[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

harddrive write causes system to tie itself to death



i have this older system with a 20Gb maxtor IDE drive. i have been
experiencing some weirdities therewith: on the largest partition
(/dev/hda8), a number of write attempts can cause the rsync (or
badblocks) process causing them to eat up all the system's resource to
a point where there is no interaction possible with the system, not
even through the tty console locally. killing the remote rsync usually
rescues it after a couple of minutes, but the processor queue will
then be in the excess of 40 or 50.

so my question: what's wrong, or should i just say bye to the
harddrive and get a new one?

and also, how can this happen? i understand that badblocks might get
locked up trying to write if the harddrive is too broken to do it, but
it's single-threaded, no? and when the process can't write, it goes
into uninterruptible sleep mode. doesn't that mode allow the scheduler
to use the (unused) processor for another process? how then can
a queuelength of 40 or 50 come about?

oh yeah, and of course there are no log entries or console output
related to this. and the harddrive does not make funny noises or
anything like that.

it's just weird.

-- 
martin;              (greetings from the heart of the sun.)
  \____ echo mailto: !#^."<*>"|tr "<*> mailto:"; net@madduck
 
NOTE: The pgp.net keyservers and their mirrors are broken!
Get my key here: http://people.debian.org/~madduck/gpg/330c4a75.asc
 
humpty was pushed.

Attachment: pgpgvPIvw5qE2.pgp
Description: PGP signature


Reply to: