Hello, On 2/26/2007 10:54 AM, Ferdinando Pasqualetti wrote: > > Hi Bacula users, > sorry if you get this message two times, I sent it with a wrong sender > (not in the list), so I am sending it again. > I am facing a problem that came out with rev. 1.38.11 (I never saw it > with 1.36.3). The problem did not happen all times, but very often. Now > I switched to 2.0.2 and this problem is much more frequent. > The problem is that when a tape was exhausted bacula changes correctly > the tape in the autochanger drive but just after get this error: > > 25-feb 02:47 bacula-sd: End of Volume "web-004" at 594:3362 on device > "LTO1" (/dev/lto1). Write of 64512 bytes got -1. > 25-feb 02:47 bacula-sd: Re-read of last block succeeded. > 25-feb 02:47 bacula-sd: End of medium on Volume "web-004" > Bytes=594,178,937,856 Blocks=9,210,362 at 25-feb-2007 02:47. > 25-feb 02:47 bacula-sd: 3301 Issuing autochanger "loaded? drive 0" command. > 25-feb 02:47 bacula-sd: 3302 Autochanger "loaded? drive 0", result is > Slot 3. > 25-feb 02:47 bacula-sd: 3307 Issuing autochanger "unload slot 3, drive > 0" command. > 25-feb 02:48 bacula-sd: 3304 Issuing autochanger "load slot 4, drive 0" > command. > 25-feb 02:48 bacula-sd: 3305 Autochanger "load slot 4, drive 0", status > is OK. > 25-feb 02:48 bacula-sd: 3301 Issuing autochanger "loaded? drive 0" command. > 25-feb 02:48 bacula-sd: 3302 Autochanger "loaded? drive 0", result is > Slot 4. > 25-feb 02:49 bacula-sd: Wrote label to prelabeled Volume "web-005" on > device "LTO1" (/dev/lto1) > 25-feb 02:49 bacula-sd: New volume "web-005" mounted on device "LTO1" > (/dev/lto1) at 25-feb-2007 02:49. > 25-feb 02:49 bacula-sd: End of Volume "web-005" at 0:1 on device "LTO1" > (/dev/lto1). Write of 64512 bytes got -1. > 25-feb 02:49 bacula-sd: webfs3-job.2007-02-24_20.03.22 Error: Re-read of > last block OK, but block numbers differ. Last block=0 Current block=9210362. > 25-feb 02:49 bacula-sd: Job write elapsed time = 06:43:26, Transfer rate > = 24.52 M bytes/second > 25-feb 02:49 webfs3: webfs3-job.2007-02-24_20.03.22 Fatal error: > backup.c:860 Network send error to SD. ERR=Pipe rotta > 25-feb 02:49 bacula-dir: webfs3-job.2007-02-24_20.03.22 Error: Bacula > 2.0.2 (28Jan07): 25-feb-2007 02:49:25 > > It seems there are two problems, the first one (and the most important > one) is that bacula get an end of volume on the new tape,
What Bacula reports as an EOT can be caused by a drive error, too, so for the time being I assume that the second error is tightly related to this one. > and the second > one is the difference in the last block (it appears to be the last block > of the previous tape). If that's the case, and your description seems quite clear, you might have found an OS or hardware bug, too. This is only guesswork, but it could be possible that, after a tape change, the hardware or the tape driver don't update their state information. If that's the case, you could try the following: - first, have a look at your system log and dmesg output. There might be errors reported there. - second, try to reproduce the problem without using Bacula. Unmount the tape drive from bconsole. Load a tape (an unused one, or one with write protection). If you use an empty tape, write some data and some file marks to it, ending with an EOT mark. dd and mt are tools for that purpose. Then, use tapeinfo or st to observe the tape status, especialy the block position reported, when doing some rewinds, fast forwards, offline, and see what happens after you used mtx to unload and reload that tape. If there really is a problem with the hardware or the OD driver, you should be able to reproduce it then. Updating the drive firmware and the OS (or, if that's up to date, filing a bug report) would be two options then. Otherwise, you should run btape again, because there are some things in the report I don't like - errors writing the last block to tape should not happen with current hardware, for example. You might try to tune your device configuration, and perhaps you'll have to set the tape driver to a different write mode. Suggesting something is difficult without seeing how it's setup now :-) > Bacula is a MySQL version on a RedHat AS 4.04, rpmbuilt on that system, > an HP proliant G3 3.2 Ghz, 2Gb RAM. > The tape is an MSL6000 with two LTO-3 drives, drived by bacula directly > (not using the autochanger as device - 1.36.3 setup). > Btape tests run correctly, including the "fill and change tape" (I am > attaching the test result, if someone is interested). > Did anyone get a similar problem? That basic setup should run ok I think... nothing unusual there. Arno > > -------------------------------------------------------------------------- > Ferdinando Pasqualetti > G.T.Dati srl > Tel. 0557310862 - 3356172731 - Fax 055720143 > > > > ------------------------------------------------------------------------ > > ------------------------------------------------------------------------- > 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 -- IT-Service Lehmann [EMAIL PROTECTED] Arno Lehmann http://www.its-lehmann.de ------------------------------------------------------------------------- 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
