Strange potential issue with BAB11.5SP2 SAN option

[FrogEater]

Hello,

My config is as follow:
1 primary server / 4 distributed, a HP MSL6030 30 slot shared over FC.

On teo of the distibuted server, the backup behave as if they aren't waiting for the lib to get free up.

In the log on "working" server i see from the time the job should start line saying "Job can not be start" then "Check device availablity"
Then the job starts properly once the library has been released by other machines.

In the log of "non-working" server I have sequence as follow:
- Jon cab not be started
- Check device availabilty
- then the job move from Ready to Active
- then I have a failure to connect to Media.
- Waiting by +/-30min+
- then job start on a blamk tape, discarded the one it should use.

so, instead of using one tape for all my job (as it fits), I use two (while I still have 100% backup anyway)

HAs anyone ever seen that before ?

Thanks,
Chris

 

[vschumpy]

OS level drivers for the tape drives need to be disabled on all servers participating in the SAN - seen this hundreds of times.

 

[FrogEater]

Hello,

OS drivers have been installed (the devices - tape lib & tape drive- are flagged "Unknown" with yellow exclamation mark in the device manager).

Chris

 

[vschumpy]

Disable the devices anyway, then re-run device configuration - using auto config on the SAN primary.

 

[FrogEater]

Apparently, I've fixed the problem by deleting all the jobs and re-creating them locally on each distributed server (I did create them connecting remotly through the Arcserve Manager). Since I've done that, device management works well.

I though about this solution when I saw the dist. server backing up \\<host>\Drive$ instaed of the local ID of the drive.

Thanks for your help !

Chris