NBU 126.96.36.199 eject tapes from Quantum i500 robot with 18 slot MAP via CLI
I am looking to automate my tape ejection from our Quantum i500 robot. Once a week, I eject, on average 90 tapes, for offsite storage.
Currently I use the web GUI for the robot and eject 18 at a time.
I've googled the commands, but everything I've found is 5+ years old.
I run a vmquery that gets me the Media ID, Barcode, and tape description.
From this output, I generate a flat txt file for the media ID, which I run through a script to generate the command to run:
/usr/openv/volmgr/bin/vmchange -res -multi_eject -w -verbose -rn 0 -rt tld -rh $RobotHost -ml $Media01:$Media02...$MediaNN
Questions I have:
01. Do I have to break up the -ml list to 18 or is the system smart enough to take all media IDs as input and pause after each 18?
02. The system times out after 5 minutes and reimports the ejected tapes from the MAP if I don't get to it fast enough. How do I get it to leave tapes in MAP indefinitely? Is it a combination of the arguments -verify_eject and -continue_ie? If so, how would the command look?
Thanks in advance,
You can only eject as many as the MAP/CAP can handle. Pull the tapes, then run the command for the next batch.
As far as timing out and re-importing, don't have that issue & don't use the extra parameters you mentioned.
And I have to ask why are you still on version 7.5? Way, way past time to upgrade.
90 tapes a week? Wow! Still using LTO2s?
If you can only choose 18 tapes for a manual eject because that is the size of the CAP/MAP, then you need to do the same in the script. So, you won't gain much trying to automate with the amount of tapes that you need to eject.
Maybe you should consider eject daily?
About the 5 min timeout - you can add MAP_CONTINUE_TIMEOUT parameter in vm.conf.
I agree. We need to upgrade. It comes down to a either we are staying with tapes or going virtual.
Averaging 90 per week on LTO4.
We are backing up OLD data that customers have not done anything with in a while.
One of my projects this year is to identify data that is beyond cold and have it cleaned up. That should reduce backup times and number of tapes.
Thanks for your reply!
Thanks for the reply.
Automating will be faster because I won't have to hunt in GUI for tapes to check off over multiple pages. It will speed the unloading considerably.
We pull tapes once a week for offsite storage. Robot is located at our colo and I am only there once a week, specifically for tapes.
Thanks for the tip on timeout. Set it on my robot host. Now we wait until Tuesday.
If that doesn't do it, try changing /usr/openv/volmgr/bin/vmchange -res -multi_eject -w to
/usr/openv/volmgr/bin/vmchange -res -multi_eject -sc
Forgot where I found the information but it works with our Quantum libraries.