ASR's with the HP Agents for VMware ESX Server 3.x 7.7.0 and the Emulex HBA API Kit 7.3.2

[Dana Swanson]

Hi All,

Is anyone out there running the HP Agents for VMware ESX Server 3.x 7.7.0 and the Emulex HBA API Kit 7.3.2?

We have a few ESX 3.0.1 servers becoming unresponsive and a ASR is kicking in. Two servers so far. One with some ECC memory errors on one DIMM but the other has no reported hardware issues.

We haven't had any issue with ESX 3.x server reboots in the past.

I un-installed the Emulex HBA API Kit and will see how this goes.

Thanks!

Dana

[Rob Buxton]

I've only ever installed the basic LPUTIL Kit, not the full one.
There is an issue with QLogic HBA's when an HBA Rescan is done that can cause a hang but not an ASR.
An ASR should leave a trace in the IML - enter cpqimlview from a ssh session to view it.

[Dana Swanson]

Thanks Rob,

I did check the IML on the System Management Homepage when I first noticed this but the only entry is the ASR.

I just run cpqimlview to see if any more information was available on the Linux side of ESX but it displays the same info as the IML on the System Management Homepage.

Thanks for the tip!

The annoying thing is I ran the un-install script for the Emulex Kit and after a reboot I can still see the presented storage. I could never see the storage on the hba's until the Emulex kit was installed. It must be running still.

I'll have to read up on the Emulex site.

Thanks,

Dana

[Dana Swanson]

I just did some research into the Emulex HBA API Kit 7.3.2.

I'm not running the version of /opt/Compaq/storage/bin/libdfc.so that came with the Emulex HBA API Kit, I'm running the one that comes with the HP Agents for VMware ESX Server 3.x 7.7.0.

I didn't stop the insight agents before running the Emulex ./Install.sh. Because of this the libdfc.so stayed at it's current rev which is a good thing since the version available on the Emulex Kit was older.

When I initially installed the HP Agents for VMware ESX Server 3.x 7.7.0 I could not see any storage via the Emulex cards and this was something we wanted enabled.

What must have happened is when I setup the management agents there was no storage presented to the Emulex card. When I looked up the Emulex Kit info someone must have enabled the storage and I thought installing the Emulex HBA API Kit discovered the storage when it was just enabled on the HBA.

I just un-installed the Emulex Kit from the ESX servers. The HP Agents for VMware ESX can see the presented storage fine.

So it's possible the Agents for VMware ESX Server 3.x 7.7.0 may be the issue.

Thanks,

Dana

[Dana Swanson]

I think I figured out the issue. There's a more recent version of the Firmware for the HP ProLiant DL585 G1 that fixes a ASR issue caused when a ECC memory condition is detected.

One of our servers is having ECC's and one is not. Both have ASRed. Perhaps the server without any ECC errors in the IML is having ECC events and can't log the event before the system becomes unstable.

Will let you know how this works out.

Dana

[Rob Buxton]

ESX should see SAN storage without the Emulex HBA API Kit.
I've only installed the LPUTIL kit as it makes updating HBA firmware easier.

ESX has all of the HBA drivers it needs for supported hardware.

[Dana Swanson]

Hello,

I'm guessing what happened was this:

The ESX admin had said it would be great to see the HBA storage on a test server I had installed the hp agents on. He assumed the storage had been assigned to the HBA.

There must have been no storage assigned to the HBA making me think there was an issue. I installed the Emulex kit and there was the HBA storage.

Someone must have assigned the storage before I installed the Emulex kit, causing me to think this fixed it.

The HP agents work fine without the Emulex kit. It was one of those timing things with multiple people configuring the test server. I should have asked if there was storage assigned first but live and learn.

See ya!

[Dana Swanson]

I'm in the process of upgrading the ESX Servers running on ProLiant DL585 G1 Systems to ROMPaq Firmware Upgrade Version 2007.02.14 A.

I'll post if this doesn't fix the problem of server ASR's. It looks like the solution.

Thanks,

Dana

[Dana Swanson]

I'm in the process of upgrading the ESX Servers running on ProLiant DL585 G1 Systems to ROMPaq Firmware Upgrade Version 2007.02.14 A.

I'll post if this doesn't fix the problem of server ASR's. It looks like the solution.