geek stuff
Mar. 21st, 2005 07:33 pmi'm normally *not* a hardware person. my usual requirements for a PC are "get me something that's sure to work".
when even those requirements can't seem to be met, I'm *usually* either in trouble, or i just live with what does work.
2 years ago, parts of my linux box seemed to die -- i couldn't find a power-cable mapping configuration that would allow both the scsi cdburner and the ide cdrom to actually both work at the same time. one, or the other, or one of my hard drives would simply not show up.
sometime last year, i started having IDE problems. the secondary IDE slot (which had one hard drive and the IDE cdrom) started having minor problems. usually a reboot with its requisite fsck would take care of things, but they would return within a couple of days. files would *seem* corrupted, though they'd actually be fine on the disk (and of course, HAD to be fine on the cdroms which worked fine in other machines). in spite of things being really ok in the end, the annoyance factor continued to grow, and i definitely decide to accept the potential loss of the 2nd ide system entirely and move the disk over to ide #0....when i found the time (hah hah).
so i finally, today, managed to get away from work with a little free geek-time to myself. i undid *everything* inside it and rewired it from scratch. in the course of that, a metric ton of dust was kicked up (i'm still sniffling), and some of that came out of the 2nd IDE cable connection when I undid that (maybe, hopefully, that was the problem, causing a weak connection?).
i then rearrainged everything, switched everybody's master/slave settings to cable-select on the grounds that i fully expected to configure the cables correctly this time. hit the 'net to make sure i was doing it right. turned on the power...
wow.
everything worked, first time, and my SCSI cd-burner seems to have come back to life! and redhat managed to re-locate my second disk's filesystem to its new location as slave 0 instead of master 1.
now that everything seems to be working (i'll give it another day or two to be sure), I feel more confident finally upgrading this box off of the ancient-history RH9 to a fedora or something else (I have *tons* of distribution cdroms to consider, thanks to picking up british magazines).
and heck, I also don't feel so nervous about going and buying some more memory for this thing...yeah, its an old box (1998), but for linux its fine if there's enough memory to handle java and firefox.
when even those requirements can't seem to be met, I'm *usually* either in trouble, or i just live with what does work.
2 years ago, parts of my linux box seemed to die -- i couldn't find a power-cable mapping configuration that would allow both the scsi cdburner and the ide cdrom to actually both work at the same time. one, or the other, or one of my hard drives would simply not show up.
sometime last year, i started having IDE problems. the secondary IDE slot (which had one hard drive and the IDE cdrom) started having minor problems. usually a reboot with its requisite fsck would take care of things, but they would return within a couple of days. files would *seem* corrupted, though they'd actually be fine on the disk (and of course, HAD to be fine on the cdroms which worked fine in other machines). in spite of things being really ok in the end, the annoyance factor continued to grow, and i definitely decide to accept the potential loss of the 2nd ide system entirely and move the disk over to ide #0....when i found the time (hah hah).
so i finally, today, managed to get away from work with a little free geek-time to myself. i undid *everything* inside it and rewired it from scratch. in the course of that, a metric ton of dust was kicked up (i'm still sniffling), and some of that came out of the 2nd IDE cable connection when I undid that (maybe, hopefully, that was the problem, causing a weak connection?).
i then rearrainged everything, switched everybody's master/slave settings to cable-select on the grounds that i fully expected to configure the cables correctly this time. hit the 'net to make sure i was doing it right. turned on the power...
wow.
everything worked, first time, and my SCSI cd-burner seems to have come back to life! and redhat managed to re-locate my second disk's filesystem to its new location as slave 0 instead of master 1.
now that everything seems to be working (i'll give it another day or two to be sure), I feel more confident finally upgrading this box off of the ancient-history RH9 to a fedora or something else (I have *tons* of distribution cdroms to consider, thanks to picking up british magazines).
and heck, I also don't feel so nervous about going and buying some more memory for this thing...yeah, its an old box (1998), but for linux its fine if there's enough memory to handle java and firefox.
no subject
Date: 2005-03-22 12:38 am (UTC)no subject
Date: 2005-03-22 01:17 am (UTC)now, with that enclosure, i'm finding the fan's already dead. i need to talk to someone @ belkin to see if there's a standard replacement part or i need to shell out another $50...