Hypersvisor Snapshots

[Mike Maggio1709162854]

We have a VM Snapshot schedule that keeps failing. We are seeing these errors:

 

The snapshot schedule failed because a snapshot could not be created 

 

SM has thrown a generic python exception 

 

This is Xneserver 7.6

 

I can't find anything on the web regarding this error that is relevant. Has anyone seen this and can you suggest a resolution?

 

Thank you,

Mike Maggio

[Alan Lantz]

Maybe rebooting the server, and you may have reached your limit on snapshots. 

Would take SM.log info to try and sort it out. If a server reboot doesn't fix it I

would export/import the VM and see if that clears it up.

 

--Alan--

 

[Mike Maggio1709162854]

Thanks Alan. Is there a limit on snapshots?

[Mike Maggio1709162854]

2 minutes ago, Luisa McLawhorn said:

 

An array-based snapshot is a copy of the image of a running VM or application server at a specific point in time and, as a result, the snapshot will appear as a “crash copy” of that VM/application if it is fully restored and accessed.

I"m not sure what you mean. The snapshots we are trying to create are not being created.

[Alan Lantz]

The snapshot limit is around 30, but you usually see a message about the snapshot chain being too long.

Just thinking out loud you hit some sort of limit on XenServer or storage preventing the snapshot from

occurring. Local or remote storage? A single VM or multiple VM's? Windows or Linux ?

 

SM.log should give you some insight.

 

--Alan--

 

[Tobias Kreidl]

SR space issues, perhaps?

[Mike Maggio1709162854]

Actually, it seems our SR's are unplugged. I ran a repair but that didn't fix it. I'm thinking the best was to fix this to to reattach each SR.

[Tobias Kreidl]

You can try that; could be that a reboot might even fix things if the SRs are not able to be manually attached.

-=Tobias

[Mike Maggio1709162854]

I will try rebooting first to see if that fixes it.

[Mark Syms]

You should also be aware that XenServer 7.6 is long out of support and is not receiving any updates, even security related fixes.

[Mike Maggio1709162854]

Thanks Mark. I am aware of that but I am reluctant to upgrade until I get this issue fixed.

I rebooted all the hypervisors last week during mainatenance but that did not resolve the problem. I also noticed that when I tried to migrate a VM from one SR to another, we get the same generic python error. I am wondering if the problem is tieh the storage device itself (Equalogic). Has anyone experience an issue like this before?

[Mark Syms]

@Mike Maggio you can email me the /var/log/SMlog file from the primary server in the pool and I'll take a quick look. Address is just my dot separated name @citrix.com 

[Mike Maggio1709162854]

Thanks Mark This is a secure environment so I don't think i can do that. Is there something i can look for? Otherwise, I"ll have to get permission and the logs will need to be scrubbed.

[Mark Syms]

Just look for exceptions in the log.

Or, run "tail -F /var/log/SMlog" while you take a snapshot and it will all be in that block (save the log to a file if you use an ssh client like putty) and then you can read just what happens in the time frame of interest.

[Mike Maggio1709162854]

Thanks Mark. I will do that and let you know the results.

[Mike Maggio1709162854]

Hi Mark:

 

I ran the command and here is the rror:

 

LVHD over iSCSI EXCEPTION <class 'journaler.JournalerException'., Failed to read from journal (followed by the name of the journal)

[Mary ONeill1709163018]

Nice and amazing post this one is, thanks for sharing…

dgcustomerfirst

[Mike Maggio1709162854]

I"m back to this one again as it has not been resolved.

Two questions:

Do you tihink thi is a Citrix issue or an issue with storage?

Do you think upgrading to 8.0 (and then 8.2), which is in the plans, will resolve this?

[Mark Syms]

It's probably both, something went wrong and then the code is not handling the failure. As XS 7.6 is long EOL (as is Citrix Hypervisor 8.0) there is now no supported upgrade path that doesn't involve re-installing from scratch.

[Mike Maggio1709162854]

I though you could go from 7.2 to 8 and then to 8.2.

[Mark Syms]

34 minutes ago, Mike Maggio1709162854 said:

I though you could go from 7.2 to 8 and then to 8.2.

no, it was only ever supported to upgrade from N-2 to N, i.e. 7.5/7.6 -> 8.0. Long Term Support releases, i.e. XenServer 7.1 CU2 are expected to be upgradable to the next LTSR release, i.e Citrix Hypervisor 8.2.

Currently there are only the two LTSRs in support, i.e XenServer 7.1 LTSR CU2 (until August 2022) and Citrix Hypervisor 8.2 LTSR

[Mike Maggio1709162854]

I figured that but I thought I'd ask. Thanks.

[cartlon flores1709163273]

You should also be aware that XenServer 7.6 is long out of support and is not receiving any updates, even security related fixes. wendy's lunch time surveyzop.com