You are not logged in.
My main system drive keeps churning over, causing programs to load slowly and, during heavy use, bringing the computer almost to a halt. Most programs are handled fine, but certain ones, particularly Nautilus, trigger up to a minute of constant drive activity before they load up but don't necessarily slow the rest of the system.
When more software is running it happens seemingly at random and sometimes when the machine's been idle for several minutes. In this scenario, the activity is as good as perpetual and interferes with normal operation to the extent that switching to a terminal, waiting several minutes between inputs, and running killall on a few programs is the only solution. Sometimes it gets to the point that mouse and keyboard control fails (though the cursor still jolts around the screen) and rebooting is necessary.
I had a similar problem in Ubuntu 9.10, but it was only when the swap was in use. I've been running sans swap since returning to Arch, and everything was pretty smooth and responsive until a short time ago.
Offline
I wonder if you're running out of hard drive space on one of your partitions? Or RAM? Open a terminal and type "df -h" and "free" and show us the output. Also, use "top" to see what process is active when you get the HD thrashing. Also use the "iostat" command to see what is doing the thrashing...
Philosophy is looking for a black cat in a dark room. Metaphysics is looking for a black cat in a dark room that isn't there. Religion is looking for a black cat in a dark room that isn't there and shouting "I found it!". Science is looking for a black cat in a dark room with a flashlight.
Offline
It's possible low memory contributes to the problem in the second scenario where a lot of programs are running and everything collapses. But this doesn't explain why programs should spend half a minute disk churning when there's only about 2-300Mb RAM usage.
df -h
Filesystem Size Used Avail Use% Mounted on
udev 10M 156K 9.9M 2% /dev
/dev/disk/by-label/archroot
12G 5.3G 5.2G 51% /
shm 502M 276K 502M 1% /dev/shm
/dev/sda6 31M 14M 15M 49% /boot
/dev/sda5 9.2G 5.1G 3.7G 59% /home
free -m
total used free shared buffers cached
Mem: 1003 803 199 0 105 379
-/+ buffers/cache: 318 684
Swap: 1427 0 1427
Normally there is a 1TB drive attached at about 50% use, and a 750Mb at 80%. I think the 1TB is ext4 and the 750Mb is NTFS, in case that might have any bearing on the issue. I've unplugged them to see if that improves things. Currently everything's pretty smooth, and there's no unnecessary disk activity.
Last edited by Jimage (2010-10-22 05:37:36)
Offline
The ntfs-3g driver causes heavy load in my experience. And maybe Nautilus just tries something on all drives when startet, maybe searching for icons, or indexing something or whatever insane stuff things like Nautilus do.
Ceterum autem censeo Systemdinem esse delendam
Offline
Well, it's nothing to do with those drives. Just arrived home from a weekend away. The computer was left idling with XChat and Chromium running. I got home, went to load up a page and the HDD churning kicked in. I've managed to drop back to a shell and login, but iotop has hung. The drive activity continues.
Alright... it just loaded up. There are about 15-20 entries running several hundred K/s disk read, many around 800 and several peaking at 1.2Mb or so.
Hmm. I ran killall chromium, which didn't help. Then killall Terminal (Xfce Terminal Emulator 0.4.5). The disk usage halted within a few seconds. iotop is back to all zeros. I doubt the problem is specifically with this program, but we'll see next time it happens. I'll log iotop to a text file next time as well.
What's particularly strange is that it doesn't happen on the EeePC with an essentially identical setup, from which I'm typing this. I imagine either a driver is playing up, or there's some resident software playing havoc that I haven't installed on the EeePC.
Last edited by Jimage (2010-10-24 12:42:09)
Offline
Have you checked out smart? Possibly there are problems with you hard drive. I think the wiki page needs to be updated for a few things. There is a link on the page of Ubuntu's version that I think is more up to date.
Setting Up a Scripting Environment | Proud donor to wikipedia - link
Offline
Careful, I had a machine with similar symptoms once and it turns out the harddisk is dying.
Offline
I doubt your issue lies with ntfs-g3. That is a VERY stable driver. I would be more concerned about a hardware failure. Backup critical data now!
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
Come to think of it, this drive did come out of a computer I ended up with that was having a problem I couldn't diagnose at the time. Was probably the same problem. Ubuntu was running off another identical hard drive. It might have come from that same computer, as it was running a RAID. Guess I'll transfer the system across to another drive ASAP.
Offline
I'll second the stability of NTFS-3G. I've used it for about 2 years with Ubuntu and haven't had problems at all. I've also been using it on my desktop which now runs Arch + KDE, and I don't have any disk thrashing whatsoever. Also, is the 1TB hard drive that you have a Seagate? There is a certain batch of 1TB Seagate hard drives that is extremely prone to very premature failing, and I'd venture to say about 2/3 of all of them die.
Here is an example of the drive, in case you want reference: http://www.tigerdirect.com/applications … No=4691121
Offline
I second the notion of looking at your drive's SMART data, namely Reallocated Sectors Count.
Offline
It's not the TB at fault, though it is a Seagate. I don't remember when I got it, but I'm pretty sure it was last year sometime, so it's survived a fair while so far. The two problematic drives are 80GB Seagates. It happened again today, and killall Terminal followed by Chromium did the job again.
Offline