|C H A P T E R 6|
This chapter contains issues that have been fixed for the Sun Blade X6220 server module. This chapter covers the following topics:
In order to create a RAID volume by migrating data from the primary disk to the secondary disk, you must preserve at least 1.6GB of disk space as metadata space for LSI RAID.
This issue was fixed with LSI BIOS Phase 10 release.
After running hundreds of sync_netperf TCP/UDP connections, the nge interface is not available.
To avoid this issue, perform the following steps:
1. Open the /etc/system file using a text editor.
2. Add the following line to the end of the system file.
3. Reboot the system.
This issue was fixed with Solaris 10 8/07.
When using the raidctl utility with a Solaris 10 11/06 operating system,
RAID-1E volumes might be reported as volume type IS instead of IME.
This issue was fixed with Solaris 10 8/07.
The Sun Blade X6220 Server Module Service Manual, 820-0046-10, currently states:
The hard disk drives have three LEDs. The order listed below is when the server module is installed in the chassis:
The correct behavior for all HDD activity LEDs is as follows:
This section, External Status LEDs, of the Sun Blade X6220 Server Module Service Manual, 820-0046 has been corrected.
This section covers the following topics:
When an uncorrected ECC error occurs on an 8-DIMM CPU module, the error might be reported on the lowest-numbered DIMM pair (the pair closest to the CPU) regardless of where the error actually occurred. For example, in a fully-populated module, an error in Pair 3 (DIMM7 and DIMM6), Pair 2 (DIMM5 and DIMM4), or Pair 1 (DIMM3 and DIMM2) may be reported as occurring in Pair 0 (DIMM1 and DIMM0).
This problem can cause the system to reboot repeatedly, with no output on the video display. If this occurs, check the ILOM System Event Log for frequent ECC errors in a lowest-numbered pair.
Solaris AMD x64 support includes a boot-time check for the presence of a BIOS workaround for the AMD Opteron Erratum 131. If Solaris detects that the workaround for Erratum 131 is needed but it is not yet implemented, Solaris logs and displays the following warning message:
The Sun Blade X6220 server module BIOS implements a superset workaround that includes the workaround required for Erratum 131. This warning message can be safely ignored.
|Note - This issue was resolved in Solaris 10 5/08.|
If you use the stop /CH command to shut down the chassis CMM, the server modules will shut down immediately.
Connect to each server module LOM and to gracefully shut it down. Then shut down the chassis.
When you power down a server module in preparation for removal from the chassis, the blue ready-to-remove light is illuminated. However, It is possible that a firmware upgrade could be in process, even after the blue light goes on.
Check with your system administrator before removing the server module to make sure there is no firmware upgrade in process.
If you reset the service processor without rebooting the operating system, you are unable to open a system console with the command start /SP/console.
When the system boots up, all four disk fault LEDs are on amber, even though the system OS is booting and operating normally.
The following issue is fixed in the phase 10 release of the LSI firmware.
The clients/syslog destination ip1 or ip2 cannot be set with a standalone zero (0) in the IP address. For example, 10.10.10.10 and 100.100.100.100 works. 184.108.40.206 does not work.
There is no workaround.
Some ntp IP address will be considered invalid when set in the CLI.
Use the WebGUI to set ntp IP addresses.
A CLI session might lose data if a new image upload fails due to a timeout tftp during download.
Close the current CLI session and open a new session.
The ILOM CLI might have issues setting multiple properties simultaneously.
You must set properties individually. For example, to set ipaddress to 10.8.168.64:
The ILOM service processor only reports information about thresholds that are used or controlled by the service processor. Therefore, fan and power supply threshold values cannot be read by the CMM and blades.
This fix is planned for a future release.
When a Raid 1 system with a hotspare is created, the system might fail to boot the Solaris OS and will keep attempting to reboot after the primary disk on slot 0 is removed.
Leave the primary disk in slot 1, then move the secondary disk from slot 2 to slot 0.
The following issue is fixed with the software 1.0.2 update. In addition, all issues that are listed in Issues fixed in Software Release 1.0.1 will also be resolved if you update from software release 1.0 to software release 1.0.2.
