Problem using VMware Update Manager

[rtmanager]

I have new installed esx 3.5 hosts together with a VCenter Server 2.5.

When I try to install the latest patches via VMware Update Manager, the install worked good up for the first 7 patches.

Between installing the 7. and the 8. Patch I will loose the connection to the esx host.

The only way for me to come out was to reboot the host and remediate again the patches.

Now VUM is installing correct the last 10 patches and (after a reboot) set the esx host to status green toward the baseline.

The error message in VCenter was: An error occured while communicating with the remote host.

Can anyone help me about this strange behaviour?

[tharpy]

We're having a similar problem last night....three new (identical) HP hosts installed 3.5 update 1 (build 64607a). Using VC 2.5 and UM to patch after initial build, host #3 remediated fine and is now "compliant". Host # 1 went into maintenance mode and appeared to install several patches before hanging at 85%. Had to reboot the host before reconnecting to VCMS; UM shows 17 patches not compliant....then removed the host from the cluster, attached the baselines to the individual host and remediated again. A few more patches installed, then the remediation stalled again....now showing only 7 non-critical patches not compliant. Rebooted the machine and re-inserted into the cluster. Ran the remediation on the Cluster again and Host #2 stopped at 81% with an error message on VCMS "Unable to access the specified host". In both cases where UM appeared to have stalled, we could ping the server by name and ip....

[Schorschi]

Suggest you open SRs with VMware and make sure VMware can find root cause for this, or both issues, if different. This is not good. We decided not to use UM for other reasons, having to do with features we need, but are not yet in UM. Appears our caution not adopting UM initially was justified.

[mike_laspina]

Hi,

On the Tasks & Events tab you should be able to see what patch it stopped on in the Events log.

The events info should be helpful in identifying what patch had the issue.

It would be helpful to the community if you can post it.

Maybe it's something that can be reordered to correct the patch install stability problem.

[ExCon]

I'm on my 19th host right now, and I've seen this behavior on every single one. After ESX350-200804403-BG, the host disconnects. After a reboot, the host does the Pegusus install queue thing for 1, 2, and 3 (failing twice on 3) and then reconnects in the VC. I remediate again, and all finishes up okay. On the final reboot, Pegusus queues 1 and 2 do not show up, and queue 3 still fails twice. As far as I can see though, everything is okay from that point on.

[mike_laspina]

That pegasus queues indicate this issue, but I don't know how that would impact UM.

http://communities.vmware.com/message/914939#914939

[ExCon]

Yeah, I wasn't meaning to imply that UM was impacted by Pegusus. Rather I was reporting what happened to me after updating via UM. The failure of one of the patches and subsequent disconnect requiring a reboot was really the meat of my post. Good info in the post you referenced, though!

[dmadden]

This is a known issue with the hostd patch (don't remember the number off the top of my head).

Here is the KB for it:

http://kb.vmware.com/kb/1004966

Apparently this only occurs with Update Manager. VMware is working on a fix for the problem.