Release Notes for Cisco UCS C-Series Software, Release 1.4(2)
Available Languages
Table of Contents
Release Notes for Cisco UCS C-Series Software, Release 1.4(2)
Overview of the Pre-Installed Cisco Flexible Flash Card
Hardware and Software Interoperability
Upgrading BIOS and CIMC firmware
Caveats from Previous Releases
Obtaining Documentation and Submitting a Service Request
Release Notes for Cisco UCS C-Series Software, Release 1.4(2)
First Published Date: Nov 04, 2011
This document describes the new features, system requirements, open caveats and known behaviors for C- series software release 1.4(2) including Cisco Integrated Management Controller software and any related BIOS, firmware, or drivers. Use this document in conjunction with the documents listed in the “Related Documentation” section.
Note We sometimes update the documentation after original publication. Therefore, you should also review the documentation on Cisco.com for any updates.
Table 1 shows the online change history for this document.
Contents
Introduction
Cisco® UCS C-Series Rack-Mount Servers extend unified computing innovations to an industry-standard form factor to help reduce total cost of ownership (TCO) and increase business agility. Designed to operate both in standalone environments and as part of the Cisco Unified Computing System™1, the series employs Cisco technology to help customers handle the most challenging workloads. The series incorporates a standards-based unified network fabric, Cisco VN-Link virtualization support, and Cisco Extended Memory Technology. It supports an incremental deployment model and protects customer investments with a future migration path to unified computing.
The Cisco UCS C460 High-Performance Rack-Mount Server is designed with the performance and reliability to power compute-intensive, enterprise-critical standalone applications and virtualized workloads. The system is a four-rack-unit (4RU) rack-mount server supporting up to four Intel Xeon 7500 series processors, up to 512 GB of DDR3 memory in 64 slots, and 12 small form-factor (SFF) hot-pluggable SAS and SATA disk drives. Abundant I/O capability is provided by 10 PCI Express (PCIe) slots supporting the Cisco UCS C-Series network adapters, with an eleventh PCIe slot reserved for a hard disk drive array controller card. Additional I/O is provided by two Gigabit Ethernet LAN-on-motherboard (LOM) ports, two 10 Gigabit Ethernet ports, and two dedicated out-of-band (OOB) management ports.
The Cisco UCS C260 High-Performance Rack-Mount Server is designed with the performance and reliability to power compute-intensive, enterprise-critical standalone applications and virtualized workloads. The system is a two-rack-unit (2RU) rack-mount server supporting up to two Intel Xeon 7500 series processors, up to 1 TB of DDR3 memory in 64 slots, and 16 small form-factor (SFF) hot-pluggable SAS and SATA disk drives. Abundant I/O capability is provided by 7 PCI Express (PCIe) slots supporting the Cisco UCS C-Series network adapters and hard disk drive array controller cards. Additional I/O is provided by two Gigabit Ethernet LAN-on-motherboard (LOM) ports, two optional 10 Gigabit Ethernet LOM ports, and two dedicated out-of-band (OOB) management ports.
This server is shipped from the factory with one pre-installed Cisco Flexible Flash card. The slots for these cards are on the I/O riser.
The Cisco UCS C250 Extended-Memory Rack-Mount Server is a two-socket, two-rack-unit (2RU) rack-mount server featuring patented Cisco Extended Memory Technology. It is designed to increase performance and capacity for demanding virtualization and large-data-set workloads. It also can reduce the cost of smaller memory footprints. This server is built for virtualized workloads in enterprise data centers, service provider environments, and virtual desktop hosting. The system also helps increase performance for large-data-set workloads, including database management systems and modeling and simulation applications. Applications that are memory bound today will benefit by the 384 GB of addressable memory that the Cisco UCS C250 server offers.
The Cisco UCS C210 General-Purpose Rack-Mount Server is a general-purpose, two-socket, two-rack-unit (2RU) rack-mount server housing up to 16 internal small form-factor (SFF) SAS or SATA disk drives for a total of up to 32 terabytes (TB) of storage. The Cisco UCS C210 server is designed to balance performance, density, and efficiency for workloads requiring economical, high-capacity, reliable, internal storage. Based on quad-core Intel® Xeon® 5500 or 5600 series processors, the server is built for applications including virtualization, network file servers and appliances, storage servers, database servers, and content-delivery servers.
The Cisco UCS C200 High-Density Rack-Mount Server is a high-density server with balanced compute performance and I/O flexibility. This price-to-performance optimized two-socket, one-rack-unit (1RU) rack-mount server is designed to balance simplicity, performance, and density for web infrastructure and mainstream data center, small-office, and remote-office applications. Its single-rack-unit size makes it useful for service providers offering dedicated or multi-tenant hosting, and its economical price makes it well suited to the appliance market.
C-series servers are capable of standalone management using Cisco Integrated Management Controller (CIMC), which is an integration of the service processor hardware and the CIMC firmware. CIMC has a Web GUI and a CLI mode.
The Cisco UCS C200 M2 Small Form Factor (SFF) server is a factory-configurable option that is an alternate to the existing Cisco UCS C200 M2 Large Form Factor (LFF) server.
The LFF version of the server can contain up to four 3.5-inch LFF hard drives or solid state drives. If a customer orders the new SFF version of the server, the chassis is configured with a drive backplane and front drive bays that can contain up to eight 2.5-inch SFF drives. The drive bays and backplane are factory-configurable, but are not supported as field-configurable.
The SFF server also allows configurable selection of front-panel control modules. Customers can choose the standard control panel module or an optional DVD-drive module. These two modules are field-replaceable and interchangeable with the supplied cables.
In addition, 2.5-inch drives have been qualified as configurable options for the SFF server. The RAID controller options have also been qualified for the SFF server (LSI MegaRAID 9280-4i4e and LSI 1068-based mezzanine card).
Overview of the Pre-Installed Cisco Flexible Flash Card
The Cisco Flexible Flash card is pre-installed with three software bundles, each on one of four preconfigured virtual drives (VDs). The fourth VD allows you to install an OS or an embedded hypervisor.
