[Cialug] Fedora - the dreaded rpmq failure

[Josh More]

RPM is based on the Berkley DB, which sometimes corrupts.  That's what
the --rebuilddb is supposed to fix.  If I recall correctly, you used to
need to go into /var/lib/rpm/ and remove the __* files.  (Best to move
them instead of deleting them.)

Interestingly, while a lot of systems use RPM that is back-ended by
Berkley, I've only seen this problem on RedHat/Fedora boxes.  My SUSE
and Mandriva boxes seem to be OK.

Another thing to consider is what used to happen with Red Carpet in the
early days.
If you are running a nightly update process, and a repository isn't
available, rpmq might start cycling.  The solution there is to either
turn off automatic checking or only use high-availability repositories.



 

-Josh More, RHCE, CISSP, NCLP, GIAC 
 morej at alliancetechnologies.net 
 515-245-7701



>>> David Champion <dchampion at visionary.com> 08/07/07 12:32 PM >>> 
Anyone have a real solution for the dreaded Fedora rpmq hanging from
the
/etc/cron.daily/rpm script (or anything else that does rpm -q)?

Here's a random googled thread on it...

http://linux.derkeiler.com/Mailing-Lists/Fedora/2007-01/msg01564.html

Noticed this happening on a FC6 box today, I'd forgotten about it, but
seem to recall this being a problem with Fedora like a couple of years
ago. Can't believe they haven't solved it yet.

I'll do a reboot and rpm --rebuilddb later when people aren't using
it... but would be nice to find a real fix for it.

-dc

_______________________________________________
Cialug mailing list
Cialug at cialug.org
http://cialug.org/mailman/listinfo/cialug