Re: [Networker] Drive unable to unload
2005-02-07 12:38:26
Matt,
We have had a problem where networker thinks it it is still mounting
the tape or "mount in progress". The drive and tape are completely idle
and can be physically load/unloaded. The drive appears stuck in this
state to Networker. Have you had any luck with the cdi commands with a
problem like this ?
Robert Maiello
Pioneer Data Systems
On Mon, 7 Feb 2005 10:38:26 -0500, Matt Temple
<mht AT RESEARCH.DFCI.HARVARD DOT EDU> wrote:
>Somen Chanda wrote:
>
>>Hi,
>>
>>We moved from Sun to Linux and facing a lot of problems. This is one of
>>them. We get this error very frequently.
>>
>> media info: failed unloading drive `rd=backup1:/dev/nst1' to slot '12',
>>error '25'
>>
>>Then the drives will be disabled by legato unless we notice and run mt
>>command to unload the tape manually. I use
>>
>># mt -f /dev/nst0 unlock
>># mt -f /dev/nst0 rewind
>># mt -f /dev/nst0 offline
>># mt -f /dev/nst0 eject
>>
>>We have a ADIC Scalar100 with IBM Ultrium 3580 drives in them. Our backups
>>are failing because of these errors and the tapes are bound to get stuck in
>>the drives every night.
>>
>>Can anyone help?
>>
>>Thanks in advance.
>>
>>Somen Chanda
>>
>>
>Somen,
>
> I, too have had similar problems which I've written about aq nyumber
>of times.
>If you look at the most recent Linux release notes (I think it wasi n
>the release notes),
>you'll see that they have a further recommendation for the stinit.def
>file. They suggest that
>you add a line that looks like the following:
> "auto-lock=0"
>Actually in the release notes, they have it as
>"auto_lock" (with an underscore) which is incorrect.
>We have AIT drives, and I wonder if this might not be relevant to your
>setup.
>Over the last 4 years here, I've had many, many problems with drive
>settings
>(variable block size) reverting to fixed, tapes not unloading, etc.
>
>I've also found that many of the /usr/sbin/cdi* utilities have been
>useful for
>workarounds, especially /usr/sbin/cdi_load_unload. But believe it or
>not, one of our
>problems is only fixed by forcing a tape into a drive by front-panel
>commands on the library,
>then running "scanner -v" on the tape. Many weirdly-spent hours of
>experimenting led to
>this discovery!
>
>
> Matt Temple
>
>--
>=============================================================
>Matthew Temple Tel: 617/632-2597
>Director, Research Computing Fax: 617/582-7820
>Dana-Farber Cancer Institute mht AT research.dfci.harvard DOT edu
>44 Binney Street, LG300/300 http://research.dfci.harvard.edu
>Boston, MA 02115 Choice is the Choice!
>
>--
>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
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
|
|
|