The VDs are configured with the following content:
- Cisco UCS Server Configuration Utility (SCU).
- Hypervisor (HV). This is a VD that you can use for your own purposes.
- Cisco Drivers (Drivers).
- Cisco Host Upgrade Utility (HUU).
Refer to the following documents for more information about these tasks:
- Replacing a card: Cisco UCS C260 Server Installation and Service Guide
- Enabling and booting a VD: Cisco UCS C-Series Rack-Mount Server Configuration Guide or the Cisco UCS C-Series Rack-Mount Server CLI Configuration Guide
- Monitoring and managing a card with CIMC: Cisco UCS C-Series Rack-Mount Server Configuration Guide or the Cisco UCS C-Series Rack-Mount Server CLI Configuration Guide
The links to these documents are in the C-Series documentation road map:
Hardware and Software Interoperability
For detailed information about storage switch, operating system, adapter, adapter utility, and storage array interoperability, see the Hardware and Software Interoperability Matrix for your release located at:
http://www.cisco.com/en/US/products/ps10477/prod_technical_reference_list.html
Transceivers Specifications
The Cisco UCS C-Series servers supports a wide variety of 10 Gigabit Ethernet connectivity options using Cisco 10GBASE SFP+ modules.
Table 2 and Table 3 details the controllers and the supported transcievers.
Firmware Files
The C-Series software release 1.4(2) includes the following software files:
Note Always upgrade both the BIOS and the CIMC from the HUU ISO. Do not upgrade individual components (only BIOS or only CIMC), since this could lead to unexpected behavior.
Host Upgrade Utility
The Cisco Host Upgrade Utility (HUU) is a tool that upgrades the following firmware:
- Cisco Integrated Management Controller (CIMC)
- System BIOS
- LAN on motherboard (LOM)
- RAID Controllers (LSI storage controllers)
- Cisco UCS P81E Virtual Interface Card (VIC)
- Broadcom PCIe adapters
– 5709 Dual and Quad port adapters
In addition, this utility now supports the following options:
- Download ISO images for a selected platform on a Windows operating system.
- Recover a corrupt BIOS in the EFI shell.
The image file for the firmware is embedded in the ISO. The utility displays a menu that allows you to choose which firmware components to upgrade. For more information on this utility see:
http://www.cisco.com/en/US/products/ps10493/products_user_guide_list.html
Starting with this 1.4 release, separate ISO images of Host Upgrade Utility are available for different server platforms.
The ISO image is now named as ucs-<server_platform>-huu-<version_number>.iso.
The Cisco Host Upgrade Utility contains the following files:
System Requirements
The management client must meet or exceed the following minimum system requirements:
Features in this Release
In this release, the features of 1.4(1) are introduced on C260 and C460 server platforms. This section briefly describes the feature set of 1.4(1) that are now available on these new server platforms:
– The CIMC XML API is a programmatic interface to the Cisco Integrated Management Controller (CIMC) in a standalone C-Series server.
– The CIMC XML API is implemented to behave similar to the UCSM XML API. Customers familiar with using the UCSM XML API can reuse the same client development techniques and tools to integrate with a C-Series server.
– The CIMC XML API only implements a subset of the methods and functionality of the UCSM XML API.
– The CIMC inventory, state and configuration information is modeled in a hierarchical tree structure known as the Management Information Tree (MIT).
– Group Authorization and searching nested groups
– Allow for multiple Domain Controllers
– Allow for multiple Global Catalogs
– Inventory and fault information delivery for fan, power supply, CPU, memory, storage physical drive, storage virtual drive, battery back-up unit for the storage controller, and supported storage controllers.
– SNMP V1, V2c and V3 authentication.
– MD5/SHA authentication along with AES/DES encryption available for SNMP V3 security.
– Access to LAN on Motherboard MACs.
– Support for configuration of FCoE over NIV for UCS P81E Virtual Interface Card.
– Support for VM FEXs on P81E Virtual Interface Card.
– Control of Hard drive locator LED (MegaRAID enabled adapters only).
– Per platform Host Upgrade Utility ISO.
– Support for firmware upgrades on plug-in PCIe adapter cards
- Broadcom 5709-Dual-Port
- Broadcom 5709-Quad-Port
- Broadcom 57711-Dual-Port
- Broadcom 57712-Dual-Port
- Intel® PRO/1000 PF Quad Port Server Adapter
SNMP
The supported MIB definition for release 1.4(2) can be found at the following link: ftp://ftp.cisco.com/pub/mibs/supportlists/ucs/ucs-C-supportlist.html
Note The above link is incompatible with IE 9.0.
BIOS/Hardware
Symptom Cisco is introducing a new generation of SATA 2.5-inch hard drives which are for the first time compatible with the new SATA Revision 3.0 standard capable of 6 Gbit/s. In particular these hard drives are SATA 500 GB (A03-D500GC3) and 1 TB (A03-D1TBSATA). Due to legacy design constraints in certain Generation M2 C-Series servers, the faster 6 Gbit/s link speeds may not always be possible in all servers. However, in those cases, the link speeds will properly negotiate the link back to the Revision 2.0 standard capability of 3 Gbit/s. This operation is seamless to the user.
Workaround This is a known behavior, as follows:
- The UCS C200 M2 SFF server supports these drives at full 6Gbit/s speeds.
- The UCS C210 M2 server guarantees support of these drives at a minimum of 3Gbit/s speeds.
- The UCS C250 M2 server guarantees support of these drives at a minimum of 3Gbit/s speeds.
- The UCS C460 server supports these drives at full 6Gbit/s speeds.
- The UCS C260 server supports these drivers at full 6Gbit/s speeds.
Note In some cases, you may find that some or all links have negotiated to the higher 6Gbit/s speeds. Those links that do, will operate properly at that speed.
Upgrading BIOS and CIMC firmware
Caution When you upgrade the BIOS firmware, you must also upgrade the CIMC firmware from the same HUU ISO, or the server may not boot. Do not power off the server until the BIOS and CIMC firmware are updated.
