ndmp error 800

[jimtheone]

Our Netapp filer connects directly to robot for NDMP backup.
when robot gets reboot, Netbackup fails with error 800 resource request failed. Backup job runs and gets as far as mounting the tape in the job details, but it never mounts the tape and errors out with NDMP backup failed, path=UNKNOWN.
If we reboot the Netapp Filer, then backup will finish successfully. Is there a way to avoid rebooting Netapp Filer?

Thanks,

 

[just2confused]

Probably a Netapp bug. You don't mention what version of ONTap you are running....but have you tried rescanning the bug on the NetApp instead of rebooting? Determine the adapter number then do:

storage disable adapter <adapter#>

then re enable using:

storage enable adapater <adapter#>

You can only do this is you have redundant adapters.

 

[xmsre]

When you reboot the robot, at a minimum scsi busreset is set and it makes sense that ndmpd would get confused. Try

ndmpd off
ndmpd killall
ndmpd on


instead of rebooting the filer.

 

[dgallegos]

Also make sure that if you have snapshots turned on, that the netapp has enough space for the back up. you can also view whats the commands are doing in verbose mode on the netapp if you log into it via puddy. let me know if you need a doc that covers this.

 

[wileywombat]

When the issue happens again, go to the filer and issue a sysconfig -t, to list the tape drives. Ensure that this matches what is defined in NBU.

Also you can manually test the drive path by loading a scratch tape and issuing the commands "mt -f nrst*a status" and then dump to it using "dump uf0 nrst*a /vol/vol0