Unsolved
This post is more than 5 years old
3 Posts
0
803
Ejecting tapes to cap
- STK L700e Library w/ networker 7.3.2
I'm taking over our sites backups and getting up to speed on Legato. Every Monday an email is sent out listing the tapes that are to be ejected from the robot.
First the operator determines if the tape is in a slot or a drive: nsrjb -C
If the tape is in a drive then the operator moves it to a slot: nsrjb -u volser
else the tape is ejected to the cap: nsrjb -w volser
Is it possible to move a tape directly from the drive to the cap?
Has anyone automated this process with a script?
I'm taking over our sites backups and getting up to speed on Legato. Every Monday an email is sent out listing the tapes that are to be ejected from the robot.
First the operator determines if the tape is in a slot or a drive: nsrjb -C
If the tape is in a drive then the operator moves it to a slot: nsrjb -u volser
else the tape is ejected to the cap: nsrjb -w volser
Is it possible to move a tape directly from the drive to the cap?
Has anyone automated this process with a script?
ble1
2 Intern
2 Intern
•
14.3K Posts
1
August 13th, 2007 14:00
pgerke
35 Posts
1
August 13th, 2007 15:00
I have provided my operators a Unix script to do some of the things you are after. First off, I run a plain nsrjb -u, with no other options, which unloads all of the drives at the same time. Then the withdraw is just nsrjb -w vol1 vol2 vol3... depending on how many volumes the operator requested.
If your Monday e-mail is generated by a script on your backup server, then you might consider modifying that for the withdraw scipt, but keep in mind that once a tape is withdrawn into the cap, it disappears from NetWorker. One time they were looking everywhere for a missing tape, only to find it the next time they opened the CAP. I keep thinking I need to update my script with an mmlocate for the withdrawn tapes, to give them some sort of location once they are withdrawn, but have not yet done so.
I hope this is useful.
Good Luck,
Paul
Petek4
3 Posts
0
August 27th, 2007 16:00
Today I ran the "nsrjb -u" command. I was a little concerned since all 6 of my tape drives had tapes in them and two were actively writing ... couldn't really discern from the man page if running "nsrjb -u" would differentiate from tapes being written to vs. tapes left in a drive when done. The experienced indicated that it leaves the tapes being written to alone.
But one of the tapes wouldn't eject.
From the GUI the status for the tape shows ""ready for writing, idle". After checking google I decided it was a hung nsrmmd process which I tracked via fuser and killed. I was then able to eject the tape. Not sure why this happen but I'm told it happens a lot here. Guess I'll be opening a call with Legato.
networker-pkain:~ ==> fuser /dev/rmt/1cbn
/dev/rmt/1cbn: 11994o
networker-pkain:~ ==> ps -ef | grep 11994
root 11994 11785 0 Aug 19 ? 258:12 /usr/sbin/nsrmmd -n 11
networker-pkain:~ ==> sudo kill -TERM 11993
Password:
networker-pkain:~ ==> ps -ef | grep 11993
pkain 5542 5231 0 09:22:50 pts/4 0:00 grep 11993
Part of our process for ejecting a tape is to run a for loop for mmlocate. Seems like the mmlocate doesn't take multiple volser's as an option so I just run a for loop to move them to site "offsite".
Thanks for the input.