At 4:49 PM +0800 10/5/05, Edward Dekkers wrote: >Spencer Kellis wrote: >> I would have thought bad memory would be manifest in many more ways than >> simply yum not working. My system is otherwise stable, and has been up >> and operating for a week currently without any other issues. I >> appreciate the idea though, and if you still consider bad memory an >> option I'd be interested to hear it (and how & why). >> >> If there are any other ideas out there, even pointers to what to look >> through for possible problems on my own, I'd appreciate it. >> >> Thanks, >> Spencer > >The reason he would have suggested it is because it's probably the #1 >cause of segmentation faults and signal 11 faults on Linux, together >with bad hardware. This isn't Windows and you aren't in Kansas any more. > >Linux is super stable and if it crashes, it is more than likely not >actually Linux's fault. > >Trust the reply to your post - run a memtest86 overnight on full testing >suite, don't question a perfectly reasonable response. I think the >person who replied has been around Linux longer than you from what I can >see. > >If the memory tests OK, fine, we'll look at something else, but you >really need to eliminate it 100% sure before we go on. My own guess is that there is something badly wrong with RPM's database, and that --rebuilddb might need some help, such as rm'ing the existing database first. Hopefully someone who knows more about RPM will chime in. ____________________________________________________________________ TonyN.:' <mailto:tonynelson@xxxxxxxxxxxxxxxxx> ' <http://www.georgeanelson.com/>
