• To add a jukebox, run jbconfig from the command line.
• When this fails, run (/etc/LGTOuscsi/)inquire, which will print the details of attached scsi devices.
• nsrim -X can also help.
• Command line tools: nsrwatch, nsrmm, nsrjb, savegrp
• mminfo -a -r state,volume,pool,%used,capacity,written

• Situtation: complete disaster recovery - only have the tapes, no server.
• Setup a new machine with the same hostname as the original backup server
• Install Legato Networker on a new machine
• Attach a tape drive and add it into Networker
• Find a tape with the bootstrap on it (You did record this BEFORE the disaster happened, didn’t you?) and load it into the drive
• Run scanner -B /tape/device/name and note down the Save Set ID and the volume label.
• Wait for the bootstrap information to be read from the tape
• Run mmrecov and answer the questions
• Restart Networker
• Run nsrck -L7 client.domain.name
• Run nwrecover and recover the data

Sometimes, depending on your setup, Networker will only mount the first tape to be read from in a recovery, and then sits there re-reading and re-mounting the tape, but not actually getting anywhere. The nwrecover process keeps reporting the Networker server as busy retrying every 30 seconds. The fix is to either manually mount the other tapes it needs (but isn’t actually waiting for in the ‘pending’ window), or to touch /nsr/debug/no_striped_recover, so it only reads from one tape at a time. I’m still investigating how to get Networker to do striped recovers.

nsrjb -l -n -vvv -S 1 -f rd=muttley.ebi.ac.uk:/dev/nst4
nsrjb -u -vvv -f rd=muttley.ebi.ac.uk:/dev/nst4
nsrjb -o recyclable -o manual -L
nsrmm -b recyclable -l -o recyclable -o manual
nsrclone -b "Maui Clone" -v
nsrjb -o recyclable -o manual -L -b "Maui Clone" 002025

Make sure that any local firewalls rules on the legato server are disabled.

After upgrading to Networker 7.3, you may get this error on some of your clients that are using RMAN backups.

nsrd: server info: Too many clients.  Maximum is 0 for client foo.bar.com

This error makes no sense, but the fix is to upgrade the networker client to 7.3 and the Oracle module to 4.50 (the latest versions on the 7.3 CDs).

Don’t upgrade to 7.3 unless you need to. nwadmin has been removed in 7.3 and the replacement java GUI is bloody awful. I still use nwadmin running from a client but it doesn’t quite work properly with 7.3 - you often can’t mount or unmount tapes. Stick with Networker 7.1.