Unified Manager data directory running out of disk space due to large number of backups
Applies to
- OnCommand Unified Manager (UM) 9.5 and below
- ActiveIQ Unified Manager(UM) 9.6+
Issue
- UM over time will fill up the data directory due to large number of backup files stored
- Acquisition will not function properly for the clusters
- Subsequently there may be old data in the reports and false positive alerts present on the UM
- Upgrade to higher versions will fail with error:
unable to perform operation due to available disk space limitation
- An error message similar to the following can be seen in the ocum-error.log:
[root@motorhead_ocum]# cat ocum-error.log | grep backup
2018-02-14 06:20:47,855 ERROR [oncommand] [taskScheduler-4] [Backup|ScheduleBackup] [c.n.d.i.backup.BackupRestoreManager] Not enough space to carry out backup operation
2018-02-14 16:44:53,591 ERROR [oncommand] [taskScheduler-4] [Backup|ScheduleBackup] [c.n.d.impl.backup.BackupScheduleJob] Error while creating the backup
[root@motorhead_ocum]#
- Some of the incremental backups will not be compressed and will stay at the database-dumps-repo directory with its original size as an .sql file and not as a .7z archive