We spend buckets of money to do this NDMP thing and

the UNIX community is rather disgusted with the restore times.

Our we doing something wrong, or is this a typical restore scenario.


NDMP filer directly connected to an AIT-2 Library.

UNIX master Server with NDMP extensions.

Restore from the following Full Backup

 Media ID N00019 58,083.85 MB
 Media ID N00020 16,948.29 MB
 # of Files   1,371,186.00


Restoring one lousy file resulted in the following :


Restore started Thu May 10 17:00:04 2001

17:00:37 (151227.xxx) Media id N00020 is needed for the restore.
17:00:37 (151227.xxx) Media id N00019 is needed for the restore.

17:00:38 Restoring from image created Fri May 04 18:10:05 2001
17:00:40 INF - Waiting for mount of media id N00019 on server xyz.
17:01:48 INF - Waiting for positioning of media id N00019 on server xyz.
17:02:18 INF - Beginning restore from server xyz to client ndmp-hawg.
17:02:46 INF - Restoring NDMP files from /vol/vol0 to
/vol/vol0/home/restores.
19:06:27 INF - Waiting for mount of media id N00020 on server xyz.
19:09:22 INF - Waiting for positioning of media id N00020 on server xyz.
19:09:23 INF - Continuing restore from server xyz to client ndmp-hawg.
19:45:21 /vol/vol0/home/onelousyfile.htm
19:45:22 Status of restore from image created Fri May 04 18:10:05 2001 =
the requested operation was successfully completed

19:45:24 (151227.xxx) INF - Status = the requested operation was
successfully completed.

UNIX admin quote:

NDMP = Need to Develop More Patience

UNIX admin endquote


Do the other NDMP backup products share this issue ?


Thanks,


Greg Guenther
NT Admin with the bright idea to do NDMP backups