On Wed April 2 2008 10:37:52 am Jon Dowland wrote: > Package: bacula-director-sqlite3 > Version: 2.2.8-8 > Severity: minor > > Hello, > > I'm still testing bacula and I wanted to go back to a blank > slate so I removed the stuff under /var/lib/bacula, > including bacula.db. > > Afterwards, /usr/sbin/bacula-dir, and the calling init > script, return 0-success, despite the fact the process > immediately exits and the daemon does not run. > > Reinstalling the package recreates the bacula.db file, but > I think in its absence the director should shout loudly.
Can you try running the director manually from the command line and see if it returns a nonzero exit code in this situation? My hunch is that it isn't, so there is no way for the init script to detect this situation. There is probably an upstream architectural reason why this is so. It is not possible to put a test directly into the init script, because users can change the location of their database using configuration directives. So my feeling is that the init script shouldn't be a "nanny script" and try to police every thing that an admin could do to break bacula :-) -- John -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
