> On Fri, 5 Dec 2003, Ross Macintyre wrote: > >> Hi, >> I have a machine that has a corrup RPM database. >> I tried the usual >> rm var/lib/rpm/__db.00*, then >> rpm --rebuilddb >> which has worked in the past. >> This time nothing. > > If rpm is getting stuck, rather than crashing, that often means some other > process has the database locked. Look for such processes, or even reboot > before removing the __db files. See also > http://www.rpm.org/hintskinks/repairdb/ > for further rebuilding hints if necessary. thanks to all who replied. Yes the problem was fixed by a reboot and a rebuild. I got caught out here because 'df' (no args) wasn't hanging and could be killed with ctrl-c. What I did find was that when I gave the verbose option to rpm, I did see df failing on a particular mount point; when I did try df on that mountpoint, it did hang and couldn't be ctrl-c'd. The trouble here was that I was loathe to reboot, and the lesson I've learned is that I shouldn't rely on this server's RPM database(I was using it to compare against other machines RPM database for consistency). Thanks, Ross > > Michael Young > > > -- > Psyche-list mailing list > Psyche-list@xxxxxxxxxx > https://www.redhat.com/mailman/listinfo/psyche-list > -- Ross Macintyre (raz@xxxxxxxxxxxxx) -- Psyche-list mailing list Psyche-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/psyche-list
