Views:

Summary



When vStor and OSS objects are in the same node group the vStor backups might be incorrectly reporting success; however are not recoverable .

Symptoms



Very first backup after deployment of vStor completes successfully (RC=0). Subsequent backups report below error (in Job log) however completed successfully after BASE backup is forced

X.X.X.X sssnap Wed Aug 29 14:43:45 2018 SNBSNP3068W 5000: Primary and Secondary are out of sync for volume [C:\]....
X.X.X.X sssnap Wed Aug 29 14:43:45 2018 SNBSNP3068W 5000: Volume [C:\] : Forcing the BASE backup...
X.X.X.X sssvh Wed Aug 29 14:43:46 2018 SNBSVH_626W  Task 7 will perform a BASE transfer for destination re-synchronization

DPX Environment:
OSS server and vStor Storage_CTL nodes are in the same node group 

Replication job reports following error:
Mi 29. Aug 13:41:09 UTC 2018 vreplicate %Performing base, no previous successful snapshots for vol01 
Mi 29. Aug 13:41:09 UTC 2018 vreplicate %There are ZERO snapshots of the source volume

Restore job reports error:
Task 1 cannot find its restore source snapshot and it's not been double protected. (snap ID=7b4245582d373636463643333033312d3542383641304434303634387d0000000000000000000000000000000000000000000000000000000000000000000000) 
Message: Task 1 NDMP_LOG: id(3.504), type(NDMP_LOG_ERROR), text(iSCSI mapping could not be performed. CDOT need iscsi service) 

 

Resolution



This behavior has been addressed in a DPX Maintenance Update. Please perform a DPX Autoupdate to apply latest DPX release.  (This can be performed from Help | Autoupdate DPX with the DPX Management Console.)
See the Software Update Release Notes for DPX 4.5.4 patch 283 referencing Issue ID DPSUST-4426 .
If the issue continues to occur after applying the update please contact  Catalogic Software Data Protection Technical Support referencing the Issue ID.


Workaround:
  • Create a new node group and move all vStor Storage_CTL nodes to this new node group
  • Create a new job definition for any backups pointing to the moved vStor Storage_CTL nodes
- Any block level jobs affected by this issue will fail with " "Reason(Job definition has storage controller target defined but has an incompatible OSS target flag set. Resave the job definition with new name to correct the issue. If the problem persists contact Catalogic Software Technical Support.)"  to ensure their job definitions are corrected.
- Start the DPX Management Console and create or resave  a new job definition for any backups pointing to the vStor Storage_CTL nodes