Cisco provides the Cisco Host Upgrade Utility to assist you in upgrading the BIOS, CIMC, LOM, LSI storage controller, and Cisco UCS P81E Virtual Interface Card firmware to compatible levels.
The correct and compatible firmware levels for your server model are embedded in the utility ISO.
To use this utility, start with the Cisco Host Upgrade Utility User Guide, which includes the instructions for downloading and using the utility ISO. Select the correct guide for your desired firmware version from this URL:
http://www.cisco.com/en/US/products/ps10493/products_user_guide_list.html
Resolved Caveats
This section lists the resolved caveats for the 1.4(2x) release, where ‘x’ indicates the revision character.
The following caveats were resolved in the 1.4(2) release:
CIMC
Symptom When “ipmitool sdr” command is entered at the command prompt, an unspecified error is observed.
[root@localhost ~]# ipmitool sdr
Get SDR xxxx command failed: Unspecified error
Symptom The CIMC firmware upgrade utility can cause an error on the CIMC that will cause System Event Log (SEL) messages to be repeated as if the CIMC had just been restarted.
Workaround None. Firmware upgrade will complete successfully, but dozens of extra redundant SEL messages can occur. (CSCtl12301)
Symptom When you reboot/upgrade the CIMC firmware through the WebGUI, the WebGUI will redirect to login page. When you try to login, you will get a message “session is already active, please logout to clear the existing session”. You will get the same message, even after you logout, and login again.
Workaround You must clear the cookies/cache to login successfully. (CSCts39518)
LSI
Symptom SNMP trap is not received when check consistency has failed. This issue is seen with certain firmware releases of the LSI Storage controller.
Workaround Update the LSI storage controller firmware. (CSCtr90017)
BIOS
Symptom Attempt to reset the C250 server from the BMC causes spurious ECC uncorrectable errors.
Workaround Do not reset the server from the BMC, instead power down and power up the server. (CSCtt35212)
BIOS
Symptom If you enter into the LSI Option ROM Configuration utility WebBIOS, by pressing <CNTRL-H> during boot, the BIOS boot order changes. So after exiting the WebBIOS and rebooting, the system might not boot in the same order as expected previously.
Workaround After exiting the WebBIOS (which reboots the system anyway), update the Boot order by entering BIOS Setup. Follow these steps:
Step 1 Press F2 to enter the BIOS Setup.
Step 2 Goto Boot Option Tab, select “Hard Disk Order” and update the boot order as required.
Step 3 Update the Boot Option #1, #2, and so on.
Step 4 Save and exit (F10). (CSCta56788)
Symptom Some boot order configurations can not be applied as expected using only the CIMC interfaces, the server BIOS settings will have an affect that may be unexpected. If for example you specify a boot order of CDROM, HDD, and the BIOS setting is HDD, CDROM, PXE, EFI in BIOS, the applied configuration will be CDROM, HDD, PXE, EFI. Another example would be that if you do not have FDD devices and specify a boot order in CISM of FDD, CDROM, HDD, the FDD will not be appllied. The resulting configuration will be CDROM, HDD.
Symptom When VT-UTF8 is selected as a terminal type in the BIOS Setup, junk characters are seen on the terminal.
Workaround Do not select the VT-UTF8 terminal type, instead use the VT100 default mode. ( CSCtb25124 )
CIMC
Symptom Immediately after switching off the main CPU in C460 M1 rack server, the CIMC becomes unresponsive.
Workaround Restart the CIMC. (CSCtl56864)
Symptom On some UCS C210 servers, FAN4 and FAN5 TACH sensors are showing status as CRITICAL and the speed is 0.
Symptom BIOS configuration is not saved in CIMC exported configuration.
Symptom Gradual memory leak over a period of time is seen in CIMC, which may lead to Out of Memory (OOM) scenario. This is observed when the SNMP feature is enabled.
Workaround None. This has been resolved in 1.3(3) and 1.4(1). (CSCto53530)
Symptom Disabling SNMP feature from Cisco CLI fails.
Workaround Use WebGUI to disable SNMP feature. This is available in the page Admin->Communications services. (CSCtq28791)
Symptom If you enter special characters for a VNIC name like “(“or”)” then a core is generated.
Workaround Special characters should be assigned in quotes. For example, If assigning cd()() as the name of the VNIC, it should be enclosed in quotes in the GUI as “cd()()”. (CSCto54995)
Symptom Image files created by the Virtual Media subsystem are created in the user's desktop directory.
Workaround There is no workaround. Currently image files are always created in the user's desktop directory. Once the image file is created a user can move it anywhere, but the creation process is currently in a fixed location. (CSCtk59127)
LOM
Symptom C460 10GE LOM ports are not supported in RHEL 5.6. Broadcom netxtreme2-6.0.XX Linux drivers cannot be compiled in RHEL 5.6.
Workaround Broadcom does not support RHEL5.6 in this release. (CSCtn17919)
LSI
Symptom LSI Storage controllers on C-460 M1, C200 M2, C250-M2 servers running ESX4.1 and ESXi4.1 cannot be managed using the host application MegaRAID Storage Manager running on remote clients.
Workaround The LSI controller can be managed using WebBIOS which can be accessed using Cntrl+H keys during BIOS bootup when LSI controller is scanned. (CSCtk06208)
Symptom The host based LSI tool MegaCli when executed will display a “Permission Denied” message on ESXi4.1 and ESX4.1 operating system.
Workaround After installing MegaCli on ESXi4.1 or ESX4.1, use the following command to change permission: chmod 755 MegaCli. Once the permission is changed, MegaCli command and options work successfully. (CSCtk11863)
KVM
Symptom User sees out of date (without KVM and VM tabs) KVM/vMedia client windows.
Workaround When running 1.3(1) software, if a KVM session is initiated, and the client does not have separate KVM and VM tabs, then it almost certainly indicates an older client is being run. The solution is to clear the java cache on the system. This is different than the browser cache. On Windows XP, follow these steps:
Step 1 Navigate to the Control Panel -> Java Control Panel.
Step 2 Click on the Settings button in the Temporary Internet Files panel.
Step 3 Click on the Delete Files button in the Temporary Files Settings popup box.