To avoid system reboots due to memory errors for server modules running Linux or Windows operating systems, you will need to disable the AMD PowerNow feature. This issue affects BIOS 0ABJT025 and is fixed in BIOS 0ABJT026 included in software release 1.0.2.
There are two options that you can use to disable PowerNow: Either disable PowerNow in the operating system or disable PowerNow in the BIOS.
To disable PowerNow in the BIOS:
1. Boot or reboot the server.
2. Press F2 to enter the BIOS Setup Utility.
3. Click on the Advanced menu.
4. Click on the CPU Configuration option.
5. Change the ACPI 2.0 Objects setting to Disabled.
6. Press F10 to save the changes and exit.
This chapter describes the issues that have been fixed in the ILOM 1.1.8 version of the Sun Blade X6220 server module software 1.0.1 release.
The CMM front panel power LED does not power on as soon as 3.3V_AUX is available because the ILOM service processor could take up to two minutes to boot after a blade is inserted.
The service processor must boot before LEDs are switched on. Please wait until ILOM has booted before checking front panel power LEDs.
The ILOM service processor might fail during the BIOS flash upgrade. If this should occur, power on the server, then retry the ILOM flash upgrade.
The service processor hangs and the JavaConsole cannot be loaded when a browser is used to start the service processor WebGUI. When the using the following selections: Remote Control-> Redirection, press Launch Redirection, the following screen prompt appears before the service processor hangs:
To avoid this issue, use one of the following workarounds:
This fix is planned for a future release.
The CLI WebGUI does not show the fan fault state. This fix is planned for a future release.
You can inspect the fan fault state through /CH/FMx/FAIL.
The ILOM service processor cannot set the LEDs on when using the WebGUI. When in the WebGUI, and the /sys/ok2rm and /sys/service sensors are turned off, there is no way to turn them on. When these two sensors are turned on using CLI, and not using the WebGUI, the sensors can be turned off.
To control the LEDs, use the IPMI or ILOM CLI to turn /sys/ok2rm and sys/service indicators on and off.
The /CH/BLn/SP/cli command is not usable when the chassis power is off. This fix is planned for a future release.
To avoid this issue, turn the chassis power on.
The ILOM WebGUI displays power supply unit (PSU) VINOK and PWROK sensor information as fault LEDs.
You can safely ignore this PSU information.
When the IPMI command is used to perform a warm reset of the BMC (Baseboard Management Controller), the CLI user might get logged out, or the following error message might be displayed:
Do not use the IPMI protocol command to warm reset the BMC (Baseboard Management Controller).
The single sign-on feature for the JavaConsole is not working.
To avoid this issue, you can log in to JavaRConsole using the same login and password as the web browser.
The following message might display when running a Linux operating system and starting the blade CLI.
You can safely ignore this message.
The SEL list periodically reports critical IPMI events about the system devices even though the system and SDR log reports that the system is okay.
For example, you might see the following messages about devices and sensors (bl8.prsnt, bl6.prsnt, ps0.vinok0, ps1.prsnt, and ps1.vinok).
43 Fri Mar 9 17:37:51 2007 IPMI Log critical ID = 29 : 03/09/2007 : 17:37:51 : Entity Presence : bl8.prsnt : Device Absent33 Fri Mar 9 15:57:52 2007 IPMI Log critical ID = 1f : 03/09/2007 : 15:57:52 : Entity Presence : bl6.prsnt : Device Absent35 Fri Mar 9 16:38:01 2007 IPMI Log critical ID = 21 : 03/09/2007 : 16:38:01 : Power Supply : ps1.vinok1: State Asserted38 Fri Mar 9 16:57:37 2007 IPMI Log critical ID = 24 : 03/09/2007 : 16:57:37 : Entity Presence : ps1.prsnt : Device Absent
You can safely ignore these messages.
When using ILOM 1.1 or higher and you attempt to flash upgrade the service processor, you might receive a partial flash failure.
Reset the service processor and reattempt the flash upgrade. You should ensure that you are using the correct pathname.