Dan Langille wrote: > Dan Langille wrote: >> Eric Bollengier wrote: >>> Le Saturday 22 August 2009 13:57:51 Dan Langille, vous avez écrit : >>>> Eric Bollengier wrote: >>>>> Le Saturday 22 August 2009 13:48:06 Dan Langille, vous avez écrit : >>>>>> Eric Bollengier wrote: >>>>>>> Le Saturday 22 August 2009 13:28:40 Dan Langille, vous avez écrit : >>>>>>>> It appears Bacula is always trying the IPv6 address first, despite no >>>>>>>> IPV6 support on this box, then failing. Is that what we want? Should >>>>>>>> it not then try IP4 given the address is there? >>>>>>> It seems to be related to the bug #1029 corrected in May. >>>>>>> http://bugs.bacula.org/view.php?id=1029 >>>>>>> >>>>>>> >From what i read in the code, it should try the IPV6 first, then the >>>>>>>> IPV4 >>>>>>>> >>>>>>>> (if >>>>>>> any). >>>>>>> >>>>>>> If one of them is usable, the connection should be ok and you should >>>>>>> see only an Error message reported by a Pmsg() about the IPV6 protocol. >>>>>>> Maybe this Pmsg() is causing the problem. >>>>>>> >>>>>>> Can you try to comment the src/lib/block.c:1025 Pmsg() call, and re-run >>>>>>> the test ? >>>>> You should read src/lib/bsock.c and not src/lib/block.c >>>> Found. Line number is not correct though. I think you mean 228. >>> Human grep works well :) >> Well, I want to make sure I'm doing as expected. :) >> >> >> === Starting pretest at 08:34:42 === >> >> Starting the Bacula Storage daemon >> Starting the Bacula File daemon >> Starting the Bacula Director daemon >> Connecting to Director localhost:8101 >> Error: Starting Bacula Failed in pretest >> scripts/cleanup >> rm -f tmp/file-list >> rm -fr tmp/* working/* Testing >> >> >> Still an issue. I'll look close at pretest now. > > Interesting. Running the commands manually succeeds: > > > [...@ducky:~/src/BaculaRegressionTesting-TRUNK/regress] $ zstat=0 > [...@ducky:~/src/BaculaRegressionTesting-TRUNK/regress] $ > [...@ducky:~/src/BaculaRegressionTesting-TRUNK/regress] $ debug_wait > [...@ducky:~/src/BaculaRegressionTesting-TRUNK/regress] $ echo $debug > 1 > [...@ducky:~/src/BaculaRegressionTesting-TRUNK/regress] $ > ${scripts}/bacula-ctl-sd start > Starting the Bacula Storage daemon > [...@ducky:~/src/BaculaRegressionTesting-TRUNK/regress] $ > ${scripts}/bacula-ctl-fd start > Starting the Bacula File daemon > [...@ducky:~/src/BaculaRegressionTesting-TRUNK/regress] $ > ${scripts}/bacula-ctl-dir start > Starting the Bacula Director daemon > [...@ducky:~/src/BaculaRegressionTesting-TRUNK/regress] $ ps auwx | grep > bacula > root 43698 0.0 0.3 5436 2608 ?? IsJ 7:27AM 0:00.02 > /usr/local/sbin/bacula-fd -u root -g wheel -v -c /usr/local/etc/bacula-f > dan 82628 0.0 0.3 4704 3024 ?? IsJ 8:55AM 0:00.01 > /usr/home/dan/src/BaculaRegressionTesting-TRUNK/regress/bin/bacula-sd -v > dan 82777 0.0 0.3 4400 2860 ?? SsJ 8:56AM 0:00.01 > /usr/home/dan/src/BaculaRegressionTesting-TRUNK/regress/bin/bacula-fd -v > dan 82860 0.0 0.3 5116 3448 ?? SsJ 8:56AM 0:00.01 > /usr/home/dan/src/BaculaRegressionTesting-TRUNK/regress/bin/bacula-dir - > dan 82896 0.0 0.1 1544 612 p1 R+J 8:56AM 0:00.00 grep bacula > [...@ducky:~/src/BaculaRegressionTesting-TRUNK/regress] $ cat > ${tmp}/bconcmds | ${bin}/bconsole -c ${conf}/bconsole.conf > Connecting to Director localhost:8101 > 1000 OK: localhost-dir Version: 3.0.3 (22 August 2009) > Enter a period to cancel a command. > messages > You have no messages. > quit > > > Still testing.
The reason it 'fails' but 'succeeds': $ run_bacula Starting the Bacula Storage daemon Starting the Bacula File daemon Starting the Bacula Director daemon Connecting to Director localhost:8101 [...@ducky:~/src/BaculaRegressionTesting-TRUNK/regress] $ echo $? 1 [...@ducky:~/src/BaculaRegressionTesting-TRUNK/regress] $ The result is 1 but there is no error shown. Still looking. ------------------------------------------------------------------------------ Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day trial. Simplify your report design, integration and deployment - and focus on what you do best, core application coding. Discover what's new with Crystal Reports now. http://p.sf.net/sfu/bobj-july _______________________________________________ Bacula-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/bacula-devel