Step 4 Be sure to check the Applications and Applets checkbox and then click OK. (CSCtn41499)
Symptom User is not allowed to choose file name when creating vMedia image files.
Workaround The .img file created when using the vMedia “Create Image” feature is the same as the directory from which the contents are created with the “.img” extension added. Before creating the file, you can change the directory name to match what you would like the final file name to be. (CSCtk59152)
Known Behaviors
The following are the known behaviors found in the 1.4(2) release:
BIOS
Symptom BIOS boot order configuration from CIMC or UCSM may not be successful sometimes.
Workaround After BIOS boot order configuration, restart the system. Or wait for the BIOS boot to complete before configuring the boot order. (CSCts70906)
Symptom Installation of any UEFI Aware Operating System may result in a complete system failure.
Workaround No workaround exists. We do not support UEFI aware Operating Systems. (CSCtd71780)
CIMC
Symptom Network mode cannot be set to cisco_card mode even though the option is available.
Workaround Install Cisco UCS P81E card to move into cisco_card network mode. (CSCth87044)
Misc
Symptom Cannot install VMware on ICH10 SATA controller in SW RAID mode.
Symptom On completion of the "Recover Corrupt BIOS" wizard, clicking on the "Finish" button brings up a dialog with the message "BIOS recovery may not be in a state in which it can be canceled. Attempt to cancel anyway?"
Workaround The message is shown erroneously. Simply click "Yes" to close the message dialog. (CSCtd84141)
Symptom A name without domain part used during login in a scenario where authentication fails over from AD to CIMC does not succeed.
Workaround To connect to an AD server you need to use the full form of a user name (like bob@domain.com, even if domain.com had been specified in AD configuration). If this fails because the AD server is unreachable then login with the partial name. (CSCtd74258)
Symptom The following message appears after installation or upgrade to ESX 4.0 vSphere:
Error:TSC: 1137350434cpu0 :0)NUMA : 827 Signficant imbalance between NUMA nodes detected. Performance may be impacted.:Workaround There are two possible workarounds.
Option 1: Ensure that memory configuration is equal across both processors. Ensure that there are equal amounts of DIMMs of the same size and speed inserted in the DIMM sockets.
Option 2: Reboot system and enter BIOS setup by pressing F2. Go to Advanced-->Memory Configuration-->Memory RAS --> Performance Configuration --> and change NUMA Optimized to Disabled. Press F10 to save and reboot. (CSCtc33846)Symptom Windows 2008 R2 installation fails with the following message: “The Computer restarted unexpectedly or encountered an unexpected error. Windows installation cannot proceed. To install Windows, click OK to restart the computer, and then restart the installation”.
Workaround Remove the Intel® PRO/1000 PF Quad Port Server Adapter for a successful windows install. (CSCtt01897)
The following are the known behaviors found in the 1.4(1) release:
BIOS
Symptom BIOS POST will generate “PCI out of resource” errors when more than three PCI cards are added to a server, and all the PCI cards may not function correctly once the server boots to an Operating System.
Workaround Disable unused PCI card option in ROMs using BIOS setup or CIMC BIOS configuration page. (CSCtg82049)
Symptom Two CPU erroneously detected with different frequency after OS reboot. During boot the server may hang with the following message during BIOS post. Unrecoverable fatal error found. System will not boot until the error is resolved.
Workaround Reset the server. (CSCti49345)
Symptom On the C series servers with UCS P81E Virtual Interface Card, the system may hang during system boot (BIOS post).
Workaround This happens only when more than two vNICs are set to bootable in UCS P81E Virtual Interface Card. Do not set more than two vNICs to be bootable in UCS P81E Virtual Interface Card. (CSCti22874)
Symptom BIOS boot order may get corrupted when you try to set the boot order from CIMC.
Workaround You can recover the boot order by configuring the boot order from CIMC with “EFI Shell” as the boot device. “EFI Shell” is provided as a boot option mainly for diagnostic and debugging purposes. This is the last option by default. In case, if it is needed to set “EFI Shell” as first boot device, then BIOS setup should be used to set the boot order. (CSCtl44067)
Symptom When BIOS recovery process is initiated to recover a corrupt BIOS, this process will alter the boot order.
Workaround The boot order has to be reconfigured using BIOS setup after doing the BIOS recovery. (CSCto04406)
CIMC
Symptom After Factory default, if the network cable is connected to management port 2. IP restored does not ping.
Workaround We need to connect to management port 1, if we are trying to use factory defaults. CIMC will retain a DHCP IP on management port 1. So we will not loose connectivity to the machine. (CSCte93298)
Symptom Occasionally when configuring a C-series rack server for in-band management (connecting to the management controller via the host LAN on Motherboard ports, or via a Cisco adapter port), rebooting the host will cause an existing vKVM session to drop.
Workaround Use the dedicated CIMC ports for management connectivity. If that is not possible in your C-series server deployment, configure the switch ports to which the host is attached for fast spanning-tree recalculations. That can avoid the usual 30-second delay many switches impose when a port bounces before they re-run the spanning-tree calculation. (CSCth98858)
Symptom After resetting a C250 M1 to factory defaults, if the network cable is connected to management port 2, the restored IP does not ping other addresses. Factory default mode is not supposed to be used for normal operation. If ethernet management port 2 is used it gets a static IP used by Cisco even if DHCP is enabled.
Workaround Connect to management port 1 instead. (CSCte93298)
Symptom SSH and HTTP both can be disabled. If both SSH and HTTP are disabled at the same time then you will loose the complete connectivity to CIMC.
Workaround Do not disable SSH and HTTP at the same time. If disabled, to recover reset the factory default jumper on the board. (CSCtf86432)
Symptom CIMC becomes un-responsive when repeated WSMAN requests are sent.
Workaround Multiple WSMAN requests timeout as current CIM object manager servers only support one request at a time. Send WSMAN requests in a staggered manner so that they do not time out. (CSCth57587)
Symptom When downgrading from CIMC version 1.2 release to 1.1, CIMC is inaccessible.
