We have just had a disk in a raid1 array fail on us but have recovered the data fine from the other disks in the array and are now happily working...

The disk that failed was also the location we backed up to before archiving backups to loads of other places. We are confident that this partition was not full, but backups from 1st and 3rd June both omitted the largest database [gbak produces a 6GB file which gzips to about 1GB] suggesting that gbak failed to complete this database. Regrettably our backup log file was also on the same drive so I've got no record of any errors!

We also have a backup from the 2nd June that will not restore.

gbak gives the error:

gbak: adjusting an invalid decompression length from 54 to 51
gbak: ERROR: invalid request handle
gbak: ERROR: gds_$send failed
gbak: Exiting before completion due to errors
Segmentation fault

Whilst the pressure is now off as the database is back on line I'd like to take this opportunity to work out why we are getting this error, and also understand why the backups on 1st/3rd failed completely but the 2nd seemed to create a backup fine.

Thanks

Ian