On Wednesday 20 June 2007 16:43, BOLLENGIER Eric wrote: > I launched more than 125 jobs under heavy conditions,
125! Yea, that should be a good test :-) > and all of them have run fine ! Ah, great, good news. > > Thanks > > > On Wednesday 20 June 2007 14:16:19 Kern Sibbald wrote: > > On Wednesday 20 June 2007 11:33, Andreas Helmcke wrote: > > > On 20.06.2007 09:31, Kern Sibbald wrote: > > > > I think I have worked out all the problems with the Storage daemon and > > > > multiple drive autochangers, so I would appreciate it if Eric, Andreas > > > > and the beta testers would try the current SVN. Given the number of > > > > changes I made, it would not be too surprising if there are still some > > > > unresolved issues, but *all* the regression tests now run perfectly > > > > fine -- here (on SuSE and CentOS). > > > > > > > :-) Looks good! :-) (svn Rev 5047) > > > > > > At least my standard three job test runs well. All jobs are using the > > > already loaded tapes. > > > > > > I also tried some combination of jobs (some with loaded tapes, some have > > > to load tapes) and had no errors. > > > > Ah nice. Finally !!!! > > > > > Well done! > > > > Thanks. > > > > > As this is true for *backup* it doesn't work for reading tapes. Restore > > > and verify jobs still use the first available drive and not the one > > > where the tape is already mounted. > > > > Yes, true, thanks for reminding me. I explicitly did not implement it for > > restore yet. I didn't want to "break" too many things at one time. > > > > I don't think it will be hard to fix, the same code *should* work. > > However, in the next few days, I would like to look at the problem you > > raised about prunning/recycling Scratch volumes. If you don't hear back > > from me on the restore drive selection issue in a few weeks, please don't > > hesitate to remind me -- though I *am* making a note of it. > > > > Thanks again for all the testing. Especially for testing so rapidly! > > Without that, this would have required months to fix. > > > > Regards, > > > > Kern > > > > > > Note, the SD reservations system should now re-use volumes in the drive > > > > on which they are mounted when the job starts. However, if during the > > > > job, a > > > > errors > > > > > > Volume is needed that is in another drive, the SD will still move that > > > > volume > > > > > > to the drive the job is using. > > > > > > I consider this a minor problem. The normal case will be concurrent > > > running daily incremental jobs where the tape stays in the drive and > > > will be reused the next day. So when starting this jobs they now reuse > > > their tape _and_ drive. At least, this is true for me. > > > If the job needs another tape it is most like because the tape is full > > > or has errors so the job will need a "fresh" (purged) tape which most > > > likely has not been used for a long time and therefor it is very > > > unlikely that the tape is still loaded. > > > > > > > > > Andreas > > > > > > > ------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ _______________________________________________ Bacula-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/bacula-devel
