TSM not checking in new or empty old tapes

[mosiac]

I've run into an odd issue currently in my tsm environment.

I used up all my scratch over the weekend and I need to check in more so I went to add tapes to my ts3310 library and move them in like always and run the command to put tapes in and it only checked in 2 of the 12 I wanted to check in.

My first thought was perhaps the library was full but I have 8 free slots, so I thought I would just load some fresh tapes in instead of reuse scratch but those won't check in either. The commands complete successfully but nothing happens.

I've checked for processes sessions and requests to see if there is anything waiting for something else before inserting new tapes and I don't see anything.

Now I think it might be because all of my current LTO5 drives are full, I need one for checkin correct? If the arm can just handle the checkin process without a drive then I have no idea where to go from here.

Nothing in my environment has changed in the past couple of weeks so I'm not sure where else to look because I'm not seeing any errors when I attempt to checkin new volumes.

 

[marclant]

move them in like always and run the command to put tapes in and it only checked in 2 of the 12 I wanted to check in.

Did you check the activity log to see if there was an explanation?

I thought I would just load some fresh tapes in instead of reuse scratch but those won't check in either.

Were they labelled? Did you check the activity log?

Now I think it might be because all of my current LTO5 drives are full, I need one for checkin correct?

If that was the case, that would cause the command to fail.

 

[mosiac]

Act log states there aren't any volumes for checkin. They all have labels we get them from the company we purchase our tapes through that way.

I made sure they're going to the proper logical library before checkin as well so the only log info I see is

ANR8431I Checkin libvolume process completed for library loglib02; 0 volumes found.

Could there be a problem with my arm not reading the labels randomly?

 

[marclant]

They all have labels we get them from the company we purchase our tapes through that way.

Not the sticker, the internal label that TSM writes on the first few blocks of the actual medium when you issue the label libvolume command. You have to label new tapes: HELP LABEL LIBVOLUME

For the tapes you reused, are they still defined to a pool or have a reuse delay? Check "query volume [volumename] f=d"

What is the command you used to checkin?

 

[mosiac]

label libvolume LOGLIB02 search=bulk labelsource=barcode checkin=scratch WAITTIME=0

is our script for running new tapes in.

checkin libvolume loglib02 search=bulk checklabel=barcode status=scratch waittime=0

is the script that adds tapes that are being reused after being expired from backupsets, copypool, archives, etc.

 

[marclant]

label libvolume LOGLIB02 search=bulk labelsource=barcode checkin=scratch WAITTIME=0

Ok, so you are labeling the new tapes. Did you put the tapes in the I/O port before issuing the command?

checkin libvolume loglib02 search=bulk checklabel=barcode status=scratch waittime=0

For the tapes you reused, are they still defined to a pool or have a reuse delay? Check "query volume [volumename] f=d"

 

[mosiac]

Yeah our process is insert tapes into I/O station, wait for on screen button for logical library to be checked into, then after that run the command, since our tape library is on a different floor the time it takes to get back to the desk is considered plenty of time for the arm to look over the tapes and start the process.

 

[marclant]

Well, there has to be something. For the new tapes, recheck the activity log line by line from the moment you issue: label libvolume until the time you see the process completed.

For the tapes you are reusing, since it checked 2 out of 12, there's something with the others:

For the tapes you reused, are they still defined to a pool or have a reuse delay? Check "query volume [volumename] f=d"

When you issued the command, does the number of empty slots correspond to the number of tapes mounted?

 

[mosiac]

Found it. Loglib02 is now "full" and I actually got a failure from it this time.

Thanks,

My library still says it has 8 free slots but I can take a look at that and see what it thinks relative to what tivoli thinks.