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:
1. Set debug level -d100 on the command line of the SD.
2. Add a "setdebug level=0 storage=DDS-4" at the beginning of the
test. I have done this and committed it.
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.
Regards,
Kern
>
> The full script output of the file is here:
>
> http://www.langille.org/tmp/2drive-incremental-2disk.output.txt
> (464KB)
>
>
>
> -------------------------------------------------------------------------
> 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
-------------------------------------------------------------------------
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