Workaround Check its current network settings through the CIMC Configuration Utility or set the CIMC network settings to default using the CIMC Configuration Utility and use default network settings for version 1.1. (CSCti48255)
vKVM/vMedia
Symptom RHEL 4.8 does not detect the USB when mounted through virtual media.
Workaround The kernel needs to be rebuilt with max_luns option set to 255 in modprobe.conf. (CSCtd03150)
Symptom Launching the KVM console with Microsoft Internet Explorer browser fails with the message “Internet Explorer was not able to open this Internet site. The requested site is either unavailable or cannot be found”.
Workaround From the Internet Explorer toolbar, select Tools -> Internet Options. The Internet Options dialog box will pop-up. Click the Advanced tab and uncheck the “Do not save encrypted pages to disk” option. (CSCtd27392)
Symptom Unable to launch KVM. While launching the KVM, the error message “Login Failed” is thrown and KVM Closed Automatically. This can occur in the following scenarios:
- When you try to launch all the Maximum KVM sessions quickly (4 KVM session) and you can see that out of 4, 2 or 3 are failed with an error Message “Login Failed”.
- When launching single session of KVM, you might get an error message “Login Failed” and KVM is automatically closed.
Workaround When the KVM launch is failed with an error message “Login Failed” then wait for 2-3 mins and then again relauch the KVM. (CSCtg22414)
Symptom Sometimes blank screen is seen on KVM after FW update.
Workaround Clear Java cache from the control panel for Windows machine. (CSCtn49617)
Symptom The virtual mouse in the KVM console is difficult to use when installing SLES 11.
Workaround Use a physical mouse instead of a virtual mouse when installing SLES 11. (CSCta55758)
Symptom On C460, firmware update might fail with an error message “System update Package not found for the platform” or “Error string not found”.
Workaround Check if the SCU boot CD / ISO is mounted / mapped in vmedia.
Symptom KVM and vMedia does not work with mismatched IE and JRE executables.
Workaround Ensure that the matching JRE (and only the matching JRE) are installed. If a mismatched JRE is installed, it needs to be removed. (CSCtn42526)
CLI
Symptom In Power restore policy, a delay value is shown even if the delay type is random. In CLI, if the power policy is set to random, ideally delay value should not be shown.
Workaround Due to CLI limitation delay value is shown. This must be ignored. (CSCtk83452)
Misc
Symptom The CLP command takes a long time to complete and/or other applications appear slow.
Workaround The -l all option instructs CLP to traverse the entire sub tree. Due to the number of objects that must be queried, this command requires a long time to execute. To reduce the time and resources required, cd to the target(s) of interest and execute a show on the individual objects. (CSCtg72073)
Symptom System reboots when one or more Intel SSD 64GB drives are plugged in when the C-460 Server is up and running.
Workaround If Intel 64GB SSD drives need to be used in the C-460 server, power down the system completely and gracefully and then plug in the drives. The drives are recognised by the Storage controller after power-on. (CSCti94758)
Symptom Rx Errors or Rx FCS Errors reported by Ethernet vNICs.
- Processor C3 or C6 Enabled in the BIOS and high rx traffic load. In this case the PCIE root complex occasionally stalls up to 100 microseconds which may cause the ingress packet buffers to fill, resulting in a partial (truncated) frame which is reported as an FCS error.
- High traffic replication load to multiple vNICs assigned to the same uplink (For example, multiple vNICs in promiscuous mode). If receive traffic is replicated to multiple vNICs this may exceed the transfer ability of the adapter, resulting in the receive buffer filling. This may cause a truncated frame which will be reported as an FCS error.
Workaround Enable per-priority pause on active COSes on the switch. (CSCti17467)
LOM
Symptom Wake on LAN (WoL) Magic packet will not wake the system if they are sent to 10 GigE LOM SFP+ interfaces.
Workaround Use 1 GigE LOM ports for WoL. (CSCtg68906)
Symptom C460 10GBaseT LOM port can not hold the link when the host is down (on standby power) and it is connected to a switch which does not support 1G or lower speed.
Workaround C460 10GBaseT LOM does not support 10G speed on standby. If the BMC network mode is “Shared 10G” and the LOM port has to stay up on standby power, the port must be connected to a switch which supports 1G or lower speed and auto negotiation must be enabled. (CSCtl53810)
Web UI
Symptom While changing network mode to cisco card mode in IE, sometimes you might get an error as not able to parse the xml response.
Workaround Update IE browser version 8.0.6001.18702 and above or use other browser clients like Firefox, or Chrome. (CSCto29382)
Symptom The SEL Viewer agent in Windows 2003 might crash/stop when accessing SEL.
Workaround This issue is caused only when a COM based service try to access the WMI namespace. Assign the service with the current account (administrator) privilege. To set the privileges, follow these steps:
Step 1 Open services console by typing “services.msc” in Run.
Step 2 Select the service with the name “Cisco UCS SEL Viewer Agent”.
Step 4 Change “Log on as” from “Local System Account” to “This account”.
It is recommended to run SEL Viewer agent from an administrator account. (CSCtj18010)
Utilities
Symptom CIMC downgrade does not work. Firmware downgrade using the fwup utility older than ver 1.1.5 is causing the failure.
Workaround Use the fwup utility above ver 1.1.5. This version of utility is available only in the software releases above 1.4.x. (CSCtr91185)
Symptom When you try to run other applications which uses the same IPMI driver used by SEL Viewer, SEL Viewer stops.
Symptom All OEM events, severity, and other fields in SEL Viewer will be displayed as Unknown. For example, Microsoft Windows 2003 R2 or Windows 2006 with Microsoft IPMI driver installed logs OEM events not known by SEL Viewer utility.
Symptom Unable to access SEL Viewer GUI. SEL Viewer may report CIMC unavailable.
Workaround Verify whether firewall is enabled and the port on which SEL Viewer is running (default: 9099) is allowed through firewall. (CSCtl08079)
Symptom The SEL Viewer agent service stops in Windows when reloading SEL, while SEL is empty. This happens in the second attempt.
