[svlug] Processes that don't die with -9

[kevin dankwardt]

A process that is a zombie can't be killed, even with 9, because it is
already dead. It hasn't been cleaned up because its parent either hasn't
done a wait() or caught the sigchild signal. Killing its parent will also
clean it up.

Tasks (processes or threads) can be blocked "noninterruptible" in the
kernel. This means that they are blocked and will not receive any signals.
Including the kill signal, number 9. This is kernel behavior and there is
not necessarily anything you can kill to kill those. You can look at ps to
see if it is blocked or zombie. NFS, is an example where sometimes things
get blocked uninterruptible and you just have to wait for it to time out as
far as I can tell.

If what you saw before was something waiting on I/O that couldn't be killed
it certainly sounds like they were blocked uninterruptible. It could be the
driver for the I/O device or something else.

On Thu, Dec 4, 2014 at 11:55 AM, Sarah Newman <sarah.newman at computer.org>
wrote:

> I've had times in the past where kill -9 did not work on a process. The
> solution was typically to
> take care of whatever IO the process was waiting on.
>
> But now I'm wondering if killing a different thread (LWP) other than the
> primary *could* have
> worked, which as I learned last night are shown by adding '-L' to 'ps'.
> Does anyone know?
>
> Thanks, Sarah
>
> _______________________________________________
> svlug mailing list
> svlug at lists.svlug.org
> http://lists.svlug.org/lists/listinfo/svlug
>



-- 
visit www.kcomputing.com for the best in Linux developer training.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.svlug.org/archives/svlug/attachments/20141204/e0fed263/attachment.htm