Re: Preventive Maintenance Question

[Jack Kingsley <iseriesflorida@xxxxxxxxx>]

Up to date on the tape drive firmware.

On Thu, May 13, 2010 at 11:19 PM, Kirk Goins <kirkgoins@xxxxxxxxx> wrote:

> Jerry, I recently had an error similar to yours, In my case a SLR60 drive
> and 2 pairs of 4326 ( 35GB ) drives mirrored, The system had been running
> without errors for months. I as hired to do an OS upgrade and I wanted a
> full system save. This 520 was a web server so it was pretty static and the
> client had not been doing backups. I had some tape interface errors just
> rying to init the tape, but did get a clean backup. I did notice after the
> backup both mirrored pairs had a failed unit. Since the internal tape drive
> and the disks share the same controller ( and SCSI Bus ), We had IBM
> repleace the tape drive. NO more errors. Rebuilt the mirrored drives did
> the
> OS upgrade via Images catalog and no errors. I can only assume the tape
> drive was the problem.
>
> You may be seeing the same type of problem, since the errors are when the
> tape is in use.
>
> On Thu, May 13, 2010 at 1:30 PM, Jerry Adams <Jerry@xxxxxxxxxxxxxxx>
> wrote:
>
> > For the second time in less than a month we had a critical system
> condition
> > last night.  Both times it occurred during our nightly backup procedure.
> >  Errors were reported for the 6384 (cartridge) tape drive and three or
> four
> > of the disk drives (4326's) of which we only have four [4].  To me it is
> > sort of funny (not in a hilarious way mind you) that the same scenario
> > occurred (i.e., trying to use tape drive, disk errors reported).
> >
> > The tape drive's error SRC was 63A09100 (a tape unit failed).  The disk
> > drive SRC's were 43263109 (May indicate a disk unit failure) and a couple
> of
> > 57099030 (which I can't find).  The tape drive went into FAILED status on
> > the WRKCFGSTS *DEV display.  The disk drives (three [3] last night, four
> [4]
> > on 20 April) never actually failed; that is, we were able to do SAVLIB's
> to
> > Save Files and FTP to a separate machine.
> >
> > Both times IBM replaced one of the disk drives (we use RAID5), and the
> > system rebuilt the disks.  Speculation (my term) is that the tape drive,
> > which is on the same bus as the disk drives (or so I was told), went into
> a
> > FAILED status because the disks were sending so many errors that the bus
> got
> > clogged up and ergo the tape drive timed out (or something along those
> > lines; I wasn't taking very good notes at the time).  In neither event
> was
> > anything done to the tape drive, except to vary it off/on (i.e., no
> repairs
> > of any kind).
> >
> > Service Tools' Product Activity Log (PAL) is also showing an error
> > (B6005120) on a 2844 device which, I think, is the RAID card (anyway, it
> is
> > a card in our 520), though WRKPRB does not list it.  There were five [5]
> of
> > these all with the same timestamp about one [1] minute before PAL shows
> the
> > 43263109 errors on the disk units.
> >
> > I know that disks fail; been there, done that, got the souvenir.  But I
> > (and my boss) was wondering if it could be something less obvious causing
> > the problem.  More to the point something that could be diagnosed using
> some
> > kind of hardware diagnostics tools either in our tool kit (iNav, SST, or
> > something to which we have access), or in the CE's tool set.  I'm a firm
> > believer in "An ounce of prevention..." adage, not the "If it ain't
> > broke..." one.
> >
> >
> > Jerry C. Adams
> > IBM System i Programmer/Analyst
> > --
> > B&W Wholesale
> > office: 615-995-7024
> > email:  jerry@xxxxxxxxxxxxxxx
> >
> > When we played, World Series checks meant something.  Now all they do is
> > screw up your taxes. -Don Drysdale
> > --
> > This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing
> list
> > To post a message email: MIDRANGE-L@xxxxxxxxxxxx
> > To subscribe, unsubscribe, or change list options,
> > visit: http://lists.midrange.com/mailman/listinfo/midrange-l
> > or email: MIDRANGE-L-request@xxxxxxxxxxxx
> > Before posting, please take a moment to review the archives
> > at http://archive.midrange.com/midrange-l.
>
>
> --
> Kirk
>  --
> This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
> To post a message email: MIDRANGE-L@xxxxxxxxxxxx
> To subscribe, unsubscribe, or change list options,
> visit: http://lists.midrange.com/mailman/listinfo/midrange-l
> or email: MIDRANGE-L-request@xxxxxxxxxxxx
> Before posting, please take a moment to review the archives
> at http://archive.midrange.com/midrange-l.