Need some clarity, Veritas 8.5

[88bw]

First-system topology, NT Domain, 4 servers(2K) + PDC(NT-4), I have veritas 8.5 running on a W2K server. here is the issue, I have 2 Exabyte EZ-17 autoloaders, each with a 7 tape magazine. SCSI id's are good, veritas sees them and will write to them. When adding the 2nd drive I decided to balance the load by having one machine do differentials(each night, directed to one autoloader) and the other doing Fulls(each weekend, 2 week rotation). So there are 4 sets of tapes. c,d,e,f. e & f are for the weekend, c & d, weekedays. the backup jobs are "server01-diff" etc, (1 tape per server except 03 which needs 3)and "server01-full" etc.(again, one tape per server, except 03) the tapes are allocated(14 each) to each set of jobs. But when a job runs, the wrong tapes are being pulled!!!! I have no idea how to resolve this, I have deleted all jobs and started all over again 3 times, and yet when I run a job, some random tape gets pulled. I have inventoried, shows correct.
*Sorry for the long post, tried to give enough details*

"peace is merely the pause between wars"

 

[packdragon]

The problem is that Backup Exec doesn't care what the label says, it grabs tapes from the media set that is specified in the job. If all your tapes are members of the same media set, then they will be pulled randomly. If you want to stick with your current setup, then you will need to...

1. Create a media for each job that needs its own tape.

2. Drag the tapes into their respective media sets.

3. Set each job to use the corresponding media set.

Hopefully I've understood the problem correctly. Good luck!

- Zoe, that's ZOH-EEE, get it right please
- Just a little ol' MCP at Solien Technology
-

http://www.solien.com

 

[88bw]

Actually I should have killed this thread 2 minutes after I wrote it......the problem was that with two machines you must designate which machine is perform which task. Drop down menu on the job creation screen. DUH!!!!

Simple oversight due to the eternal search for the greater problem.