Hello,
On 03.11.2005 21:51, Vadim A. Umanski wrote:
DIR dies some time in the night...
That's how it died on cgatex job last time.
02-Nov 03:21 nfs4p-dir: Begin pruning Jobs.
02-Nov 03:21 nfs4p-dir: No Jobs found to prune.
02-Nov 03:21 nfs4p-dir: Begin pruning Files.
02-Nov 03:21 nfs4p-dir: No Files found to prune.
02-Nov 03:21 nfs4p-dir: End auto prune.
02-Nov 07:05 nfs4p-dir: Start Backup JobId 3961,
Job=cgatex-full.2005-11-02_07.05.00
02-Nov 07:05 cgatex-fd-fd: Since time adjusted by 0 seconds.
02-Nov 07:05 s10-sd: Volume "Vol0086" previously written, moving to end of data.
02-Nov 07:06 s10-sd: User defined maximum volume capacity 734,003,200 exceeded
on device /d/0/bacula.
02-Nov 07:06 s10-sd: End of medium on Volume "Vol0086" Bytes=733,941,548
Blocks=11,378 at 02-Nov-2005 07:06.
02-Nov 07:06 nfs4p-dir: Recycled volume "Vol0087"
...
...
02-Nov 07:35 s10-sd: Recycled volume "Vol0091" on device "/d/0/bacula", all
previous data lost.
02-Nov 07:35 s10-sd: New volume "Vol0091" mounted on device /d/0/bacula at
02-Nov-2005 07:35.
02-Nov 08:04 s10-sd: User defined maximum volume capacity 734,003,200 exceeded
on device /d/0/bacula.
02-Nov 08:04 s10-sd: End of medium on Volume "Vol0091" Bytes=733,952,897
Blocks=11,377 at 02-Nov-2005 08:04.
03-Nov 01:05 nfs4p-dir: Start Backup JobId 3962, Job=nfs4p.2005-11-03_01.05.00
03-Nov 01:05 nfs4p-fd: Since time adjusted by -1095 seconds.
I don't understand the last string...
That simply indicates that bacula detected a time difference between the
DIR and the FD it just backs up. In that case, bacula adjusts the time
stamps it creates so that, when later referring to this job, all parties
talk about the same system time.
...
AL> It looks like that job hasn't failed but got cancelled - hat status
AL> should, as far as I know, only happen as a direct result of user
AL> intervention.
No one but me could intervent. I did not. There was no manual
cancel... it would be too simple...
I wonder ... when DIR falls, what's going on then...
Interesting. It would be interesting to find out in which cases the
status is set to canceled. I've never seen a cancel except when, well,
someone used the cancel command. An error status is much easier to
produce :-) and when the DIR crahes I usually find the status Running in
the catalog and nothing new in the status output. Like today after a
power failure :-( And of course I've not got an UPS installed myself, I
only recommend them to customers :-|
...
AL> Well, start with the debug log and probably the debugger. That should
AL> help understanding what happens when Bacula crashes. Or upgrade to 1.38
AL> and see if that fixes your problem (which might easily happen). The
AL> upgrade itself is not a problem as long as you know how your installed
AL> version was built (options to configure)
I'll probably have to investigate it.
Good luck - that can be a difficult jobs if there are no records. On the
other hand, if Bacula was installed from a package or you find a shell
script to configure, make and install it's easy...
AL> and have the necessary toolchain and libraries installed. The
AL> catalog upgrade can be a problem as you can not easily revert to
AL> an older version...
That matters.
Note that I didn't write impossible... and, of course, you could keep a
backup copy of your existing catalog and would only have problems with
the volumes written between upgrade and restore of the catalog backup.
Arno
--
IT-Service Lehmann [EMAIL PROTECTED]
Arno Lehmann http://www.its-lehmann.de
-------------------------------------------------------
SF.Net email is sponsored by:
Tame your development challenges with Apache's Geronimo App Server. Download
it for free - -and be entered to win a 42" plasma tv or your very own
Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php
_______________________________________________
Bacula-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bacula-users
