vbr.py failes when disk containers are different during full backup/ restore

When doing full backup and restore using vbr.py, Vertica reports a problem in catalog bootstrap if the disk locations used in the backup server are not the same as that in the restore server. We had a secondary disk added using ADD_LOCATION in the database that was backed up. The IP address and the node name are the same between the restore and backup servers, only the disk structure is different. This problem was fixed by making the disk containers identical, but that severely limits the usage of the backup function. I couldn't find this requirement to have the same disk structure in the documentation as well.


  • Options
    Hi Sajan, thanks for bringing this problem to our attention! We have made our support team aware.
  • Options
    Thanks Danielle
  • Options
    The vbr.py utility requires that the backup server disk locations are identical to the restore server.  If you add a storage location after creating a backup, restore will fail, because vbr rebuilds the catalog from the backup, which does not include the new location.  To restore successfully from a backup, do not change either the backup or restore servers.       
    A documentation enhancement request has been entered to clarify this.
  • Options
    Thanks for the clarification Amy.

Leave a Comment

BoldItalicStrikethroughOrdered listUnordered list
Align leftAlign centerAlign rightToggle HTML viewToggle full pageToggle lights
Drop image/file