Performance Issues since update to B&R 12.3.1.1139

[nintehol]

Hi all,

we have performance issues since upgrading to B&R 12.3.1.1139

We start the backup job (8 conncurent VM's) VM's stucking at 0% with 0KB/s for around 10 Minutes after that time backup starts with regular performance comparing to speed prior updateing.
next conncurent VM's during jog stucking again 0% with 0KB/s for around 10 Minutes and so on

B&R 12.3.1.1139 simple deployment on local disks (SSD), W2K22

vCenter 8.0.3.00400

3x hosts @ VMware ESXi, 8.0.3, 24585383

SAN (Datacore SANsymphony 10.0 PSP19) with Fibre Channel SAN 16G

Great Performance befor updating to B&R 12.3.1.1139

Does anyone have comparable problems?

We have raised support ticket as well.

thanks
Holger

[david.domask]

Hi nintehol, welcome to the forums.

Sorry to hear about the troubles and thank you for opening a case in advance. May I ask you to share the case number please?

I'm not immediately familiar with any known issues that match the above description, so let's see what Veeam Support sees in the logs.

Out of curiosity, since you have Datacore, are you using the Veeam Datacore plugin to allow Backup from Storage Snapshot as the transport mode or you are using DirectSAN/Hotadd/NBD?

If you are using Backup from Storage Snapshots, can you check some of the VMs and see if they have leftover snapshots? (Check both snapshot manager and on the datastore folders themselves as leftover snapshot delta files may not show in the Snapshot Manager)

I ask on snapshots as they will increase the processing time for Backup from Storage Snapshots.

[nintehol]

Hi David,

Case #07680407

Just want to unterstand if anybody elase is strugeling with with issue.

We are using Datacore in DirectSAN (FC) transport mode without any plugins.

The interesting thing:

Backupjob just "waiting" at 0% with 0KB/s for around 12 minutes then it starts with great performance and then performance went down again just to speed up again after the next concurrent set and so on.

I can summarise:

WAIT - GOOD PERFORMANCE - WAIT - GOOD PERFORMANCE and so on
Per each concurrent object group, job "stucks" for around 12 minutes at 0% - 0KB/s

[nintehol]

If anyone runs in this issue as well

For us worked this: Within backup job - "Guest Processing" - "Guest interaction proxy" - change from "Automatic selection" and manual select "SERVERNAME" of primary backup server

Performance restored to good performance prior updateing to B&R 12.3.1.1139

[david.domask]

Thank you for sharing the results of the investigation nintehol, and glad to hear Support was able to assist.

Guest Interaction Proxies make sense as the culprit here, so very glad you shared your solution.