On Sat, Jan 27, 2007 at 03:30:48AM +0000, Tyler wrote: > In the meantime, how does one check for bad blocks and bad ram? I have > the pre-installed fsck running every 30th boot, and so far no errors > have ever been reported. >
Check the man page for the filesystem-specific fscker (honest, I'm not swearing) for your system. For e2fsck its -c -c. Note that this isn't done from shutdown -F. The way to run this is either from a rescue system (e.g. the install media) or from booting with the init=/bin/sh command line. This leaves the / fs mounted ro where you can safely run fsck on all your filesystems. This also has the added bonus of triggering the drive's automatic badblocks reallocator, e.g. badblocks tries to write to a sector, the drive tries to do that, gets an error, tries again a few times, gives up and returns a new block with the data written. All transparently to you until the drive runs out of spare blocks. So, in effect, you should never even __find__ a bad block until the drive is at the end of its life. I know you solved the origional problem by using gimp instead of gv, but what happens if you have an xterm on top (above gv) with top running? Watch the amount of swap used. If its much, then you should add ram. Its cheap. Doug. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]