Proxy Unavailable and failed to start backup process

[LunenK]

Hello,
i installed new Veeam B&R v12 and AHV Plugin 4.0 and then i added nutanix cluster (AHV6.5).
after above, i published Veeam AHV Proxy.

now here is problems.
When i create backup job and start it, Proxy is going unavailable on Veeam B&R Console
and backup job failed with below message.

"Backup failed. Failed to start a backup process: Can't get policy session with tag [476cf53d-1282-419b-acc1-24dd4cf11f0e]"

This message only can see on proxy web console, never on B&R Console.
At Prism, created some tasks.
"Snapshot Created"
"Snapshot Deleted"
"Volume group Created"
"Volume group Deleted"
but all task logs are not designate any VM, and there is nothing on VM Tasks what i try to backup.

anyway, i still can access Veeam AHV Proxy web console.
so i create snapshot job and start it. works well with proxy unavailable states.

something wrong on networks? or AHV policies?
Proxy is on same network with CVM.

[LunenK]

Unavailable Proxy never comeback to normal.
the only way to make normal is re-install proxy or remove from B&R and add existed proxy again.

[Mildur]

Hi Lunen

Do you have a case number for this technical issue? Please provide a support case ID for this issue, as requested when you click New Topic.
Without case number, the topic will eventually be deleted by moderators.

Our customer support can provide you an answer after analyzing the log files.

Best regards,
Fabian

PS: support can only help if you upload logs https://www.veeam.com/kb1832

[LunenK]

here is my case number 05918392.
thanks.

[dinglin]

Got the same issue
Try reinstall VBR with SQL Server database, and it works

[magnusvr]

Any updates on this? Getting the same with Veeam 12, fresh Postgres install, and AHV Proxy 4.
After a while they appear as Unavailable, and backup jobs fail with "Backup failed. Failed to start a backup process: Can't get policy session with tag xyz"

Do I need to reinstall Veeam?

[Mildur]

Hi Magnus

We only got 1 case number and it was for the community edition.
If you face the same issue, please open a support case and let us know the case number.

Best,
Fabian

[magnusvr]

Hi Fabian,

case 06094908
Installed AHV Proxy for two cluster. Install went fine and they were OK in the Repository pane in Veeam.
As soon as we try to start a job, it kicks off and starts to do things in Prism, but it also fails with
6/2/2023 1:47:54 PM SERVERZAPR : Backup failed. Failed to start a backup process: Can't get policy session with tag [daef3cfd-19b5-4c38-87da-1cada732c15e] —

Then it goes "unavaialable" in Veeam. I can do the Properties-next-next-Finish just fine, all is accepted and connected, but still unavailable.
Tried so far.
Rebooting the Veeam server and AHV Proxy.
Made a Windows firewall rule, then disabled the whole Windows Firewall, no diff.
Removed the Proxy from Veeam and re-added it. Comes up as unavailable straight away.

We have two fresh AHV 6.5.2.5 clusters, same on both.
Everything is in the same LAN/IP-range, CVMs, Veeam, AHV Proxies etc.

Thanks,
Magnus

[magnusvr]

#06094908
Just had a good session with Alexandra, we checked everything, redeployed a new proxy etc, everything is green and fine, but same message. As soon as we run a backup it goes unavailable (but available from the proxy end), and the same error.
Workaround is to redeploy the whole environment, new veeam server etc. Afternoon and evening activities secured.
It seems there is a bug, since using it with postgres is supported but just seems to not work. There were two more cases support new about where they ended up using SQL express again as the resolution.

[Mildur]

Hello Magnus

Thank you for the update.
So you are saying, with MSSQL everything should work everything and with PostgreSQL, the integration with Nutanix stops working? Let me check with support. I want to see the other cases.
Redeploying everything is the last thing we should try. If it works with MSSQL, then I would spin up a new MSSQL (Express) installation and migrate the database.

Best,
Fabian

[magnusvr]

Hi Fabian,
as this is a completely fresh install and no backups have been taken, there is nothing to migrate. Easier to just reinstall.
I'm keeping the postgres server for the first cluster to be able to trobleshoot further, just for "fun".
All my important VMs are currently in the second AHV cluster, so I'm installing a new Veeam server with SQL for that, to get backups for the weekend.

[Mildur]

Hi Magnus

Ah. Thanks for the update.
Please let me know if a new VBR server with MSSQL works.

Best,
Fabian

[magnusvr]

#06094908
Alexandra came back with a finding! It's the region and language of the Veeam Server, AHV Proxy and Postgres DB.
Our server was set to English OS, but region and keyboard was Swedish. This confuses the software, as its trying to talk to the database with a different date/time format.

This fixed it;
- Install US English language pack
- Change Region to United States
- Make a config backup.
- Uninstall Veeam Backup & Replication completely! ( also manually uninstalled a few services that was left)
- Uninstall Postgres. Delete the Postgres-folder in C:\Program Files (not sure if needed).
- Install Veeam12 and go for a new Postgres install. Let Veeam install.
- Install the AHV package.
- Start Veeam and restore configuration.
- Right click your AHV Proxy and do a next-next-finish walkthrough. It will pop back to "available" status after a shotrt while.
- Backup and rejoice!

Thanks Alexandra in support!

/Magnus

[Mildur]

Hi Magnus

I'm glad you could find a solution together with Alexandra.
Meanwhile I discussed the case with our support management. It seems that we hit a known issue which RnD is already aware off. We will fix the date format issue in combination with PostgreSQL server in one of the next updates.

Best,
Fabian