Losing pings during creation/deletion of snapshots

[rcustersp]

We are using Vizioncore vReplicator to replicate one of our file servers every hour. This server is experiencing network timeout in applications/reporting services that connect to other servers. I just ran a constant ping during the last replication and found that while vCenter was deleting the snapshot for this server it lost 9 pings. Is this normal? Is ther anything I can do to prevent this?

[DSTAVERT]

Especially when you have a quiesce memory operation the VM will experience a short "freeze" while memory is committed to disk.

http://kb.vmware.com/kb/1015180

[rcustersp]

Guest quiescing is not being used on this snapshot.

[Borja_Mari]

Hi,

this behavior you are talking about can be "normal".

I will try to explain it.

When a snapshot disk is merging with its base disk (delete), while this task is going on a temporal snapshot is created just to hold all the disk operations in this windows time.

Then after the snapshot disk consolidating, the temporal snapshot is merged with the base disk too. While this happens the vm is in a frozen/stopped state. This usually just happens in 2 seconds.

But if the temporal snapshot was a little big (the vm was suffering a moderate-big usage while the first snapshot merging task), then the final merge can be more than 2 seconds. Then this can explain the 9 lost pings you are talking about.

If you want to gain more advanced knowledge about snapshots, i would recommend you to check this post.

Hope this helps

Regards/Saludos,

Pablo

Please consider awarding any helpful or corrrect answer. Thanks!! - Por favor considera premiar cualquier respuesta útil o correcta. ¡¡Muchas gracias!!

Virtually noob blog

[petedr]

If your snapshot is large I have seen lost pings on the snapshot deletion from Vmware.

www.phdvirtual.com, makers of PHD Virtual Backup for Vmware and Xen Server, formally esXpress

[rcustersp]

With all of it's drives the server is about 960 GB.

[petedr]

If you have extensive changes occurring inside the VM during the time the snapshot is in place that would cause large snapshot growth. That could cause a ping loss on snapshot deletion..

www.phdvirtual.com, makers of PHD Virtual Backup for Vmware and Xen Server, formally esXpress

[VMmatty]

The freezing when you take a snapshot is usually caused by the "Snapshot virtual machine memory" option. This causes all of the contents of the VM to be written to the disk so it can be restored intact. Depending on how much RAM the guest has it can take a decent amount of time and all operations inside the guest will be halted.

The alternative is to deselect that option, but that really limits your ability to revert to that snapshot. If you roll back to a snapshot where you did not snapshot the virtual machine's memory it will be as if you powered it up - it will boot up and continue where you left it before the snapshot. That means that any running applications, etc, will stop and need to be restarted. It's fine if you know you're going to delete the snapshot but if there is a chance you won't then you really need to snapshot the VM's memory also.

[rcustersp]

In this case the memory snapshotting is not being used.