We are using DFM 5.2 and have OSSV deployed to abou 80 to 100 hosts.
The backup is managed via DFM usign a few different datasets.
Datasets with around 20 or less relations are working without any issues.
Once getting to 25 and more relations in one backup set several systemstate backups will fail.
When doing a "snapvault update" via the console of the storage system to all 25 or more relations there are no issues.
Of cause some of the updates are getting queued and will take time before they start.
In the job I can find the messages Error Message: <storage system>: Qtree <server name>:SystemState is involved in another transfer
Scource Qtree: <server name>:SystemState
Destination Qtree: <storage systems>:/<datasetname>/<server name>_SystemState
Bytes Transferred: Approximately 0KB
Error Message: <storage systems>:/<datasetname>: NDMP error: GENERIC
The error happens each time exactly 6 min after start of the dataset.
It looks like when DFM starts the job there is a kind of a time-out which leads DFM to report the job as partially failed.
I did find some information on
- VSS snapshot creation timeout (sec) which we still have on the default of 180 sec
- dpMaxGetStatusRetries or dpGetStatusRetryInterval
Would it help to modify these seting, is that the right way?
Is there anyone seeing similar issues?
Any input is appreciated.