I had a similar problem this morning. I noticed my load average going
thru the roof and the suspect process just wouldn't die, and then saw
in the console that the /dev/hdd was giving error messages. Rebooting
immediately showed up a problem with the whole IDE1 bus. Open the
case, reseat the cables, and whoohoo all is well again.
Howard.
>>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<
On 11/1/00, 4:16:43 AM, Matthew Dalton <[EMAIL PROTECTED]>
wrote regarding Re: [SLUG] Killing *really* dead proccess.:
> Steven downing wrote:
> > So finally my question is, is there a way to kill something which
isn't responding to those signals. And zombies?
> You know, the strangest things happen when your reading a really good
> O'Reilly Unix book...
> >From "Unix Power Tools"
> ----------
> Cleaning Up an Unkillable Process
> You or another user might have a process that (according to ps) has
been
> sleeping for several days, waiting for input. If you can't kill the
> process, even with kill -9, there may be a bug or some other problem.
> - These processes can be unkillable because they've made a request for
a
> hardware device or network resource. UNIX has put them to sleep at a
> very high priority and the event that they are waiting on hasn't
> happened (because of a network problem, for example). This causes
*all*
> other signals to be held until the hardware event occurs. The signal
> sent by kill doesn't do any good.
--
SLUG - Sydney Linux User Group Mailing List - http://slug.org.au/
More Info: http://slug.org.au/lists/listinfo/slug
