On Friday 06 October 2006 02:21, Michael Brennen wrote: > > Per Kern's request I've been trying to get the regress scripts set up to find > a bug in multi-drive usage. In a 4 tape autochanger library only one tape is > being used. Up to and including the 9/9/2006 beta multiple drives were being > used for archives in different Pools That seems have been lost in recent > versions, with no configuration changes. This is a status report on > diagnosing the multi-drive problem. > > Kern, you mention in your README that running on a production system will > overwrite the database, unless a different engine is used. I've set up with > sqlite to avoid conflict with the mysql production system. What you didn't > note is that the regression tests will overwrite tapes loaded in an > autochanger! I saw what it was doing and stopped the test in time. You > might want to add a warning about running regress on an autochanger with live > archive tapes; it will use what is there without asking which ones to use. I > now have a test system running with some scratch tapes.
I've added the following to the README: ======== Important !!! ============ When running the tape tests, Bacula will write on any tape that is in the tape drive that you have configured. If it is a production Bacula tape, it will be destroyed. If you have configured an Autochanger, Bacula will write on the tapes in slots 1 and 2 thus destroying any information on those tapes, even if they are Bacula production tapes. =================================== > > After a couple of hours I have the regress scripts built against sqlite and > they are trying to run. The default test tape scripts are hard coded in many > different places to run against a single drive 'DDS-4' device setup. I am > modifying the various scripts and configurations to put in my 4 drive library > configuration. That is slow sledding but is starting to come together. When > I have something that looks like like the bug I am seeing I will turn in a > bug report. I hope I understood what you are doing correctly. You should be creating new regression scripts that use a 4 drive library configuration rather than modifying any existing tests. If you need to modify something in the scripts directory, you need to either ensure that the modification doesn't break any existing tests, or you need to create new scripts. Normally this is rare but involves creating new bacula-dir.conf and bacula-sd.conf files (with different names of course) You might start by looking at the test named 2drive-incremental-2tape, which does test multiple drive usage. Also, I would recommend that you attempt to reproduce the problem on a two drive autochanger, otherwise, I won't be able to use the test unless you send me both the test and a four drive autochanger. Even more ideal would be to duplicate the test using my autochanger simulator, where you can have up to 10 drives. For example, see the test 2drive-incremental-2disk. I suspect that you should be able to reproduce it with this script, but it is an autochanger simultor, so there are possible subtle differences. Another approach would be to demonstrate the problem on a 4 drive tape autochanger, then modify it to use my simulator, and if that shows the problem you can avoid sending me a four drive autochanger, since I lack space to for installing it :-) Regards, Kern ------------------------------------------------------------------------- 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
