Veeam and Server "Hang" on Backup Copy Job.

[StevenMeier]

I have a case open : Case # 03512997

This has been a tricky one as the "Hang" is intermittent.

We are running 9.5 Update 4a

2 x HPE Apollo 4510 Gen Boxes, both with 128 GB of Ram , 2 x 12 core , 60 x 12 TB Disks, 1 x Fibre Card , 1 x 2 port 10 GB NIC.(There is 1 Box at each Data Center, connected by 10 Gb layer 3 Link)

The physical Boxes run Windows Server 2019, and Veeam options , Proxy, repository and Gateway service for connection to HPE storeOnce 5250 boxes.

Backup jobs run fine.
The failures only seem to occur on Backup Copy Jobs. Where data is going from site A to B and vice versa for the backup Jobs.
This like I said can be very intermittent .

Windows is up to date and there is nothing in the event logs
I have logged calls with HPE and while No fault was found they did recommend putting the latest ProLiant support Pack released in April this year , which both have had done.

The symptom is basically the Backup copy job is running (in the latest crash today) only one Backup copy job and no other Veeam tasks Running at all.
The Destination Server for the backup copy job Hung.

ILO I can only see the login Screen and have to power off.

When Server comes back up I disable in Veeam management the job and restart it and it finishes fine.

The Veeam mgmt. Server is a virtual so does not run on ether of these work horse boxes.

Any Ideas would be much appreciated

I Have logged the job when the crashes started and we have been slowly adding jobs and turning them back on , so this is the first crash in 3 to 4 weeks.
Cheers
Steve

[HannesK]

Hello Steve,
I suggest to continue working with support as it is not really possible to find a solution for such issues in the forums.

What I cannot find in your text is, whether your backup copy jobs go to the StoreOnce boxes or Apollo-to-Apollo. If BCJ is going to StoreOnce, then I can just add a guess to try active full backup copy jobs ("Read the entire restore point from source backup instead of synthesizing it from increments" checkbox ) instead of of synthetic (full) BCJ

Best regards,
Hannes

[StevenMeier]

Ok Thanks,

The techs have been very helpful, But because it does not do it all the time its very hard to track Down. The Tech is escalating to Level 2.
But to answer your Question.
The backup Copies go from Site A to Disk on the Apollo ie
we store 4 weeks of Backups local at site A: those backups do a backup copy to the Apollo at site B
Monthlies from Site A go to the storeOnce at Site B and the source is set to the repositories at Site B and Site A (they Generally pull from the repositories at Site B in this example). The jobs to storeOnce never seem to Fail or cause a fail.
It only Seems to be backup copies from Site A to Site B which is Apollo to apollo