Workaround Start the service named “Cisco UCS SEL Viewer Agent” manually through service control panel. (CSCtl12945)
The following are the known behaviors found in the 1.3(2d) release:
CLI
Symptom There’s a message discrepancy between user interface and CLI when an invalid special character is entered.
Workaround Special characters should be entered with double quotes to be considered a valid string. (CSCto11022)
The following are the known behaviors found in the 1.3(1c) release:
CIMC
Symptom KVM console is blank after updating CIMC firmware from AP to BP.
Workaround
Option 1: From control panel clear the cache of java and try performing the reproduction step.
Option 2: Move CIMC to 1.4(x) release. (CSCtl93037)vKVM/vMedia
Symptom When you try to use virtual media to map both a .img file and a physical USB stick (pen/thumb drive), the following message (or a similar one) pops up: “Either Virtual media is detached or virtual media redirection for the selected virtual disk drive is already is in use by another user”.
Workaround The workaround is to only map one or the other drive at a time. (CSCtk54308)
Symptom When creating a vMedia image file, a popup with the following message is displayed, “Create Image Failed. An error occurred in the Image library”.
Workaround The image creation feature only supports image sizes up to 4 GB (32-bits) in size. Trying to create images from filesystems larger than that will result in an error. There is no direct workaround. Break data into smaller chunks, and represent each chunk as a separate image. (CSCtk61700)
Symptom When shutting down a vKVM session, you will notice a popup with the following message, “Closing this session will also close the locked virtual media session. Do you want to continue?”
Workaround None. Click the “Yes” button. (CSCtk61907)
The following are the known behaviors found in the 1.2(2d) release:
CIMC
Symptom With maximum hardware acceleration in some Windows OS installs, the vKVM cursor disappears.
Workaround If the slider is moved even one notch to the left of the maximum acceleration, then the remote cursor appears. (CSCtj30195)
The following are the known behaviors found in the 1.2(1a) release:
BIOS
Symptom After upgrading to BIOS version 1.2.1a.0, one of the two onboard LOMs may get disabled.
Workaround After upgrading to BIOS version 1.2.1a.0, reboot the server, go to the BIOS setup screen, navigate to the Advanced -> PCI Configuration screen, and set the “Onboard Gbit NIC2” to enabled. Then save and exit the BIOS setup. (CSCti71226)
CIMC
Symptom The CIMC Web GUI doesn't come up after certificate import. The browser displays the following Error message: Error Invalid Server Certificate. A request failed because the server's certificate was invalid.
Workaround Provide different values for the corresponding fields for “Subject Information” and “Issuer Information” during the certificate import process. (CSCth63837)
Symptom After using the WSMAN command the session displays the following error message: Error: Unable to create Resource Locator object.
Workaround There is no workaround. This request is not valid. You will need to provide a valid target so that CIMC can return appropriate information. (CSCth43327)
Symptom Occasionally IPMI LAN interfaces will become unresponsive (to the point of lost functionality) under stress.
Workaround The workaround is not to use IPMI queries during times of high SOL traffic. (CSCtd05874)
Symptom The uboot version shows extra character after normal CIMC version.
Workaround The extra character is displayed after updating to 1.2.x revisions of the firmware. This is a new feature added to the release and it specifies a specific uboot version. (CSCti49855)
Symptom Two active media sessions observed in the webUI/CLI.
Workaround Reset to default through the webUI or CLI so that the new changes take effect. (CSCti42465)
LSI
Symptom When the installation of the LSI SNMP agent on a Windows OS is complete, an error message that reads as “Fatal Error” may be displayed, indicating an unsupported version of firmware is installed on the storage controller.
Workaround The firmware version installed on the storage controller must be a supported version. To view the list of supported firmware versions for the controller, see: http://www.cisco.com/en/US/products/ps10477/prod_technical_reference_list.html (CSCti35162)
Symptom When there are 8708 and 1064E LSI controllers on the same system, the product name of the 1064E controller is displayed as “UNUSED” on the MIB walk.
Workaround Remove the 8708 LSI controller and place only the 1064E to view the product name in the MIB walk. (CSCti64538)
Open Caveats
This section lists the open caveats for the 1.4(2) release:
BIOS
Symptom When the BIOS parameters are configured from CIMC CLI or WebUI interfaces, the configuration might not take effect.
CIMC
Symptom A Cisco C200 or C210 rack server may have a "Disturbingly high number of I2C errors" message in the systems logs; when the installed OS is ESXi 4.1 or ESXi 5.0.
Workaround None known at this time. (CSCtt70347)
Symptom CIMC power capping is not supported on VMware ESXi 5.0.
Workaround When CIMC is upgraded to 1.4(2), the CIMC will automatically disable power capping. Power capping must manually be re-enabled to use it. (CSCtt08424)
LSI
Symptom The hard disk slot numbering on C260 in the LSI Host Applications such as MSM and BMC WebUI/CLI may not match with the Physical Slot numbers on the silkscreen in the front panel of the servers. The condition occurs in C260 configurations which have 1 or 2 LSI controllers with Non-Expander Backplane. If there is one LSI controller, there will be only one enclosure and the disk numbering would be 0-7. If there are two LSI controllers, there will be two enclosures and the disk slot numbering would be 0-7 on Enclosure 1 and 0-7 on Enclosure 2.
Workaround The hard disk slots 0-7 on Enclosure 1 from Host Application MSM corresponds to 1-8 Physical HDD Slot numbers and the Hard disk slot 0-7 on Enclosure 2 from Host Application MSM corresponds to 9-16 Physical HDD Slot numbers on the Front Panel Silkscreen. (CSCtq03202)
Caveats from Previous Releases
BIOS
Symptom BIOS setup parameters configuration interface from CIMC does not validate runtime parameters of the system and also it is not aware of the interdependency between them. For example,
- If the memory configuration does not support memory mirroring, then BIOS setup does not show mirroring as a setup option. However, CIMC would still show the option to set mirroring.
- When the system is equipped with a 4 core CPU, BIOS setup option to disable the cores would only show upto 4 cores. However CIMC would still show the option upto 6 cores.
