Symptoms
Legacy scheduled VPS backup tasks fail due to insufficient limit for backup space on subscriptions.
The failing backup tasks keep rescheduling, with each new run they create a new backup on the backups storage.
The backups do not fit into the ordered backup space:
1001324 Scheduled backups diskspace Usage 8.5 GB Limit 10 GB Available 1.5 GB
/var/log/pa/core.log contains the following entries:
May 25 12:07:25.187 : DBG [task:159053600:11675 p:-default-threadpool;-w:-Idle:494 pau]: c.p.p.s.r.e.ResourceUsageMutatorBean checking subscription resource #172055 for overuse before applying delta: 1779512
May 25 12:07:25.187 : DBG [task:159053600:11675 p:-default-threadpool;-w:-Idle:494 pau]: c.p.p.tracer exit by exception: com.parallels.pa.service.resources.ejb.ResourceUsageMutatorBean.beforeCompletionjavax.ejb.EJBException: Plesk.ErrorHandling.ResourceManagement.Overuse: IDL:Plesk/ExSystem:1.0
May 25 12:07:25.189 : DBG [task:159053600:11675 p:-default-threadpool;-w:-Idle:494 pau]: c.p.p.s.o.OTSTransactionImpl commit failedjavax.transaction.RollbackException: ARJUNA016053: Could not commit transaction.
May 25 12:07:25.190 : ERR [task:159053600:11675 1:22731:7f462ef07700 lib 1109728921]: [ CosTransactions::Current_impl::commit] caught an exception, rethrowing
Cause
The behavior is acknowledged as a software issue CCU-15494 (Could not commit transaction. taskScheduledBackupVPS fails)
Resolution
There is no way to prevent task failures, so it is required to delete the backups manually.
To avoid the massive problem, it can be suggested to increase the backup space limit for all subscriptions temporarily.
Contact TAM/PTA to obtain details regarding the issue resolution in product updates.