On Tuesday 28 February 2006 15:32, Ian Levesque wrote: > Hello Kern, > > On Feb 28, 2006, at 3:19 AM, Kern Sibbald wrote: > > In fact, the output that I saw seemed to me to be non-fatal > > What do you suppose worst-case scenario is here -- that roughly 64512 > bytes may be lost at the end of the tapes?
That should be the worst case. Much more likely, nothing is lost. However, if there is a Bacula bug, the probability of data corruption/loss goes up. The comments you make below, make me think that there may be a problem ... > > > My best guess is a tape defect. If it happens with more than one > > tape, either you have a bad batch of tapes, or you have a defective > > drive or > > the logical end of tape reader head is dirty. > > Bad tapes might be the problem, as this not only happens with more > than one tape, but it happens with more than one drive. One other > thing I just noticed after you pointed it out is that '-1' is > returned every time I reach the end of the tape, though quite often > bacula reports success so I hadn't noticed... > > 26-Feb 11:13 sbgrid-sd: End of Volume "000032L1" at 148:7121 on > device "LTO-1" (/dev/nst1). Write of 64512 bytes got -1. > 26-Feb 11:13 sbgrid-sd: Re-read of last block succeeded. > > > On the other hand, I cannot 100% exclude the possibility that some > > lower > > level Bacula subroutine releases the device lock thus allowing some > > race > > condition with other threads (jobs). I'll take a look at the code > > with that > > in mind in the near future -- this week I'm tied up with family > > visitors so > > don't have much time. > > No worries! Let me take this opportunity to properly thank you for > all the work you've done on this project. I'm planning on putting > together a writeup (once I get the backups working without a hitch) > to document the successful transition we made from a commercial > application, etc. > > > Unfortunately, testing such a problem is not easy as it takes a > > long time to > > write most tapes (especially LTOs). > > Yep :) But I've got a few jobs coming this weekend that will go > through some volumes, if you'd like me to test anything out. > > Cheers, > Ian > > > > ------------------------------------------------------- > This SF.Net email is sponsored by xPML, a groundbreaking scripting language > that extends applications into web and mobile media. Attend the live > webcast and join the prime developer group breaking into this new coding > territory! > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642 > _______________________________________________ > Bacula-users mailing list > Bacula-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/bacula-users ------------------------------------------------------- This SF.Net email is sponsored by xPML, a groundbreaking scripting language that extends applications into web and mobile media. Attend the live webcast and join the prime developer group breaking into this new coding territory! http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642 _______________________________________________ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
