Qlibv bring back no volumes

[heath156]

I have had problems with the SCSI ids on my server, for some reason they changed, after working with IBM i re-created the drive and drive paths and the library and library paths with the new SCSI id's. But now when i do a q libv nothing is showing in the library, any ideas as to how I can get this back, i've tried an audit library but this didn't work.



Any help would be appreciated



Thanks

 

[Harry_Redl]

Hi,



in THIS order (assuming you have barcode reader equipped library)



checkin libvol <libname> search=yes checklabel=barcode status=scratch

checkin libvol <libname> search=yes checklabel=barcode status=private



Hope it helps



Harry

 

[heada]

Well, heres the rub. You can check all those volumes back into the library as private, but there might have been some scratch volumes in there that will now be marked as private and you wont have access to. The other issue is if you are using library sharing (which I doubt with a SCSI library)



You can try this. It wont be pretty but it should get you back the private tapes and the scratches.



checkin libvol {libr} search=yes status=private checklabel=no



Then compare the list of all volumes in your storage pools to the list of libvolumes and any volumes that are libvolumes but not stgpool volumes, get put into a second list. The second list is compared to the drmedia list (for db-backups, etc) to create a third list. You can then take all the volumes in that third list and update the libvolume to be scratch.



update libvol {libr} {volume} status=scratch



Most of this can be automated with for loops in a OS-based script.



-Aaron

 

[heath156]

thanks, would it be possible to just enter all the information into the devcnfg.out file of where the volumes are within the library or would this not work?

 

[Harry_Redl]

Hi,



no, devconfig.out file won't help you - it is used for DB restore purposes only - when TSM server is up, it is not used - all date is read from DB.



When you do it as I suggested (checkin with "status=scratch" first) and TSM finds already used tape (member of stgpool, dbbackup etc.) it will display message like "volume XYZ123 cannot be assigned status scratch" - so in the first step only real scratch tapes are checked in

In the second step you check in the private ones

When you do it in oposite sequence (status=private first) then all tapes will be checked in as "private" and you are going to update them to scratch manually.



I did it this way several times and it works - just use that two commands in the right order



Hope it helps



Harry