On Mon, 26 Feb 2007, Kern Sibbald wrote:
> On Sunday 25 February 2007 09:42, Michael Brennen wrote:
>> Hello again,
>>
>> The long weekend backups started running Friday night. This is a 4
>> drive 60 slot autochanger. The bacula version is "archive-dir
>> Version: 2.1.4 (21 February 2007)" to get the encryption patch.
>>
>> Consistent with what I posted a couple of days ago the tape drives
>> are still a bit confused, and I've had to manually unmount and mount
>> tapes to accomodate the needs of a few jobs.
>>
>> I now have a job that is waiting for a particular mount on drive 0,
>> but thd drive is unmountable. I unmounted a volume FNI0005 in drive
>> 0, used by a different job, and unmounted volume FNI0006 in drive 3
>> in order to mount FNI0006 on drive 0, which is what the waiting job
>> wanted. That process left drive 0 in the following state, taken
>> with debug 200:
>>
>> ===============================================
>> Device "Drive-0" (/dev/nst0) is not open.
>> Device is BLOCKED. User unmounted during wait for media/mount.
>> Drive 0 status unknown.
>> Configured device capabilities:
>> EOF BSR BSF FSR FSF EOM REM !RACCESS AUTOMOUNT !LABEL !ANONVOLS ALWAYSOPEN
>> Device state:
>> !OPENED TAPE !LABEL !MALLOC !APPEND !READ !EOT !WEOT !EOF !NEXTVOL !SHORT
>> !MOUNTED num_writers=1 block=5
>> ===============================================
>>
>> If I read this correctly, the lack of a bang in front of TAPE
>> indicates that bacula thinks there is a tape in the drive. There is
>> not; a raw mtx status query shows drive 0 empty. Trying to unmount
>> FNI0005 left the drive in this state, and it is now impossible to
>> mount another drive, as a mount command results in trying to unmount
>> a non-existent tape into an already full slot 1, which of course
>> fails.
>>
>> When I try to mount FNI0006 on drive 0, this is the result:
>>
>> ===============================================
>> forked processes:
>>
>> /bin/sh /path/to/bacula/mtx-changer /dev/sg3 unload 1 /dev/nst0 0
>> \_ mt -f /dev/nst0 offline
>> ... eventual timeout...
>> 3995 Bad autochanger "unload slot 1, drive 0": ERR=Child exited with
>> code 1
>> Results=/dev/nst0: No medium found
>> Storage Element 1 is Already Full
>>
>> 3901 open device failed: ERR=dev.c:424 Unable to open device
>> "Drive-0" (/dev/nst0): ERR=No medium found
>> ===============================================
>>
>> Every time I try to mount a tape to the drive it tries to unload a
>> non existent tape and will go no further. I've tried releasing the
>> drive, unmounting it, etc. and nothing has succeeded so far. I've
>> had no results googling for a resolution either; this drive status
>> does not come up much in searches.
>>
>> The autochanger and drive 0 config are below; the other drives are
>> identical with appropriate name changes. This configuration has
>> been working for weeks, and it just began showing problems. The
>> pools just now began recycling volumes, and it seems to be the low
>> numbered recycled tapes that are problematic. I do not know if that
>> has anything to do with this or not.
>>
>> Autochanger {
>> Name = Q47
>> Device = Drive-0, Drive-1, Drive-2, Drive-3
>> Changer Command = "/path/to/bacula/mtx-changer %c %o %S %a %d"
>> Changer Device = "/dev/sg3"
>> }
>>
>> Device {
>> Name = Drive-0
>> Drive Index = 0
>> Media Type = DLT-7000
>> Archive Device = /dev/nst0
>> AutomaticMount = yes;
>> AlwaysOpen = yes;
>> RemovableMedia = yes;
>> RandomAccess = no;
>> AutoChanger = yes
>> LabelMedia = no
>> # Enable the Alert command only if you have the mtx package loaded
>> Alert Command = "sh -c 'tapeinfo -f %c |grep TapeAlert|cat'"
>> SpoolDirectory = /some/dir
>> }
>>
>> If anyone has any insight on how to clear this up, or how it got in
>> this state, I would appreciate it. For now I will muddle through
>> the long backups and manually rerun the failed/blocked jobs after I
>> can restart things to get it to an unblocked state. Thanks...
>
> The first thing to do is to get back on a stable Bacula. I have not tested
> the trunk with a tape drive for well over a month, and yet there have been
> some *major* modifications to the device handling.
>
> If you don't know how to get onto Branch-2.0, then I recommend going back to
> the released 2.0.2. (Sorry I don't have the proper command just in front of
> me). Hopefully that will clear up all your problems. If you need the fix for
> the encryption problem, download the updated restore.c file from the bug.
> By the end of next week, I hope to have version 2.0.3 ready for release ...
I went back to 2.0.2 source with the filed restore.c, and all is
working well now. I did not know the difference between the branch
and trunk; this is my first exposure to an svn tree. That's what I
get for working with things I don't fully understand. :)
-- Michael
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
