In high active datazone, it can happen that nsrd forget to update
jukebox map (fields loaded barcode, loaded volume and loaded slot)
while unloading volumes. Most of the time, it's done when multiple
nsrjb are running concurrently, the changes done by the first nsrjb not
being commited yet into config files and in memory jukebox maps.

Jukebox map not being updated, networker still think a tape to be loaded
in the device and try to unload it. But reality is such no tape to be
loaded.

Result is nsrmmd fails with an opening I/O error. Then nsrjb tries, during
20 minutes to unload the tape. After that, it reports an error and mark
the slot as unloaded. (In some time, you can also lose the inventory of
the slot containing the ghost volume before second nsrjb).

I never found a way to reduce these 20 minutes to something more
acceptable but am working with it. It has nothing to do with load/unload
sleep, these values just being the time networker waits between jukebox
PUT to device or UNLOAD from device before opening the OS driver
(/dev/rmt/...., \\.\tapex)

HTH,

Th


> I have a similar problem on an ATL-P7000 tape library.  However I *am*
> still on a Sun server platform.  Does anyone know what the unload sleep
> for that type of library should be?  Mine is set at "5".
>
> I get these types of errors almost every weekend:
>
> 02/05/05 06:38:22 nsrd: /dev/rmt/45cbn Eject operation in progress
> 02/05/05 06:42:33 nsrd: media warning: /dev/rmt/45cbn opening: I/O error
> 02/05/05 06:42:53 nsrd: media info: unload error for jukebox `ATL-P7000'
> detected.  Retrying
> 02/05/05 06:42:53 nsrd: media info: unload retry for jukebox
> `ATL-P7000': sleeping 30 seconds
> 02/05/05 06:43:38 nsrd: media info: unload retry for jukebox `ATL-P7000'
> failed - will retry again.
> 02/05/05 06:43:38 nsrd: media info: unload retry for jukebox
> `ATL-P7000': sleeping 30 seconds
> <snip>
>
> This continues until the drive goes into service mode.  Is this possibly
> related to the unload sleep?  Or is there something more wrong here?
>
> Thanks,
> Brian
>
> --
> Note: To sign off this list, send a "signoff networker" command via email
> to listserv AT listserv.temple DOT edu or visit the list's Web site at
> http://listserv.temple.edu/archives/networker.html where you can
> also view and post messages to the list. Questions regarding this list
> should be sent to stan AT temple DOT edu
> =*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
>

--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listserv.temple DOT edu or visit the list's Web site at
http://listserv.temple.edu/archives/networker.html where you can
also view and post messages to the list. Questions regarding this list
should be sent to stan AT temple DOT edu
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=