Oops, forgot part of the sentence; the post should read: I did the strace thing (thanks for the info there), and I found that eggcups, magicdev, fam, and xscreensaver were doing something at exactly the same time that the accesses were showing up in gkrellm, so I killed them, and that stopped the regular accesses of the hard drive. Each one was responsible for some amount of the total. Both eggcups and xscreensaver were checking local time, among other things. How is it that all of these processes access the hd at exactly the same time? On Mon, 2003-04-28 at 20:40, * wrote: > I did the strace thing (thanks for the info there), and I found that > eggcups, magicdev, fam, and xscreensaver, and that stopped the regular > accesses of the hard drive. Each one was responsible for some amount of > the total. Both eggcups and xscreensaver were checking local time, among > other things. > > How is it that all of these processes access the hd at exactly the same > time? > > On Mon, 2003-04-28 at 09:00, Eric Wood wrote: > > Well, according to http://bugzilla.gnome.org/show_bug.cgi?id=42242 this bug > > was squashed and they're polling /proc/mounts now. > > > > Robert Day, can you strace your running applications one at a time until you > > find the app that keeps accessing your drive? > > > > # strace -p <pid> > > > > Just do the obvious pid's and you should see a several system calls fly by > > every 5 seconds - then you know you've got the right application and we'll > > go after them! > > > > -eric wood > > > > Klaasjan Brand wrote: > > > On Mon, 2003-04-28 at 04:26, Eric Wood wrote: > > > > > >> Grodon is right. I complained about this a year ago, how can a linux > > >> hard drive ever go into power saving mode when keep stat()ing > > >> /etc/mtab. There is another stat() which should have been used by > > >> Nautilus that doesn't change the atime of a file (according to the > > >> author of the ext filesystem, Steven Tweedie). Nautilus should > > >> have been using that other function or switch to using fam. Fam > > >> does not access the disk itself, it tells a subscribing program > > >> that a certain file has changed. I have no idea why the nautilus > > >> guys haven't fix this. > > > > > > Maybe because they are "buried in bug reports" and have more to do? > > > > > > Anyway, I didn't find a report on bugzilla.gnome.org nor on > > > bugzilla.redhat.com, so you might file a bug or 2... > -- > * <redhat@xxxxxxxxxxxxxxxxx> -- * <redhat@xxxxxxxxxxxxxxxxx>