- When SATA controller is disabled, BIOS setup will hide the options to set other SATA controller parameters. However CIMC would still show the other SATA controller parameters.
Workaround Even though the options are shown and can be configured from CIMC, it does not create any side-effects, nor do any harm to the system. BIOS will handle them gracefully. (CSCtl21208)
Symptom BIOS downgrade using the iFlash32 utility, from 1.4.x to the older version 1.2.x fails.
Workaround Use the startup.nsh script available in the 1.2.x container for the downgrade. This script will execute the BIOS downgrade successfully. (CSCtr93601)
CIMC
Symptom SNMPV3 traps are not receiving in Net-SNMP receiver. When V3 settings are chosen for trap delivery, the SNMP trapd CLI utility fails to receive V3 traps.
Workaround There is a possibility that the SNMP trap utility might not work for V3 authentication/privilege settings. (CSCtr83298)
Symptom The AES encryption field does not show the encryption level (128/256/ etc) on the Web UI and CLI.
Workaround The SNMP V3 encryption key length must be clearly indicated. The SNMP agent supports AES - 128 bit encryption by default. (CSCtr31577)
Symptom SNMP V1 test traps are sent even when V2 or V3 traps are enabled.
Workaround There is no provision for V2 and V3 test traps in the present implementation of the SNMP agent. (CSCtr37876)
Misc
Symptom Ports on the N2XX-ABPCI02 do not come up after a reboot or power cycle when running Linux.
LOM
Symptom Broadcom adapters and LOMs based on BCM5709, BCM57711 and BCM57712 configured for iSCSI and FCoE boot, iSCSI and FCoE boot configuration is lost after firmware upgrade. This is observed in the following cases:
- Using a Broadcom adapter based on BCM5709, BCM57711 or BCM57712.
- Having either iSCSI boot or FCoE boot configuration in the option ROM.
- Upgrading the firmware with HUU.
Workaround Record the configuration before the firmware upgrade and reapply it after the upgrade. (CSCtr85387)
CIMC
Symptom When you power on the chassis with some PS power cables disconnected, the system health LED on the front panel stays green, though some power supplies have no input voltage.
Workaround Connect all cables from APC power to the power supply securely. (CSCtg92856)
Symptom CIMC Web GUI and CLI correctly reports the number of failed DIMMs, but does not display which of the DIMMs have failed.
Workaround BIOS Setup (F2) or CIMC System Event Log can display the exact failed DIMMs. (CSCto09153)
Symptom On C460 servers, the CIMC Web GUI does not display Cisco P81E card details when the main CPU is powered off.
Workaround Use the CIMC CLI or power on the main CPU. (CSCtn75815)
Note Release 1.3(1) does not support UCS integration. Customers who want to integrate their C-Series server with a UCS instance must use release 1.2(x). If your server is currently running 1.3(1), you need to downgrade it using the 1.2.2 Host Upgrade Utility. For more information, see the Hardware Installation Guide for the type of server you are integrating.
CIMC
Symptom CIMC storage view of LSI 9260 card occasionally reports 0 Cache Memory size.
Workaround There is no workaround for CIMC storage management views. You can use host-based MSM tools to corroborate data. (CSCtn08982)
Symptom When NIC mode is in shared_lom and the boot order is such that EFI is not the first one in the order and there is no other media to boot from, the following message is displayed - “Select proper media else press a key”. If you enter a key, the KVM session terminates.
Workaround Ensure that you have the proper media in place before trying to boot from the device. If not, when prompted with above message, you will have to place the required media and then reboot the host. Else have the system in dedicated mode. (CSCtl82492)
Symptom PI2CPerformOP ioctl driver failures occur in CIMC log.
Workaround This failure message is seen when CIMC couldn't communicate to certain devices, due to the devices being missing or if there is a momentary communication failure. If the messages are constant, this may indicate a problem with the server. There is no workaround for this issue. (CSCtn55177)
MISC
Symptom SMASH memory enumeration can fail on UCS-C460-M1.
Workaround None for SMASH interfaces. Memory related data can be retrieved through CiIC CLI or WebUI. (CSCtn29771)
vKVM/vMedia
Symptom KVM fails to launch with a pop-up “Unable to launch Application”.
Workaround Follow these steps:
Step 1 Update the JRE to http://www.java.com/en/download/inc/windows_upgrade_ie.jsp
Step 3 Clear the cache (Internet Options -> Tools).
Open a new browser window and start a CIMC session. KVM should now be able to launch. (CSCtl70994)
Symptom Garbled WebUI output when LSI 9280-4i4e card is installed and external JBOD connected.
Workaround There is no actual data integrity problem, but the displays can be significantly garbled. One workaround is to only use external or only internal on any given controller card. Another is to use host based tools to display information on controllers where both internal and external storage is connected. (CSCtn03816)
CIMC
Symptom C460 CIMC network mode shared_lom_10g with active-standby redundancy can stop working when the host OS is also using 10GE LOM port and then is rebooted.
Workaround Do not use the CIMC network mode shared_lom_10g. Use one of the other modes instead. (CSCtj57061)
Symptom C460 CIMC network mode shared_lom_10g with active-active redundancy will lose connectivity temporarily when the host OS is also using a 10GE LOM port.
Workaround Do not use CIMC network mode shared_lom_10g. Use one of the other modes instead. (CSCtj58245)
Symptom LED sensor color is displayed as Red or Amber or Blue (or any supported color) even though LED state if OFF.
Workaround Ignore LED color when LED state is OFF. (CSCth84883)
Symptom If you try to move to cisco_card mode from the CIMC Configuration Utility, you see the error “Net mode invalid”.
Workaround Install either just one or two R2X0-PSU2-650-SB power supply units in the server. Restart the CIMC and move to cisco_card mode. (CSCti70359)
CIMC
Symptom When updating CIMC firmware through TFTP, if the image file is corrupted, the update status indicator is the same as if the file does not exist.
Workaround Be aware that this error message can actually indicate either of the above conditions and should make sure that the file both the exists, and is a valid firmware image for the CIMC being upgraded. (CSCti17492)
BIOS
Symptom If the current CIMC networking mode is shipping mode, then the BIOS F8 CIMC configuration utility does not allow a new networking mode and IP address to be set at the same time.
