Has anyone run int a situation where Avamar VM backup snapshot removal is causing VM issues such as application hangs or "heartbeat timeouts"?
Working with a customer and implementing Avamar VM backups for set of VMs, and they are running into this issue - and it is severe enough that they have backed out of using Avamar for these particular VMs and have gone back to using NetBackup.
Is this something that could be alleviated by more creative scheduling, or is it more complicated than that?
All comments/feedback appreciated - thanks.
I have run into this issue before. Most likely a storage or hypervisor issue not Avamar. What hypervisor and version is being used?
From the customer:
"VMWare 6.x ... VM’s are on various hardware ... Dell Servers connected to SAN, plus some Dell Servers running Nutanix (Disk in part of Nutanix) and then some Nutanix Hardware running Nutanix."
(FYI - the customer providing this info is not directly involved in management of the VMware environment, at least in terms of its underlying resources - that is a different group with different people)
If they're using NFS datastores, the stun / unstun cycle can be really long. This applies to any application that takes VM snapshots, not just Avamar. Do they seem the same behaviour if they manually create a snapshot and then consolidate it some time later (i.e. after the snapshot has existed for the same amount of time as a typical backup snapshot)?
Simply put, we had the same issue with our environment. If we kick-off too many backups at once, we run into performance issues with the number snapshots being created at once. We opted to spread out backups across our window, running backups on different datastores at different times. It's been pretty simple and easy for us to simply add new VMs to different schedules.
Hope this helps!