'show physical-server' Displays Missing System ID for a Server and Oracle Fabric Manager Does not Find Server

Asset ID:	1-72-1578890.1
Update Date:	2016-10-13
Keywords:

Solution Type Problem Resolution Sure

Solution 1578890.1 : 'show physical-server' Displays Missing System ID for a Server and Oracle Fabric Manager Does not Find Server

Applies to:

Oracle Fabric Interconnect F1-15 - Version All Versions and later
Oracle Fabric Manager - Version 3.3.0 to 4.0.2 [Release 3.0 to 4.0]
Oracle Fabric Interconnect F1-4 - Version All Versions and later
Information in this document applies to any platform.

Symptoms

When running the 'show physical-server' command from the Fabric Interconnect CLI, some servers may display 'missing System ID" message and show up in red. Also the server never shows up in Oracle Fabric Manager even though the Fabric Interconnect CLI reports the server, and can install server-profile, vnics and vhbas.

Cause

This has been reported in some HP systems, but there may be other vendors this applies to as well. It was found that the system needed to have its Serial Number input into the system BIOS to clear the issue. Previously there was a workaround to perform from ESX/ESXi server CLI that would termporarily fix this issue, but it wasn't permanent through host reboots. Therefore the workaround had to be applied every time the server was rebooted.

Solution

Somewhere on the outside of the server there should be the Serial Number of the server. Write this down and save it. Next boot the system and enter the system BIOS. The screenshot below is taken from an HP system, other vendors may locate the system S/N in another location in the system BIOS. The screenshot is supplied as an example only and my not be valid in every instance:

HP BIOS S/N Screenshot

Save and Exit the BIOS. Upon host reboot, the Fabric interconnect 'show physical-server' entry no longer shows the server in red (instead of black) and the error message is gone. Also Oracle Fabric Manager is now able to see the server, negating the need for this previous workaround from ESX/ESXi (no other workarounds existed for any other Operating System):

1) cp /usr/sbin/xsigo-hotplug /tmp
edit /tmp/xsigo-hotplug and change esxcfg-info -u to esxcfg-info -b
2) cp /tmp/xsigo-hotplug /usr/sbin
3) watchdog.sh -k xsigo-hotplug
4) watchdog.sh -s xsigo-hotplug /usr/sbin/xsigo-hotplug &

NOTE: The above workaround is not persistent through host reboots.

Attachments

This solution has no attachment