B&R 9 -very Poor Random I/O on ISCSI Repository to a FAS2554

[WinstonWolf]

Hello ,

A Question , we have an W2008R2 Veeam B&R 9 Server.
We have an FAS 2554 from Netapp , on this FAS we create 5 ISCSI Luns which are connected Via 10GBit Network Connection to the Veeam Veeam Backup Server .
On this Server we have therefore 5 Drives
The Problem is that Windows File Restore from a VM ( Repository ) needs very long time or hangs up ( over 1 Hour ) . And the merging for the virtual Full Backup to Tape needs also very long time 26 Hours .


A Case is Open .
Today we make a Measurement with the Windows diskspd Tool .

We have
26 MB/ s Random Access on ISCSI Drive M:
261 MB/s Sequenz. Read Access on M:

16MB/ s Random Access on Drive P:

M and P are Repositorys for the Veeam Backup Server .

Are this values ok ? I think not .

Thanks
Michael

[PTide]

Hi,

The Problem is that Windows File Restore from a VM ( Repository ) needs very long time or hangs up ( over 1 Hour )

Where do you have your mount server configured?

16MB/ s Random Access on Drive P:

M and P are Repositorys for the Veeam Backup Server .

Are this values ok ? I think not .

26 MB/ s random access does not seem very bad, however 16MB/s is a poor performance, indeed. Since you used a third-party tool to measure your storage performance and it showed poor results there is not much that we can do here.

Thank you.

[alanbolte]

Do you have the repository volumes configured for NetApp ONTAP deduplication and compression?

[jveerd1]

My best recommendation is to deploy NetApp Harvester with Graphite and Grafana. Within an hour you will know if your FAS the bottleneck or not. Just google for it, lots of guidance available on the net.