs
The Reset button is just inside the chassis and must be pressed using the tip of a paper clip or a similar item. Hold the button down for five seconds, and then release it to restart the server if other methods of restarting are not working.
The power button and LED allows you to manually take a server temporarily out of service but leave it in a state where it can be restarted quickly. If the desired power state for a service profile associated with a blade server or an integrated rack-mount server is set to "off", using the power button or Cisco UCS Manager to reset the server will cause the desired power state of the server to become out of sync with the actual power state and the server may unexpected shutdown at a later time. To safely reboot a server from a power-down state, use the Boot Server action in Cisco UCS Manager.
The console port allows a direct connection to a blade server to allow operating system installation and other management tasks to be done directly rather than remotely. The port uses the KVM dongle cable (N20-BKVM) which provides a connection into a Cisco UCS blade server; it has 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. A KVM cable ships standard with each blade chassis accessory kit.
Hard Drive Replacement
You can remove blade server hard drives without removing the blade server from the chassis.
The drives supported in this blade server come with the drive sled attached. Spare drive sleds are not available. A list of currently supported drives is in the specification sheets at: http://www.cisco.com/en/US/products/ps10280/products_data_sheets_list.html
Before upgrading or adding an HDD to a running system, check the service profile in Cisco UCS Manager and make sure the new hardware configuration will be within the parameters allowed by the service profile.
Each blade contains an LSI SAS 2002 RAID controller embedded in the motherboard that is not separately replaceable. The controller supports RAID 0 and 1.
If the drive being replaced was part of a RAID array, Cisco recommends using a new drive of identical size, model, and manufacturer to replace the failed drive. This recommendation comes from the industry standard practice of using drives of the same capacity when creating RAID volumes. If drives of different capacities are used, the useable portion of the smallest drive will be used on all drives that make up the RAID volume.
If you ever need to move a RAID cluster from one server to another, both the old and new servers for the cluster must use the same LSI controller. For example, migration from a server with an LSI 1064E controller to a server with an LSI MegaRAID controller is not supported. Similarly, migrating a RAID cluster from a B200 M3 to a B420 M3 is not supported.
For hard disk and RAID troubleshooting information, see the Cisco UCS Manager B-Series Troubleshooting Guide.
Removing a Blade Server Hard Drive
Installing a Blade Server Hard DriveProcedure
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.
You can use Cisco UCS Manager to format and configure RAID services. For details, 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
If you need to move a RAID cluster, see the Cisco UCS Manager B-Series Troubleshooting Guide.
Blade Server Removal and Installation
Powering Off Blade Servers Using the Power ButtonProcedure
You can also shut the server down remotely using Cisco UCS Manager. For details, 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
Removing a Blade ServerProcedure
Step 1 Loosen the captive screw on the front of the blade. Step 2 Remove the blade from the chassis by pulling the ejector lever on the blade until it unseats the blade server. Step 3 Slide the blade part of the way out of the chassis, and place your other hand under the blade to support its weight. Step 4 Once removed, place the blade on an antistatic mat or antistatic foam if you are not immediately reinstalling it into another slot. Step 5 If the slot is to remain empty, install a blank faceplate (N20-CBLKB1) to keep dust out of the chassis.
Installing a Half-width Blade ServerProcedure
UCS B200 M3 and UCS B22 M3 half-width blade servers are interoperable in a UCS chassis with any other UCS blade servers, including prior generation B200 M2 and B200 M1 servers, or other UCS B-Series blade servers. To install a half-width blade server, follow these steps:
Step 1 Grasp the front of the blade server and place your other hand under the blade to support it. Step 2 Open the ejector lever in the front of the blade server. Step 3 Gently slide the blade into the opening until you cannot push it any farther. Step 4 Press the ejector lever so that it catches the edge of the chassis and presses the blade server all the way in. Step 5 Tighten the captive screw on the front of the blade to no more than 3 in-lbs. Tightening with bare fingers only is unlikely to lead to stripped or damaged captive screws. Step 6 Power on the server. UCS Manager automatically reacknowledges, reassociates, and recommissions the server, provided any hardware changes are allowed by the service profile.
Secure Digital (SD) Card Access
Removing a Blade Server CoverProcedure
Cisco UCS-USBFLSH-S-4GB= is recommended, but if another USB drive will be used it must be no wider than .8 inches, and no more than 1.345 inches long in order to provide needed clearances to install or remove the USB drive.
