Backup Job IO stops after Disk Finishes

[rgarvelink]

This is the strangest thing but I noticed that when a job is running if the disk has multiple vmdk's both will be backing up at normal performance. When one of the vmdk's finishes all read and write IO seems to stop for 10 - 20 minutes. The processing rate says it isn't changing and the duration continues to go up, but both the performance and throughput graphs show nothing is occurring.

Eventually it seems to snap back and the job begins to run again, but the processing rate has dropped drastically and the performance is cut by like 90% and then slowly begins to improve.

VEEAM is running most current, v8 Update 2, but it was doing this on Update 1 also. Hardware is a Dell R720, Perc H310, RAID 5, local repository. Veeam is on a Server 2012 R2 VM, fully patched, and it has 1 single proxy and the local repository on this server. Both repository and proxy are set to handle 3 concurrent jobs.

Does anyone have any thoughts?

[Dima P.]

Hello Ryan,
Could you please check and let us know what Data Processing Settings are you using? Thanks!

[foggy]

I also recommend to let our engineers know about the observed behavior by opening a case with technical support. Thanks.

[Delo123]

We are also seeing this, i believe since veeam8.
*Reminder to myself, open a Ticket tomorrow morning...