On 13 Sep 2007 at 16:46, Kern Sibbald wrote:
> On Thursday 13 September 2007 15:15, Dan Langille wrote:
> > On 12 Sep 2007 at 22:01, Dan Langille wrote:
> > > I'm just documenting this test failure. It's is more of a freeze,
> > > than a failure, but I have seen this issue before. Waiting for an
> > > appendable volume.
> > >
> > > On the console, I see:
> > >
> > > === Starting 2drive-incremental-2disk at 20:40:06 ===
> > >
> > > That's about an hour ago. The test should take about 6 minutes.
> > >
> > > I am going to control-c this, and run that given test about 200 times
> > > overnight, with debug on.
> >
> > This cycle of tests failed rather early. There are no core files.
> >
> > ===== 2drive-incremental-2disk OK 22:11:38 =====
> > ===== 2drive-incremental-2disk OK 22:12:26 =====
> > ===== 2drive-incremental-2disk OK 22:23:27 =====
> > ===== 2drive-incremental-2disk OK 22:24:16 =====
> > ===== 2drive-incremental-2disk OK 22:35:17 =====
> > ===== 2drive-incremental-2disk OK 22:36:05 =====
> > !!!!! 2drive-incremental-2disk failed!!! !!!!!
> > !!!!! Restored files differ !!!!!
> > ===== 2drive-incremental-2disk OK 22:57:59 =====
> > ===== 2drive-incremental-2disk OK 23:11:04 =====
> > ===== 2drive-incremental-2disk OK 23:27:24 =====
> > ===== 2drive-incremental-2disk OK 23:43:52 =====
> > ===== 2drive-incremental-2disk OK 23:56:28 =====
> >
> > Of note:
> >
> > 12-Sep 22:22 localhost-fd: NightlySave.2007-09-12_22.12.48 Fatal
> > error: Authorization key rejected by Storage daemon.
> > Please see http://www.bacula.org/rel-
> > manual/faq.html#AuthorizationErrors for help.
> > 12-Sep 22:23 localhost-fd: NightlySave.2007-09-12_22.12.48 Fatal
> > error: Failed to authenticate Storage daemon.
> > 12-Sep 22:23 localhost-dir: NightlySave.2007-09-12_22.12.48 Fatal
> > error: Socket error on Storage command: ERR=Broken pipe
> > 12-Sep 22:23 localhost-dir: NightlySave.2007-09-12_22.12.48 Error:
> > Bacula localhost-dir 2.3.4 (12Sep07): 12-Sep-2007 22:23:00
> >
> > If you look at the test.out results above, there is no failure/error
> > that corresponds to the above times. That is, I think "2drive-
> > incremental-2disk OK 22:23:27" is a false positive. It should report
> > a failure.
> >
> > This "Fatal error: Authorization key rejected by Storage daemon."
> > failure occurs 5 times during the above tests.
>
> Dan,
>
> This looks like some sort of bug to me. Perhaps it is a race condition or
> something that is not being properly cleaned up.
>
> I propose the following for running your tests:
Just so I get this right:
>
> 1. Set debug level -d100 on the command line of the SD.
from bin/bacula-ctl-sd
${BACSDBIN}/bacula-sd $2 ${OPTIONS} -v -c ${BACSDCFG}/bacula-sd.conf -
d100
> 2. Add a "setdebug level=0 storage=DDS-4" at the beginning of the
> test. I have done this and committed it.
--- 2drive-incremental-2disk (revision 5518)
+++ 2drive-incremental-2disk (working copy)
@@ -39,6 +39,7 @@
@output /dev/null
messages
@$out tmp/log1.out
+setdebug level=0 storage=DDS-4
label storage=DDS-4 volume=TestVolume001 slot=1 Pool=Default drive=0
label storage=DDS-4 volume=TestVolume002 slot=2 Pool=Default drive=1
status storage=DDS-4
> 3. Re-run the tests and repost it as you did.
>
> The above will give us more information about what is going on during the
> authentication process. It *may* help point us to a potential problem.
I'll start them running now.
--
Dan Langille - http://www.langille.org/
Available for hire: http://www.freebsddiary.org/dan_langille.php
-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Bacula-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bacula-devel
