cancel
Showing results for 
Search instead for 
Did you mean: 

Upgrade from 7.1.0.3 TO 7.5

Michael_Blizzar
Level 3

Hello Backup Gurus,  I wanted to share my experience with my upgrade a few days ago.  After the upgeade my MSDP diskpool would not come up.  Here are the steps I did to fix the issue.  Hopefully This will help other that may run into this issue.

 

Disk volume is DOWN (status = AdminUp, InternalDown)

 

The command "nbdevconfig -changestate -stype PureDisk -dp <DiskPoolName> -dv PureDiskVolume -state UP"  failed to set state of disk volume to UP.

 

On the MSDP host we found that the deduplication engine (spoold) service was in a stopped state.  The following was seen in <storage path>\log\spoold\spoold.log when attempting to start the service:

 

March 21 08:58:07 ERR [0000000000DF98E0]: 25017: FileCopyA: could not open destination file S:\DEA\data/12198.bin
March 21 08:58:07 ERR [0000000000DF98E0]: 25017: FileMoveA: copy failed
March 21 08:58:07 ERR [0000000000DF98E0]: 25001: dcOutPlaceUpdateReplay: could not move S:\DEA\data/journal/12198_12198.bin to S:\DEA\data/12198.bin (object already exists)
March 21 08:58:07 ERR [0000000000DF98E0]: 25001: _storeSingleDCCompactRecover: dcOutPlaceUpdateReplay (object already exists)
March 21 08:58:07 ERR [0000000000DF98E0]: 25001: _storeInit: failed to recover from single container compaction
March 21 08:58:07 ERR [0000000000DF98E0]: 25001: Could not initialize DataStore Manager
March 21 08:58:07 INFO [0000000000DF98E0]: WSRequestExt: submitting &request=4&login=agent_3_23945&passwd=********************************&action=newevent&data=EVENT%7Bversion%3A1%3Btype%3A1%3Bid%3A0%3Bdate%3A1332334687%3B%7BLEGACYEVENT%7Bpayload%3Asev%3D1%3Btype%3D1017%3Bmsg%3DFailed%20to%20Activate%20the%20Store%20Manager.%0A%0APlease%20check%20the%20server%20log%20for%20additional%20information%20and%20probable%0Acause%20of%20this%20error.%0A%0AThe%20application%20has%20been%20terminated.%3B%7D%7D
March 21 08:58:07 INFO [0000000000DF98E0]: sessionStartAgent: Server is Version 7.0000.0012.011, Protocol Version 6.6.1
March 21 08:58:08 INFO [0000000000DF98E0]: WSRequestExt: submitting &request=4&login=agent_3_23945&passwd=********************************&action=newevent&data=EVENT%7Bversion%3A1%3Btype%3A4%3Bid%3A0%3Bdate%3A1332334688%3B%7BSERVERSTATUSEVENT%7Bstate%3A1%3Bserver%3ASymantec%20PureDisk%20Content%20Router%3B%7D%7D
March 21 08:58:08 ERR [0000000000DF98E0]: 26016: Store Manager: Activate failure.

 

This tells us that container file from S:\DEA\data\journal\12198_12198.bin cannot be copied to S:\DEA\data\12198.bin because a container file with that name already exists and either the file in the journal folder should be removed, or the the file in the data folder.

To verify, S:\DEA\data\journal\12198_12198.bin  was moved from its folder and we ran

"dcscan --verify 12198.bin"

to verify the container contents.  This produced CRC errors so S:\DEA\data\12198.bin was removed and replaced with S:\DEA\data\journal\12198_12198.bin (and renamed to 12198.bin).

We again ran:

"dcscan --verify 12198.bin"

This time there were no CRC errors and so the deduplciation engine could be restarted.

 

We then reset the disk pool with:

 

Ran nbdevconfig -changestate -stype PureDisk -dp DEA_MSDP -state RESET

 

The disk pool and volume were then reported to be in an UP state.

Initially backup failed with status 84, but after some time the disk pool was accepting data (I had checked for correct response with crcontrol --getmode, crcontrol --dsstat and initiated queue processing with crcontrol --processqueue and this was all as expected)