fontyyy
Contributor
Contributor

Virtual Center error

Jump to solution

Hi there,

I had a slight problem this morning, I noticed that one of our esXpress helpers said it was still running a backup of one of our servers. This server (in the Virtual Centre) said it had a snapshot.

Further investigation (on the FTP server and logging directly into the ESX box itself rather than to the VC machine with the Infrastructure client) revealed this was not true, the esXpress backup had completed and the snapshot had been deleted. so i'm guessing here there was some kind of problem with the VC database (held on a 2k SQL server).

I "fixed" this by removing the cluster and rejoining all the hosts to a "new" cluster, making all the resource pools again etc. which isn't a big deal to me as we only have 4 ESX boxes, 30odd VM's and only one pool actually has resources set, the rest are just to keep things tidy. Howerver I could imagine for some of you this could be a bit of a disaster.

On a side note I've seen esXpress hang part way through a backup, refuse to delete the snapshot and had to migrate everything else off the host and restart the host (thus downing the VM for 10 mins or so) to get anything back. This has only ever occured with VM's running database systems, both MS SQL and Postgre.

Anyone seen anything similar? Anyone got any better solutions?

0 Kudos
1 Solution

Accepted Solutions
kharbin
Commander
Commander

Hello,

Couple of things I see already. First you state, "one of our esXpress helpers said it was still running a backup of one of our servers. This server (in the Virtual Centre) said it had a snapshot." This is a result of a communication problem between Vitual Center and the ESX host. esXpress does not talk directly to VC. In fact this problem is addressed on our web forum at with the solution being simply log in to the VC host and restart the Virtual Center service. You should have done this first, then you would not have had to remove the cluser, rejoin, etc., etc.

You also state, "I've seen esXpress hang part way through a backup, refuse to delete thesnapshot and had to migrate everything else off the host and restartthe host". We see this happen when the ESX host iSCSI or SAN connection is interupted. We also see this with W2K3 VMs running MS/SQL. This is not exclusive to esXpress, simply adding a snap shot to a W2K3 VM with SQL can casue it to freeze. You can easily verify if this VM has the problem using the VI3 client.

Another thing I see is you are using v3.0.7 of esXpress released in March 2007. While this version is functionally stable, our latest version has vastly improved the logging, reporting and error detection. I would suggest going to your personal download page to retrieve the 3.0.22 release.

And last, I see no record of your contacting our support staff. Our staff would have been able to quickly resolve your problem with no interption to your operations, no rebooting hosts, etc. Please contact our support before resorting to a reboot.

If you would like, you can run the 'phd-support' program and our staff can review the findings/output with you.

Thankyou

Ken Harbin

www.esXpress.com

View solution in original post

0 Kudos
4 Replies
kharbin
Commander
Commander

Hello,

Couple of things I see already. First you state, "one of our esXpress helpers said it was still running a backup of one of our servers. This server (in the Virtual Centre) said it had a snapshot." This is a result of a communication problem between Vitual Center and the ESX host. esXpress does not talk directly to VC. In fact this problem is addressed on our web forum at with the solution being simply log in to the VC host and restart the Virtual Center service. You should have done this first, then you would not have had to remove the cluser, rejoin, etc., etc.

You also state, "I've seen esXpress hang part way through a backup, refuse to delete thesnapshot and had to migrate everything else off the host and restartthe host". We see this happen when the ESX host iSCSI or SAN connection is interupted. We also see this with W2K3 VMs running MS/SQL. This is not exclusive to esXpress, simply adding a snap shot to a W2K3 VM with SQL can casue it to freeze. You can easily verify if this VM has the problem using the VI3 client.

Another thing I see is you are using v3.0.7 of esXpress released in March 2007. While this version is functionally stable, our latest version has vastly improved the logging, reporting and error detection. I would suggest going to your personal download page to retrieve the 3.0.22 release.

And last, I see no record of your contacting our support staff. Our staff would have been able to quickly resolve your problem with no interption to your operations, no rebooting hosts, etc. Please contact our support before resorting to a reboot.

If you would like, you can run the 'phd-support' program and our staff can review the findings/output with you.

Thankyou

Ken Harbin

www.esXpress.com

View solution in original post

0 Kudos
fontyyy
Contributor
Contributor

Thank you very much for your response, I have edited the title as this is clearly not a problem caused by your software.

To be honest restarting stuff here isn't as a big a deal as it would be at many commercial organisations and from reading your 2nd paragraph I'm presuming there is no fix for the "SQL hang" issue, however as the VM isn't actually working at the time (although it does respond to pings) this isn't a big deal for me now all our VMs are on the SAN and can be migrated off.

So discounting the slight issue this morning (there was little point calling anyone, to fix it took 20 mins max, nothing went down and just to make a support call to the states I'd have had to ask my manager who's faith in VM's in general is poor) is there no way to backup the SQL servers that are inclined to hang at snapshot? Does CB (which i stil have not got round to getting working) do the same thing?

Thank you again.

0 Kudos
kharbin
Commander
Commander

VCB does the same thing. In fact any backup software that performs a hot backs up of the VMDK file will have that problem with certain VMs (everyone has to use the VMware snapshot manager). But, we have had very encouraging results from users running ESX 3.0.2 with all patch updates installed. VMs that would lock consistantly by merely adding a snapshot, now run just fine. Try upgrading a single host to v3.0.2 with all patches, migrate the VM to it and see if its happy there. Might be a simple fix for the problem.

Ken

fontyyy
Contributor
Contributor

Thanks for that, I've upgraded one box and left it running nothing much, just a linux email filter I can recreate with with an ISO and the SQL2005 server that the backup failed on every time.

The upgrade went fine and couldn't be easier, all I did was migrate everything off the host, boot the CD, press enter and click upgrade.

The VM now snapshots fine and esXpress did back it up. I'll look into moving to the latest version of esXpress soon.

Thank you again.

0 Kudos