Re: Barman was hung after a Transport endpoint error

On Thu, Nov 20, 2025 at 9:29 AM Sri Mrudula Attili <sri(at)ebi(dot)ac(dot)uk> wrote:

> Hello Team,
>
>
> This is regarding our Barman backup host, which seems like hung for 5
> days with no backup, streaming, cron etc happening.
>
> We have checkmk monitoring that notifies us if any one of the parameters
> of Barman check fail, but we havent hot any events from checkmk as barman
> check was showing all OK and no FAILED.
>
> What do you explain this situation in Barman.
>
>
> *To provide further details about the issue:*
>
> - We user Data Domain as barman backup home for storage.
> - On 5th November data domain endpoints is not connect but the mount
> was still up on the host. We still dont know from the data domain side what
> caused this issue.
> - Barman log has errors* : [2800226] barman.cli ERROR: [Errno 107]
> Transport endpoint is not connected:
> '/nfs/DDbackup/pgsql-****-****/.active-model.auto'*
> - If the end point was missing barman should fail completely and the
> checks should fail as well, Which in turn would have helped us to identify
> the issue immediately.
> - The only way we figured out that there was an issue, on November
> 11th one of barman backup maximum age parameter failed saying that weeks
> backup dint happen.
> - Upon checking why the backups failed etc, I went to barman logs and
> found out that all this time until the endpoint issue resolved barman was
> not doing the barman checks or barman check was hung at a state
>
> *Below are the configuration details:*
>
> - Barman version : 3.10.0 Barman by EnterpriseDB (www.enterprisedb.com)
> - Postgresql version: (PostgreSQL) 16.9
> - Number of hosts configured: 26
> - Red Hat Enterprise Linux release 8.10
>
>
Is NFS hung?

--
Death to <Redacted>, and butter sauce.
Don't boil me, I'm still alive.
<Redacted> lobster!