C H A P T E R 4 |
Software Issues |
This chapter describes software issues related to the Sun Blade X6450 server module.
After performing a software upgrade, there is a delay of about three minutes before the server can be powered on.
Wait three minutes or more before powering on a server after a software upgrade.
If you stop and start the host more than once, PCcheck does not run after the first time, and an error message appears in dmesg:
Retry several times until it runs.
On systems with a CF card, SUSE 10/9 reports errors.
These messages can be ignored. They do not affect system performance.
While upgrading to SW3.0, it can take about five minutes to create the SSH key. The screen displays:
SSH RSA Host Key files don't exist. Creating... Starting sshd
While this might take awhile, it should still work. No action is necessary.
The Solaris x86 X-Server might fail to start after it is enabled with the dtconfig -e command.
Add the following option to the xorg.conf file:
When using Receive Side Scaling (RSS) in Windows Server 2003 SP2, the interface fails, but shows no indication of failure other than a blank IP address, 0 packets received from the adapter, and a blank ARP table.
Disable and re-enablethe interface to restore normal operation.
Install the Installpack from the SW 2.0.3 or later ISO image. For example:
/windows/w2k3/packages/InstallPack_2_0_3.exe
Note - SW2.0.3 is released as a downloadable ISO image only. It will not be released on physical media (DVD or CD). |
For more details, please refer to the following Intel doc.
Intel® 82563EB/82564EB Gigabit Ethernet PHY- Customer letter for ESB2 / Gilgal issues - Customer Communications - 07-Nov-2008.
A server module that is fully populated with DIMMs can take up to three additional minutes to boot, due to the time required to run POST tests on the DIMMs.
VMware ESX/ESXi releases provide a health status and monitoring feature. Incompatibility between the ILOM and VMware IPMI interfaces causes ILOM IPMI sensors to be partially displayed in the VMware health status window. Sensors might also appear and disappear, and updates occur slowly.
Use the ILOM GUI, CLI, SNMP, or IPMI interfaces to monitor server health. Also, the ILOM provides remote syslog and remote alert interfaces to enable remote monitoring.
The VMware ESX 3.5 process occasionally uses 100% of a core.
# cat /proc/vmware/interrupts Vector PCPU 0 PCPU 1 PCPU 2 PCPU 3 PCPU 4 PCPU 5 PCPU 6 PCPU 7 0x21: 7442833 0 0 0 0 0 0 0 VMK ACPI Interrupt |
1. Edit the file /etc/vmware/esx.conf and add the line /vmkernel/mpsIntRouting = "TRUE".
The arrow keys do not work when using an RKVM on a SunRay.
Use alternative keystrokes, or remap the arrow keystrokes in the SunRay X session.
During RKVM sessions with RHEL4 and ESX 3.5 u2, the Alt key appears stuck.
If you are installing the Windows OS with the QLogic Fibre Channel Adapter, do not use the drivers on the Windows DVD.
When the installation wizard informs you "The driver you provided seems to be newer than the Windows default driver.", select "S" to use the driver on the DVD.
These are not actual floating point errors and the unexpected exit can cause data corruption.
Two patches to the Solaris OS must be applied, in sequence: patch 125370-06, followed by 127112-10.
The Pc-Check test User Pattern Test does not finish.
There is no functional impact. If this problem is seen, simply exit the test.
The message Panic on CPU0: Not enough RAM for domain 0 allocation is displayed. This panic is caused by certain allocations taking up too much memory in the lower memory of the kernel.
To fix the problem, update the OS with Red Hat Security Advisory RHSA-2007:0099, or later, which contains a fix for this issue.
https://rhn.redhat.com/errata/RHSA-2007-0099.html
If you are unable to update the kernel successfully, there is a workaround available. Boot with the system with the kernel parameters dom0_mem=512M mem=64G. For example, in /boot/grub/grub.conf file, enter the parameter like this:
title Red Hat Enterprise Linux Server (2.6.18-8.el5xen) root (hd0,0) kernel /xen.gz-2.6.18-4.el5 dom0_mem=512M mem=64G module /vmlinuz-2.6.18-4.el5xen ro root=LABEL=/ module /initrd-2.6.18-4.el5xen.img
When using ESX 3.5 there is an interrupt flooding at vector 0x21 which results in more than 90% usage of one core.
This issue has been fixed with BIOS version 3A08 of SW release 1.0.1.
The Solaris OS cannot be installed on a RAID array larger than one terabyte.
The Solaris installer OS doesn’t support it.
There is no workaround for this issue.
Servers running the Solaris 10 5/08 OS may experience reduced network performance.
Systems with the Intel® Xeon® Processor MP7400 series running the Solaris 10 10/08 OS might experience slightly reduced performance or increased power consumption under light utilization due to an issue where the CPUs might not idle fully.
No error message is displayed.
Add the following line to /etc/system file:
When using RKVM, the Ctrl+Alt+Delete hot key does not function for the SLES10 SP1 OS.
Use SP to reset your system, or reboot it using OS commands.
Copyright © 2010, Oracle and/or its affiliates. All rights reserved.