Virtual machine disks consolidation is needed

[eric]

Hello All,

I have a machine that is replicating off-site. In vCenter the machine has this notice: Virtual machine disks consolidation is needed

The only snapshot is the Veeam snapshot.

Is this going to be an issue?

~eric

[Gostev]

Hi Eric, are you talking about the source VM? I believe this message means that there are one or more hidden "worker" snapshots that must be consolidated. So, it is not referring to the Veeam snapshot. Thanks!

[kellogs]

Hi,

I have been using VBR 7 Enterprise Plus to replicate VMs and from time to time I am getting this on my source VMs "virtual machine disks consolidation is needed"

All our snapshots are triggered by VBR only.

This is ESXi 5.1U1 hosts running Vmware VSA (2 nodes cluster)

Anyone know why VBR 7 is triggering this error virtual machine disks consolidation is needed?

At the moment we have disabled our replica job.

Please help.

[foggy]

Most likely there're some "hidden" snapshots that are not visible with vSphere Client and should be consolidated.

[kellogs]

Hi Foggy,

I am wondering why would VBR created this "hidden" snapshot?

I thought after VBR completed the replication the snapshot supposedly to be removed automatically?

Thanks much

[foggy]

Yes, they are supposed to be deleted. However, occasionally can be left over due to some kind of issues (job crash, locked disk file preventing snapshot commit, etc.).

[wa15]

FYI, we had this issue recently and it was due to high latency on the SAN. Basically the latency was so bad on the storage backend that removing the snapshot was timing out. So that may be something to check as well.

[Vitaliy S.]

Farshid, thanks for letting us know the reason for this issue in your situation. In order to be notified about this problems in advance, I would recommend installing Veeam ONE free edition to monitor your datastore performance.

[trackstar]

Hello,



We have a Windows machine 2008 R2 with 1.7TB of total size (made up of 5 disks). In the last few months we have seen this issue often where the snapshot removal process takes forever (after the disk consolidation needed message notification). In snapshot manager, it doesn't show a snapshot but when you browse the directory, there is a orphaned disk. I have opened cases with VMware and Veeam. Veeam support told me the removal process is passed to VMware and it should be their issue. VMware took a few test snapshots for testing and then removed them (delete all). Some snapshots took a long time to remove.

I have been told/read that version 8 of Veeam has an indicator (snapshot hunter) to show any orphaned but would it help in our situation? The removal process is still taking place at the VMware level?

ESXi 5.5

Windows 2008 R2 with Exchange 2010 with 1.7 TB
Veeam 7 patch 4.
Thanks,
TT

Thanks,
TT

[Gostev]

Hello,

Yes, it would have helped in your situation for sure.
http://blog.vmjoes.com/2015/04/veeam-sn ... unter.html

Thanks!

[trackstar]

If I do an in place upgrade from 7 patch 4, do I have to make any setting change? Also, does version 8 have any other up to date critical patch?

[Gostev]

No setting changes are needed.

For patches, see the sticky topic (latest version).
v8 also has an automatic update check.

[trackstar]

Thank you. We successfully installed version 8 and patch 1.

Last question. We have days where the job ran over 24 hours (the removal of hidden snapshot). If a job named "X" is currently running and almost reach 24 hours and hasn't finished, would the next scheduled run attempt to start the same job? That would be bad. Just making sure.

[Gostev]

No, job cannot be started until it finishes running.

[vipes27]

We use Veeam for backups and replication of our VMware environment. Lately, I've been getting a couple of errors on just a couple of our servers, saying that it can't do the backup because it needs snapshot consolidation, but all automatic snapshot consolidation attempts have failed. It says the most likely reason is a virtual disk being locked by some external process. Is it safe to just consolidate the snapshots during production? Thanks.

[PTide]

Hi,

Is it safe to just consolidate the snapshots during production?

It is generally safe, however if the VM requires intensive I/O, a performance degradation may happen.

[bgalante]

Hi,
I have a VM with a snapshot that's a month old, there is a "VEEAM TEMP SNAPSHOT" that is stuck in the mix that occurred after the said snapshot.

Vsphere produced an alarm that a vm consolidation is needed. I tried to run a consolidation, it failed due to "Downtime exceeded 12 seconds".

It looks like I can go into vsphere and do a "delete all snapshots" which I have not tried yet.

Do you think it's safe to go into vsphere and do the delete all snapshots with a VEEAM TEMP Snapshot present?

[ACC]

If there are no Veeam backups running then yes its safe to delete the snapshot, however given the fact that the snapshot is month old i would recommend deleting it outside working hours. For one its takes a lot of IOs to merge the snapshot(s) back into the base disk and two you could end up with downtime of the VM when the stun occurs in the merge, this downtime can take some time if the storage isnt able to perform the merge fast.

[ACC]

If youre using hotadd also check the proxy/proxies that there arent any disks still connected to the proxy. If so disconnect those disks or the snapshot consolidation will fail.

[Vitaliy S.]

Brian, running a production VM on the 1 month snapshot cannot be considered as best practice. If it is an orphaned VM snapshots, then consolidate it outside business hours (as ACC says). On a side note, what vSphere and Veeam B&R are you at? BTW, do you have Veeam ONE installed that monitors all stalled VM snapshots?