Use of this server may require an upgrade to the IOM in the chassis. This server only supports third generation adapter cards, which have features requiring a Cisco 2204 or 2208 IOM, and are not backward compatible with the Cisco 2104 IOM.
Diagnostics Button and LEDs
At blade start-up, POST diagnostics test the CPUs, DIMMs, HDDs, and adapter cards, and any failure notifications are sent to UCS Manager. You can view these notifications in the System Error Log or in the output of the show tech-support command. If errors are found, an amber diagnostic LED also lights up next to the failed component. During run time, the blade BIOS, component drivers, and OS all monitor for hardware faults and will light up the amber diagnostic LED for a component if an uncorrectable error or correctable errors (such as a host ECC error) over the allowed threshold occur.
LED states are saved, and if you remove the blade from the chassis the LED values will persist for up to 10 minutes. Pressing the LED diagnostics button on the motherboard causes the LEDs that currently show a component fault to light for up to 30 seconds for easier component identification. LED fault values are reset when the blade is reinserted into the chassis and booted, and the process begins from its start.
If DIMM insertion errors are detected, they may cause the blade discovery to fail and errors will be reported in the server POST information, which is viewable using the UCS Manager GUI or CLI. UCS blade servers require specific rules to be followed when populating DIMMs in a blade server, and the rules depend on the blade server model. Refer to the documentation for a specific blade server for those rules.
HDD status LEDs are on the front face of the HDD. Faults on the CPU, DIMMs, or adapter cards also cause the server health LED to light solid amber for minor error conditions or blinking amber for critical error conditions.
Working Inside the Blade Server
Installing a Motherboard CMOS BatteryProcedure
There is danger of explosion if the battery is replaced incorrectly. Replace the battery only with the same or equivalent type recommended by the manufacturer. Dispose of used batteries according to the manufacturer’s instructions. Statement 1015
Step 1 Remove the old CMOS battery: Step 2 Install a motherboard CMOS battery:
- Push the battery socket retaining clip away from where the battery fits in the housing.
- Insert the new battery into the socket with the battery’s positive (+) marking facing away from the retaining clip. Ensure that the retaining clip can click over the top of the battery to secure it in the housing.
- Replace the top cover.
- Replace the server in the chassis and power on the blade by pressing the Power button.
You can order your blade server with two CPUs, or upgrade later to a second CPU. Both CPUs must be of the same type, and memory in slots intended for the second CPU will not be recognized if the second CPU is not present. You may need to use these procedures to move a CPU from one server to another, or to replace a faulty CPU.
Removing a CPU and Heat SinkProcedure
The CPU pick and place tool is required to prevent damage to the connection pins between the motherboard and the CPU. Do not attempt this procedure without the required tool, which is included with each CPU option kit.
Step 1 Unscrew the four captive screws securing the heat sink to the motherboard. See callout 1.
Loosen one screw by a quarter turn, then move to the next in an X pattern. Continue loosening until the heat sink can be lifted off.
Step 2 Remove the heat sink (UCSB-HS-01-EP). See callout 2 in the following figure.
Remove the existing thermal compound from the bottom of the heat sink using the cleaning kit (UCSX-HSCK= ) included with each CPU option kit. Follow the instructions on the two bottles of cleaning solvent.
Step 3 Unhook the socket hook. See callout 3 in the following figure. Step 4 Open the socket latch. See callout 4 in the following figure. Step 5 Place the CPU and tool on the CPU socket with the registration marks aligned as shown.
The CPU pick and place tool is included with each CPU option kit, or the tool may be purchased separately. Be sure to use the tool for the Intel Xeon E5-2400 Series processors.
Step 6 Remove an old CPU as follows:
Installing a New CPU and Heat SinkProcedure
- The CPU is supported for that given server model. This may be verified via the server's Technical Specifications ordering guides or by the relevant release of the Cisco UCS Capability Catalog.
- A BIOS update is available and installed that supports the CPU and the given server configuration.
- If the server will be managed by Cisco UCS Manager, the service profile for this server in UCS Manager will recognize and allow the new CPU.
