Gene Heskett wrote: snip:
Just for grins, Chris, re-seat the dimms and rerun memtest86 if any errors show up. I've had cheap dimm sockets get flakey even when gold plated dimms were plugged in.
Ok, well would have emails sooner, but with work; time isn't always available, specially when I get to do 2 separate jobs at one time. Anyways; on running the memtest86; I did get some errors, at 18.0MB memory sections and 18.2MB memory sections (address range always different; ran 28 passes through the tests; ran it about 10 hours straight). I haven't gone and resit the chips yet, as with time and the anoayance of getting to the chips gets to change the decision a lot. Anyways; if I this memtest86 is completely accurate, then I may have a chip that is starting to go bad. I should probably just find which it is and replace, but I am not because of lack of money. Already had one stick go bad, and haven't replaced that either.
Now as for the rpm database, going through and verifying it; it is still in good condition; backed it up again and rebuilt it. Still didn't help any. Just playing around, I checked to see if the database even knows what package owns that file, and it says bcel; which as I gather from several places on the web; that is correct. Tried reinstalling the bcel package a couple times; and also done ldconfig again; still doesn't see the file. Any ideas on how to fix that, so the computer will see it? So far it seems that, the rpm database is good, the ldconfig cache is good (checked the logs to make sure that it sees the file, and it does), and the package is good. Tried removing that file, and reinstalling that package, so the file most likely is good. The only this left that hasn't done yet that I can think of, is reinstalling the package that needs that file. Would that even have any effect in that?
Chris T.
