Post DR issues

[asteng5000]

We had a DR scenario where we had to restore the DB. However the DR process was not straight forward. We had to recreate the DB volumes and the Log volumes and then restore the latest backup copy of the DB.



All worked fine but then we started getting these error messages during migrations:



Invalid frame detected. Expected magic 53454652 sequence number 00000001 server id 00000000 segment id 0000000070740958. (PROCESS: 225)

2005-10-13 13:46 ANR1330E The server has detected possible corruption in an object being restored or moved. The actual values for the incorrect frame are: magic 5B3855D8 hdr version 39016 hdr length 50810 sequence number 1192815965 data length C63B859C server id -1836211519 segment id 10321090521601071163 crc 7AD892CD. (PROCESS: 225)

2005-10-13 13:46 ANR1331E Invalid frame detected. Expected magic 53454652 sequence number 00000001 server id 00000000 segment id 0000000070740960. (PROCESS: 225)

2005-10-13 13:46 ANR1330E The server has detected possible corruption in an object being restored or moved. The actual values for the incorrect frame are: magic 6E958C06 hdr version 11501 hdr length 10898 sequence number 1124953744 data length 64EA0519 server id 682445064 segment id 9149496896754115484 crc 0CD2CD9A. (PROCESS: 225)

2005-10-13 13:46 ANR1331E Invalid frame detected. Expected magic 53454652 sequence number 00000001 server id 00000000 segment id 0000000070740962. (PROCESS: 225)

2005-10-13 13:46 ANR1330E The server has detected possible corruption in an object being restored or moved. The actual values for the incorrect frame are: magic C02DEF52 hdr version 5623 hdr length 28391 sequence number -1281831699 data length E76813FA server id -708628760 segment id 18086761016193466693 crc EE3E0565



At this stage we recreated all diskpools with data in it which was yet to be migrated because of the TSM server crash. The discard option was set to YES. This was done because we thought that some of the data may have been migrated prior to the server crash and because we had restored a DB previous to those migrated data, no reference was available in the DB.



Now the same kind of messages are appearing during the reclaim process in the copypool.



Can someone suggest how to tackle this.



We have TSM 5.2.3.3 server

 

[sgabriel62]

Sounds like you have a few things happening (1) bad media (2) corrupted DB on disk or tape (3) bad disk.

I would T-shoot hardware first, then move onto auditing the volumes with the repair option set to yes. Perform the audit against all erroring volumes.

 

[Harry_Redl]

Hi,

I have seen these messages when I have restored the DB and there was a change in the volume content after the DB was backed up (data were migrated, reclamation occured and the tape was reused ....)

It seems to me you have this problem

IMHO you have to accept you have lost a few files - to get rid of these messages you have to audit your volumes with fix=yes to have the DB corresponding with the volume content.

Using the REUSEDELAY parameter in your tapepool definitions can prevent this thing to happen in the future.



Hope it helps



Harry

 

[asteng5000]

I think the best way forward is to run an audit the volumes with fix=yes. And yeah- I've resingned to the fact that some backups have been lost and the management is OK with it



As suggested by sgabriel62 I've checked and rechecked the server hardware thoroughly and all diagnostics reports come out clean.

 

[asteng5000]

Thanks for the suggestions..I identified the volumes which were reporting currupt data(instead of running the audit as we have around 120 tapes in the copypool). An attempted move of data to the lto pool to which the tape was associated. after the data move the culprit volumes were deleted and returnrd to scratch state.



Not happy with loosing the backed up data but hey that's life!!