Skip to main content
Solved

Backup PostgreSQL 2Node Cluster (active-standby)

  • June 13, 2023
  • 5 replies
  • 456 views

Forum|alt.badge.img+14

Hello, 

One of our customers approached us with a PostgreSQL cluster (active-standby). How can we best set up the backup here, since Commvault cannot create a backup from standby.

Kind Regards

Thomas

Best answer by Punit Koujalgi

Hello,

Going by logs, seems like replication is lagging behind a lot. Backup process only continues when Standby is up-to-date but here it seems we keep waiting for a long time for Standby to catch up.

 

29774 744e 06/14 12:48:34 5345052 PostgresStandbyBackup::WaitForLogTobeApplied() - Replay location [0/2C0000A0] log Location to be applied [2/4F000028]

29774 744e 06/14 12:48:34 5345052 PostgresStandbyBackup::WaitForLogTobeApplied() - Log NOT yet applied [2/4F000028]. Sleeping for 1 minute.

 

Can you please address the Replication to Standby and retry backup?

 

Thanks,

Punit

5 replies

Meera
Vaulter
Forum|alt.badge.img+7
  • Vaulter
  • June 13, 2023

Forum|alt.badge.img+14
  • Author
  • Novice
  • June 13, 2023

Hello @Meera

Thanks for the fast respond. I'll try to set it up tomorrow and report back if everything worked.

Kind Regards

Thomas


Forum|alt.badge.img+14
  • Author
  • Novice
  • June 14, 2023

Hello @Meera

We tried to set up the whole thing today. The backup can be started, but it does not continue and gets stuck in the "Running" status. I have attached a few logs here, maybe someone can help us. 

Kind Regards

Thomas

 


Forum|alt.badge.img+1

Hello,

Going by logs, seems like replication is lagging behind a lot. Backup process only continues when Standby is up-to-date but here it seems we keep waiting for a long time for Standby to catch up.

 

29774 744e 06/14 12:48:34 5345052 PostgresStandbyBackup::WaitForLogTobeApplied() - Replay location [0/2C0000A0] log Location to be applied [2/4F000028]

29774 744e 06/14 12:48:34 5345052 PostgresStandbyBackup::WaitForLogTobeApplied() - Log NOT yet applied [2/4F000028]. Sleeping for 1 minute.

 

Can you please address the Replication to Standby and retry backup?

 

Thanks,

Punit


Forum|alt.badge.img+14
  • Author
  • Novice
  • June 16, 2023

Hello @Punit Koujalgi

Thank you for pointing this out. Our database administrators were able to fix the problem now. The reason was that the two nodes were not in sync. 

Kind Regards

Thomas