Step 1 (Optional)If you are installing a CPU in a socket that had been shipped empty, there will be a protective cap intended to prevent bent or touched contact pins. The pick and pull cap tool provided can be used in a manner similar to a pair of tweezers. Grasp the protective cap and pivot as shown. Step 2 Release the catch on the pick and place tool by pressing the button/handle. Step 3 Remove the new CPU from the packaging, and load it into the pick and place tool as follows:
- Confirm that the pedestal is set up correctly for your processor. The pedestal ships set to the markings “LGA1356-2,” and this is the correct setting.
- Place the CPU on the pedestal. The CPU corners should fit snugly at the pedestal corners and the notches should meet the pegs perfectly.
- Place the CPU pick and place tool on the CPU pedestal aligned with the A1 arrow pointing to the A1 registration mark. Make sure the tabs on the tool are fully seated into the slots on the pedestal.
- Press the button/handle on the tool to grasp the CPU.
- Lift the tool and CPU straight up off of the pedestal.
Step 4 Place the CPU and tool on the CPU socket with the registration marks aligned as shown. Step 5 Press the button on the pick and place tool to release the CPU into the socket. Step 6 Close the socket latch. See callout 1. Step 7 Secure the hook. See callout 2. Step 8 Using the syringe of thermal grease provided with replacement CPUs and servers (and available separately as UCS-CPU-GREASE=), add 2 cubic centimeters of thermal grease to the top of the CPU where it will contact the heat sink. Use the pattern shown below. This should require half the contents of the syringe. Step 9 Replace the heat sink. See callout 3.
Caution Step 10 Secure the heat sink to the motherboard by tightening the four captive screws a quarter turn at a time in an X pattern as shown in the upper right of the figure.
Step 1 Open both DIMM connector latches. Step 2 Press the DIMM into its slot evenly on both ends until it clicks into place.
Note Step 3 Press the DIMM connector latches inward slightly to seat them fully.
The DIMMs supported in this blade server are constantly being updated. A list of currently supported and available drives is in the specification sheets at:http://www.cisco.com/en/US/products/ps10280/products_data_sheets_list.html
Low-Voltage DIMM ConsiderationsProcedure
Step 1 Enter the BIOS setup utility by pressing the F2 key when prompted during bootup. Step 2 Select the Advanced tab. Step 3 Select Low Voltage DDR Mode. Step 4 In the pop-up window, select either Power Saving Mode or Performance Mode. Step 5 Press F10 to save your changes and exit the setup utility, or you can exit without saving changes by pressing Esc.
- Only use Cisco-provided DIMMs.
- Mixing different speed DIMMs causes the server to set the memory speed to that of the slowest installed DIMMs.
- If memory mirroring is used, the total memory capacity is reduced by 50%.
- CPU1 and CPU2 (if used) must always be configured identically.
Each channel is identified by a letter—B, C, D for CPU1, and F, G, H for CPU 2. Each DIMM slot is identified by a number, either 0 or 1. Note that each DIMM slot 0 is blue, while each slot 1 is black.
The following figure shows how DIMMs and channels are physically laid out on the blade server. The DIMM slots in the upper right are associated with CPU 2, while the DIMM slots in the lower left are associated with CPU 1.Figure 18. Numbering of DIMMs and Channels
Each CPU in a Cisco UCS B22 blade server supports 3 channels of 2 memory slots each. In a 1 DPC configuration, DIMMs are in slot 0 only. In a 2 DPC configuration, DIMMs are in both slot 0 and slot 1. The table below shows the preferred order for installing upgrade DIMMs, and while other configurations may work if problems arise moving them to the preferred arrangement should help.
Bandwidth and Performance
- Ensure that both the installed CPU and the selected DIMMs support operation at 1600 MHz. If either cannot support this, the pair will run at the highest speeed of the slower of the two.
- Ensure the server is running the 2.0(2) or later BIOS version. If a BIOS upgrade is needed, do it before installing processors or memory.
- Use only Cisco certified DIMMs that support 1600 MHz speeds. DIMMs do not have to be identical in type or capacity, but beware of the caveats listed in the section below regarding performance degradation.
- Always set the system BIOS to operate the DIMMs in "Performance" mode in order to run at 1600 MHz.
- Fully populating 1 logical bank or 2 logical banks with DIMMs will ensure optimal memory bandwidth running at the 1600 MHz speed. If DIMMs are partially populated in 1 bank (less than 6 DIMMs) or 2 bank patterns (less than 12 but greater than 6 DIMMs) the 1600 MHz speed can be used, but the overall memory bandwidth will not be optimal.
