We use a different volume id for each tape we use.
When we initialize the tape before backup, we use the volume-id that is on
the sticky label on the case the tape is kept in.
We can then go into SST to view history of all errors by tape volume-id, to
find out if one particular tape is having more trouble than others. SST
stores error logs on all devices, forever, until a CE or someone erases the
older stuff.
Depending on how the backup is launched, you can specify what table of
automatic error message responses are to be taken ... the system's default,
or one you crafted for this particular job. Then if you get a particular
error message, the job will take the response you wanted it to take.
WRKPRB charts most recent errors any hardware connections, for as long as
you have set in GO CLEANUP (30 days minimum). If you have QSYSMSG setrup,
then it gets a copy of the most serious error messages. I suggest setting
that up, and checking it regularly. Last month we had a hard disk crash,
and I did not know it for 2 days, because I was too busy to check that
stuff every day. As soon as I saw the message, I e-mailed tech support ...
late one nite ... they showed up & fixed the problem next morning.
You can configure your system to automatically notify your tech support of
serious problems.
The tape drive is cleaned a certain way. After a mandatory cleaning (the
orange light came on, or we have backed up 10 tapes since the last time it
was cleaned), we stick the cleaning tape in a second time & clean again
(better to be more clean than needed, than not clean enough ... and the
tape drive lasts longer).
Depending on how long you keep your general system messages
(GO CLEANUP settings) you can DSPLOG F4 from *BEGIN available data F10 next
screen & specify some message or messages you want to see how often THAT
message has happened in the history. You can also write a CL to do this
for up to 10 different related messages.
Our LOG shows what got backed up, where there were problems.
Different types of messages get copied to different kinds of message
queues, which you can view via DSPMSG QSYSOPR or whatever. I suspect all
the tape error messages go one place, but what error message queue will
depend on your system setup.
Depending on what all gets saved (only changed objects, or weekly all
objects), it could be that one backup needs a multi-volume tape set bigger
than the one that got prepared.
You could use a display library objects info to *OUTFILE & query which have
not been backed up in a reasonable time period (where the multi-volume set
ran out), or GO BACKUP, change what to be backed up there, just for the
purpose of scrolling thru list of libraries charting when last backed up,
if changed since that backup, to see if any are seriously missing a recent
backup.
I have been seriously pre-occupied, even to follow threads that interest
me, so sorry if my answer duplicates already answered
I know this has been addressed but I am not using the correct search words.
What command can I use to see how many and what tape errors were generated
during a save? Like write errors.
I have a client whose Save Changed Object Save gets CPA4086 (Device not
ready or next volume was not loaded) once a week.
All production files are touched daily and no massive amounts are added by
Thursday. This does not happen on Friday.
I suspect the tapes used on Thursday.
Thank you,
Bonnie Lokenvitz
midrange.com
