<[EMAIL PROTECTED]> aka Dan Langille schrieb mit Datum Sat, 09 Feb 2008 22:46:53 -0500 in m2n.bacula.users:
Alright, Dan, lets get realistic again. |> |You think there may be a bug. Supplying details will help others decide |> |whether or not they agree with you. |> |> Alright, Dan. Now one especially for You, and You decide: |> |> A problem we once had with one of our clients, was: when extending the |> tape library to more than 15 meters, the robot got problems grabbing |> the tapes, due to floor unevenness. |> |> Now, is this a bug? | |Umm, why are you doing this? I'm trying to help you figure out a bug and |you appear to be mocking me. No, sorry, this was more-or-less an answer which Your "for starters" provoked me to make. :-/ It is true, I know Bacula only for 14 days now - but I am in the business for more than 20 years, and while I certainly do make mistakes, and certainly am thankful for critical feedback, I also do know what a "bug" is, what an "ambitioned usage pattern" is, what "works as designed" is and what a "complex coincidence of influences" is, and I normally do not discuss about making these differentiations in general anymore. Instead, I just do act upon the matters as they arise, in order to try and solve the problems. The specific matter which has lead to this discussion thread here, I have already tracked down, documented, and handed over to the "devel"-channel - and Kern seems to appreciate it. Another matter of similar nature is currently on the way to be posted (at the moment I do verify if my remedy does really work). Both of these matters are not really dramatic, and both can easily be *workarounded* by fixing some values in the catalog (if one knows what to do) - but currently the focus goes onto fixing the code, which is the much better solution in any case. But for both of these matters I will certainly not discuss if they are bugs or not - because they are. And then there are some totally different things, which call into the realm of "ambitioned usage patterns": these are personal preferences which are more or less specific to me and my site. I give You an example. As I said before, I want to use the "Backup to disk to tape" strategy quite exclusively. This puts some additional covetousnesses onto the disk storage. For instance, if I cancel a backup in midflight, Bacula will keep that record in the database, and will keep the list of files which are already saved from the cancelled job. And this makes perfect sense for *tape* backups - because you cannot remove that stuff from the tape anyway. But if I have such a piece on the *disk* storage, I do *not* want to have it copied from disk to tape! And the most simple and straightforward solution that I have found is to just delete these records from the database in the runBefore Option of the disk-to-tape job. The point is: I am doing this setup once, and maybe after that it will run for a rather long time, while I take care of other businesses. There is no offense meant to You, Dan - it's just that I want to be taken seriously. |IIRC, and I apologise if I'm mistaken, earlier in this thread you spoke of |manually adjusting the catalog to compensate for a migration issue. If |you'd like to share details of this manual adjustment so we can delve into |the code and find out where the migration is going wrong, great. This is already solved. (As I posted before, I was working on it until yesterday.) best regards, PMc ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