Memory Mirroring and RAS
The Intel CPUs within the blade server support memory mirroring only when no more than two Channels are populated with DIMMs. If three Channels are populated with DIMMs, memory mirroring is automatically disabled. Furthermore, if memory mirroring is used, DRAM size is reduced by 50% for reasons of reliability.
Installing a Modular LOMProcedure
The Cisco VIC 1240 is a specialized modular Lan on Motherboard (mLOM) adapter that provides dual 2 x 10 Gb of Ethernet/ or Fiber Channel over Ethernet (FCoE) connectivity to each chassis. It plugs into the dedicated mLOM connector only. It is currently the only card that can be plugged into the mLOM connector and it will provide connectivity through either a 2100 series or 2200 series IOM.
Step 1 Position the mLOM’s board connector above the motherboard connector and align the captive screw to the standoff post on the motherboard. Step 2 Firmly press the modular LOM’s board connector into the motherboard connector. Step 3 Tighten the captive screw.
To remove a modular LOM, reverse the above procedure. You might find it helpful when removing the connector from the motherboard to gently rock the board along the length of the connector until it loosens.
Installing an Adapter CardProcedure
The network adapters and interface cards all have a shared installation process and are constantly being updated. A list of currently supported and available models for this server is in the specification sheets at this URL:
Use of the adapters available for this server might require an upgrade to the FEX in the chassis. The 2104XP FEX is not compatible with any Cisco-certified adapter. If a VIC 1240 mLOM card is installed, you will have connectivity through the mLOM but other adapters will not be recognized. Use of all slots requires Cisco UCS 2200 series FEXes.
If you are switching from one type of adapter card to another, before you physically perform the switch make sure that you download the latest device drivers and load them into the server’s operating system. For more information, see the firmware management chapter of one of the Cisco UCS Manager software configuration guides.
The Cisco UCS 785GB or 365GB MLC Fusion-io Drive and LSI 400GB SLC WarpDrive have the same form factor as M3 adapter cards and can be installed and removed using the same procedures. Using these drives in a B200 M3 or B22 M3 blade server requires the presence of a VIC 1240 mLOM to provide blade I/O. They will not work in M1 and M2 generation Cisco UCS servers. These drives appear in Cisco UCS Manager as regular SSDs.
Step 1 Position the adapter board connector above the motherboard connector and align the two adapter captive screws to the standoff posts (see callout 1) on the motherboard. Step 2 Firmly press the adapter connector into the motherboard connector (see callout 2). Step 3 Tighten the two captive screws (see callout 3).
Removing an adapter card is the reverse of installing it. You might find it helpful when removing the connector from the motherboard to gently rock the board along the length of the connector until it loosens.
Enabling a Trusted Platform ModuleProcedure
The Trusted Platform Module (TPM, Cisco Product ID UCSX-TPM1-001) is a component that can securely store artifacts used to authenticate the server. These artifacts can include passwords, certificates, or encryption keys. A TPM can also be used to store platform measurements that help ensure that the platform remains trustworthy. Authentication (ensuring that the platform can prove that it is what it claims to be) and attestation (a process helping to prove that a platform is trustworthy and has not been breached) are necessary steps to ensure safer computing in all environments.
A TPM is a factory-configurable option for this server. It is a requirement for the Intel Trusted Execution Technology (TXT) security feature, which must be enabled in the BIOS settings for a server equipped with a TPM. A TPM can not be added later by customers, or moved from one server to another.
Step 1 Enable Quiet Mode in the BIOS policy of the server’s Service Profile. Step 2 Establish a direct connection to the server, either by connecting a keyboard, monitor, and mouse to the front panel using a KVM dongle (N20-BKVM) or by other means. Step 3 Reboot the server. Step 4 Press F2 during reboot to enter the BIOS setup screens. Step 5 On the Advanced tab, select Trusted Computing and press Enter. Step 6 Set the TPM Support option to Enable. Step 7 Press F10 to save and exit. Allow the server to finish booting.
For general troubleshooting information, see the see the Cisco UCS Manager B-Series Troubleshooting Guide.
Cisco UCS blade servers are intended to be configured and managed using Cisco UCS Manager. For details, 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
Physical Specifications for the Cisco UCS B22 M3
Copyright © 2012-2013, Cisco Systems, Inc. All rights reserved.