Workaround Set the new networking mode, save, then set the new IP address and save again. (CSCth71350)
Release 1.1.(2)
This section lists the open caveats for the 1.1(2a) release and applies to the M1 version of the C460 server.
MISC
Symptom If the number of Virtual Drives created in the LSI MegaRAID controller is greater than or equal to 50, the system will not boot from any of these Virtual Drives.
Workaround None. The system boots from MegaRAID Virtual Drives only if the number of Virtual Drives are lesser than or equal to 49. (CSCtg25373)
This section lists the known behaviors for the 1.1(2a) release and applies to the M1 version of the C460 server.
BIOS
Symptom Serial port B cannot be enabled for console redirection in the Server Management —> Console Redirection page of the BIOS setup.
Workaround Serial port B is primarily used for SOL functionality. The BIOS will start redirecting console messages to serial port B if SOL is enabled. You should enable SOL through BMC to get console redirection messages through serial port B. (CSCtf54851)
Symptom System reboots during EFI Windows 2008 R2 installation.
Workaround EFI OS installation is not supported by the BIOS. You should disable the EFI- Optimized mode in the BIOS setup and install Windows 2008 R2 in legacy mode. (CSCtf87728)
Symptom On completion of the "Recover Corrupt BIOS" wizard, clicking on the "Finish" button brings up a dialog with the message "BIOS recovery may not be in a state in which it can be canceled. Attempt to cancel anyway?"
Workaround The message is shown erroneously. Simply click "Yes" to close the message dialog. (CSCtd84141)
Symptom A name without domain part used during login in a scenario where authentication fails over from AD to CIMC does not succeed.
Workaround To connect to an AD server you need to use the full form of a user name (like bob@domain.com, even if domain.com had been specified in AD configuration). If this fails because the AD server is unreachable then login with the partial name. (CSCtd74258)
CIMC
Symptom The CIMC time/date is set to 1970.
Workaround Do not reboot the CIMC after the BIOS sets the time/date for the CIMC. (CSCtc21237)
Release 1.0(1)
The following caveats were opened in UCS software release 1.0(1) and are still unresolved.
This section lists the open caveats for this release. Unless otherwise noted, the caveats are for all rack server platforms.
KVM
Symptom Launching the KVM console with Microsoft Internet Explorer browsers fails and causes the message “Internet Explorer was not able to open this Internet site. The requested site is either unavailable or cannot be found”. This error occurs with Microsoft Internet Explorer browsers with the advanced “Do not save encrypted pages to disk” option set.
Workaround From the Internet Explorer toolbar, select Tools -> Internet Options . The Internet Options dialog will pop up. Click the Advanced tab and uncheck the “Do not save encrypted pages to disk” option. ( CSCtd19439 )
Documentation Updates
Active Directory
In the section titled “Active Directory” in the Cisco UCS C-Series Servers Integrated Management Controller CLI Configuration Guide, Release 1.2(1), and the Cisco UCS C-Series Servers Integrated Management Controller GUI Configuration Guide, Release 1.2(1), the following statement is incorrect:
“When Active Directory is enabled in the CIMC, all user authentication and role authorization is performed by Active Directory, and the CIMC ignores the local database. If the CIMC cannot connect to Active Directory, it reverts to the local database.”
The correct behavior is as follows:
“When Active Directory is enabled, user authentication and role authorization are performed by Active Directory for user accounts not found in the local user database.”
Symptom Linux mouse emulation is calculated, unlike Windows mouse emulation. There are situations when the KVM Linux mouse emulation gets out of sync with the host's mouse: Novell/SUSE's install DVD for SLES doesn't include mouse acceleration. The X setup for the install DVD is very basic, unlike RHEL's. A mouse will work on SLES once it is installed but lags during the installation.On Linux, you can switch between graphical and text mod using the Ctrl-Alt-F1 for text and Ctrl-Alt-F7 to go back to graphical mode. The KVM Linux mouse emulation sometimes gets out of sync when switching between modes.
Workaround When the mouse gets out of sync in Linux, the user must force KVM to recalculate it. There are two ways to force KVM to recalculate the Linux mouse: 1) In KVM, go to the Tools->Session Options->Mouse tab and press the button for Linux Mouse Acceleration and then press "Apply" and "OK". If the Linux Mouse Acceleration is already selected, the user can still press "Apply" and then "OK" to force it to recalculate. 2) Move the mouse out of the KVM window to the right of the window and back into the KVM window, forcing it to recalculate the mouse position again. (CSCtc12265)
Related Documentation
For configuration information for this release, please refer to the following:
- Cisco UCS C-Series Servers Integrated Management Controller CLI Configuration Guide, Release 1.1(1)
- Cisco UCS C-Series Servers Integrated Management Controller Configuration Guide, Release 1.1(1)
- Cisco UCS C-Series Servers Integrated Management Controller CLI Command Reference, Release 1.1(1)
The following related documentation is available for the Cisco Unified Computing System:
- Cisco UCS C-Series Servers Documentation Roadmap
- Cisco UCS Site Preparation Guide
- Regulatory Compliance and Safety Information for Cisco UCS
Refer to the release notes for Cisco UCS Manager software and the Cisco UCS C Series Server Integration with Cisco UCS Manager Guide at the following locations:
Obtaining Documentation and Submitting a Service Request
For information on obtaining documentation, using the Cisco Bug Search Tool (BST), submitting a service request, and gathering additional information, see What’s New in Cisco Product Documentation at: http://www.cisco.com/en/US/docs/general/whatsnew/whatsnew.html .
Subscribe to What’s New in Cisco Product Documentation , which lists all new and revised Cisco technical documentation, as an RSS feed and deliver content directly to your desktop using a reader application. The RSS feeds are a free service.
This document is to be used in conjunction with the documents listed in the “Related Documentation” section.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: www.cisco.com/go/trademarks . 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. (1110R)
FeedbackContact Cisco
- Open a Support Case
- (Requires a Cisco Service Contract)