We have recently discovered that our automated checkin script in perl
is failing because of the drive availability issues.

By default a checkin command uses checklabel=yes
which requires a mount of the volume.

I  personally use checkl=no
when I'm just moving a tape.  But when a tape returns from our
offsite (Arcus) (and its returning because it's been reclam'd and has no
'real' data on it) I like to check the label and be sure its not be
de-gaussed.

Therefore "stacked" checkin commands will sometimes fail if a process
requires a tape and pre-empt's it.  Its a sortof different mount queue
from the one for normal tape mount use.

... joe.f.

Joseph A Faracchio,  Systems Programmer, UC Berkeley


On Fri, 14 Jul 2000, Daniel Swan/TM wrote:

> Thank you everyone for your replies.  The problem has been solved, but the
> information you have all provided is invaluable to me
> in understanding other potential issues I should investigate, or may
> encounter, with ADSM.  I am keeping it on hand.
>
> The problem is that when tapes are being checked in, they are not being
> properly changed from vaultretrieve to onsiteretrieve
> status.  Oddly, it seems to do a few, but not all.  I'm still investigating
> this.
>
> I have recovered almost 100 tapes now with:
>
> move drmedia "*" wherestate=vaultretrieve tostate=onsiteretrieve
> checkin libv libname search=yes status=scr
>
> Thanks again folks,
> Dan.
>