Backup exec 2012 - Error backing up exchange (Writer status)

[soulseeker]

Dear Forum,

we are using BE2012 for backing up our server farm and mainly the jobs run without issues (or at least solvable issues) but backing up our exchange server sometimes gives me troubles. We are using two jobs - one SDR-Job for backing up the whole server and one job for backing up the exchange database - full on weekend and incremental during work days. The sdr-job and the database full backup often fail with error messages like these:

Auftrag beendet am Montag, 2. Dezember 2013 um 21:04:51 Abschlussstatus: Fehlgeschlagen Endgültiger Fehler: 0xe000fed1 - Beim Abfragen des Writer-Status trat ein Fehler auf. Endgültige Fehlerkategorie: Ressourcenfehler Zusätzliche Informationen zu diesem Fehler finden Sie unter der Verknüpfung V-79-57344-65233

We use an smb-share on a NAS-storage. The incremental database job runs always without troubles. The full jobs do not always fail, but with a 50% chance.

What makes me wonder - when I open the backup sets I can see that the jobs wrote gbs of data, but not as much as they should with a fault free backup. Nonetheless I can choose a date for recovery of system state with a timestamp of a failed and incompleted backup. But what would happend if I do?

Thanks and regards

Marcel

[Jaydeep_S]

Do both the jobs backup Exchange databases? Are both the jobs running at the same time or do they overlap. Cause concurrent backup of Exchange database is not supported by VSS.

http://msdn.microsoft.com/library/office/bb204080(v=exchg.140).aspx

If this is not the case, you need to confirm if the VSS writers are stable at the time of backup. Also, need to investigate what is causing the writers fail incase of an error.

[soulseeker]

Thanks for helping me!

Yes, both jobs backup the exchange database ...one job runs saturday, the other one sunday. How do I check, that the vss writers are healthy on the exchange server? Strangely, this only seems to affect  the full backups.

By the way - is it reasonable to backup the database AND do an SDR? I created the jobs because I thought that in case of an exchange crash it would be easiest to restore the SDR first as a virtual machine and afterwards restore the incremental backups.

[Jaydeep_S]

It should work or you could have the jobs run at a seperate time.

[soulseeker]

Sorry, still confused. I try to explain better :)

We run a virtual exchange 2010 with all roles on one server.

We do an SDR backup with all all drives, system state and exchange databases that runs as full backup on saturday. No incremental backups on weekdays.

On sunday we run an exchange database full backup, that does incremental backups on weekdays. Is this "overkill"? I read around the best practise documents from symantec, but I couldn't find an answer.

As you can imagine this creates a lot of data ... and I am not sure if it is necessary to backup SDR weekly.

Does it make sense to use only one SDR backup-Job with incremental weekday-jobs and don't do separate informationstore backups?

[Jaydeep_S]

My suggestion on this would be to perform a SDR Full on weeekends without Exchange Databases. And keep the Exchange Full - Incremental as is.

As for the SDR backup set, you really only need tro keep 1 backup set at any given time. So for the retention on that backup, you set it to overwrite the following week after the next full backup.

[soulseeker]

Thanks a lot ... I'll check if this will cause less troubles. Maybe the source for the problems was the "double exchange database backup". At least it will free a lot of space on our storage ;).