problems setting up a linux master server

[sql_noob]

did some testing last week for moving our current NB from a Sparc/Solaris server to a x85_x64/Red Hat server

installed RHEL 5.3
hardware is a DL380 G5 with the SC44 something SAS HBA and a HP MSL 8096 LTO-4 tape library

RH found the tape drives and the SAS adapter

after installing NB 6.5 it found the robot and i added the drives from /dev/tape/by-id/ and the tape drive ID's

created a SG, inventoried the robot and dumped all the tapes into the scratch pool. set up a test policy to back up a db

it started the backup, but never mounted the media. said mounting media for 30 minutes or so and then failed with a status code 6. i saw the robot insert the tape into the drive, but the backup never started.

i'm thinking it could have been the driver for the SAS hba. it said mptsas and the HP version said mptlinux. i installed the HP rpm, rebooted but it still said mptsas.

Tested today with a Windows master/media server and everything worked perfectly

[FrSchind]

can you access the drives with the mt command? insert a tape into a drive and issue mt -f <device> status  - it should show you some information about the loaded tape. mt -f <device> rewoffl should rewind the tape and unload it.

[Nicolai]

You need to check the mapping between Netbackup and the physical location in the robot. As Frerk suggest the best way to do this is using the mt command.

A example of a mapping that will time out after 30 minutes :

Netbackup       Real local in robot

nst0                          nst2
nst1                          nst1
nst2                          nst0

Regards

Nicolai

[scorpy_582]

one more suggestion:

Try robtest

/usr/openv/volmgr/bin/robtest

select the robot

s s <--- status of slots
s d <--- status of drives
m s1 d1 <--- move media from slot 1 to drive 1
m d1 s1 <--- move media from drive 1 to slot 1

Check if any of these gives you an error.

Further.. Since media is getting mouted, checking bptm log will be a good idea

/usr/openv/netbackup/logs/bptm

[Yasuhisa_Ishika]

As guys said, logical configuration on NetBackup may be differ from physical configuration.
Easy way to correct this is to delete devices an reconfigure from Device Configuration Wizard.
To wizard detece devices correctly, update external_types before.
http://support.veritas.com/docs/320857

If you get failed after reconfigure, show following output

# /usr/openv/volmgr/bin/scan
# /usr/openv/volmgr/bin/tpconfig -d
# /usr/openv/volmgr/bin/vmoprcmd