"Daiajo Tibdixious" <[EMAIL PROTECTED]> posted [EMAIL PROTECTED], excerpted below, on Mon, 19 Jun 2006 16:02:57 +1000:
> # grep hda6 /etc/fstab > /dev/hda6 / ext3 noatime > 0 1 > I'm also puxxled why hda6 is never fsck'd during boot, I though having > that last 1 would check it every time. That's because it's root, and because it's ext3 -- the journalling changes the way things work a bit and full fscks aren't needed as often. (I run reiserfs, not ext3, so don't know the exact details, but IIRC, ext3 has a counter that will run the fsck every X number of boots, with X tweakable using I believe tunefs.) I'd suggest you read up on ext3 if interested. As it happens, Gentoo's own founder Daniel Robbins has one of the better known series comparing the various filesystem types. He wrote it for IBM DeveloperWorks back when 2.4 was new. (Reiserfs at least is much improved since then, with the data=writeback/ordered/journalled choices of ext3, defaulting to ordered. The others will have changed some as well, but probably not as dramatically.) In any case, he explains journalled filesystems quite well, as well as the writeback/ordered/journalled distinction and the series still serves as a very good intro to the various filesystems so it's still well worth reading. I believe Gentoo still links to it in their "Other papers of interest by Gentoo devs" or whatever they call it section, so that's where I'd look. > After the fsck the system works normally, I emerged system & world with > no problems. > > firefox-bin still refuses to save links, but no longer crashes, but > after some time idling, it eats all CPU and has to be kill -9'd. You may want to verify that you've remerged the 32-bit compatibility libraries that firefox-bin depends on. As for why this might be happening, I'd say there's a fair chance your disk is getting ready to die. I'd DEFINITELY recommend getting backups ASAP, if you don't have them, and don't want to lose whatever. Also start thinking about a new disk, altho as long as you are keeping good backups, the disk you have might be fine for awhile, even if it's starting to go. You'll likely just gradually have more and more problems with it. Of course, a failing power supply or overheated CPU or memory (you aren't overclocking are you?) could also do it, but aren't generally as catastrophic to lose in terms of data loss as losing a hard drive without backups can be. Depending on the problem, it can still cause crosslinking and the like, so you /can/ lose files, but it's far more likely a file here and a file there than whole sections of the disk as it can be if the disk is the problem (and that's assuming the entire disk doesn't simply die, all at once). No matter /what/ it ends up being, I'd be verifying your backups, if you don't want to lose your data. That's for SURE. I had a disk overheat last year due to an AC going out -- summers here in Phoenix can be brutal 45C in the ventilated shade, so the AC goes out and the computer room could easily hit 55C, the computer ambient 70C, and the drives... WAYYY too high!!! Now I'm running 4x300 gig Seagate SATAs in RAID, a bit slower, but the best warrantee (5 yr.) and I've got the critical stuff on RAID-6, so can lose two of the four drives without data loss. Oh, new AC as well! Anyway, I know a bit about drives going out, unfortunately. Fortunately, mine didn't go out all at once and I had the critical stuff backed up, tho the backups weren't as upto date as they should have been, but I got what I had to and most of what I wanted off. Actually, the drive still mostly works, but for the couple partitions that got killed in the heat. I expect if I take it apart there will be head-crash rings where the heads were when it was hot. As long as I kept it cool after that, however, the damage didn't seem to spread, but I still got the RAID set up as quickly as I could. -- Duncan - List replies preferred. No HTML msgs. "Every nonfree program has a lord, a master -- and if you use the program, he is your master." Richard Stallman -- [email protected] mailing list
