Skip to content

wal dir full

reproduce the issue

Try executing the backup again and check if the issue still persists...

hostname:~ # /usr/NBPostgreSQLAgent/nbpgsql -o backup
Assuming PostgreSQL installation mode
Failed to execute PostgreSQL query: SELECT pg_stop_backup()
ERROR:  canceling statement due to statement timeout

Failed to execute PostgreSQL query: SELECT pg_stop_backup()
ERROR:  exclusive backup not in progress

understand the issue

Check the file system is fine?

hostname:~ # df -Ph
Filesystem                   Size  Used Avail Use% Mounted on
/dev/mapper/vgroot            25G   12G    5G  48% /
/dev/mapper/vgdata-lvpghome  5.0G    1G  4.0G  20% /home/postgres
/dev/mapper/vgdata-lvpgwal    20G   20G   84K 100% /srv/postgres/wal
/dev/mapper/vgdata-lvpgdata   45G   29G   17G  64% /srv/postgres/data
hostname:~ #
Congrants! You found the nasty guy. WAL dir full is bad. If you urgently want to fix something:

  • Goto the WAL directory
  • Check with backup team which WALs are already backed up
  • cleanup old WALs with: pg_archivecleanup /srv/postgres/wal/ xxxxx.xxxxx.backup (includes the latest backup)

!!!IF YOU'RE NOT 100% SURE WHAT WILL HAPPEN: DRY-RUN this command with -n option!!!!