Re: Unkillable process
Oki DZ wrote:
>
> "Karsten M. Self" wrote:
> > Using kill -9 on a process means you may have to clean up the pieces.
> > Signals 15, 1, and 2 (TERM, HUP, and INT), are generally considered to
> > be polite requests to jobs to get the hell over it already, but to clean
>
> I think Unix designers were having mixed feelings about "kill"; "kill
> -15", killing me softly... ha, just like a song, ...killing me softly
> with his words ...killing me softly with his songs...
>
> > up on the way out. SIGKILL is nonmaskable, and a process *can't*
> > perform cleanup or garbage collection even if it wants to.
actually kill is not really kill, it just send signals, some of them
ask or force process to die...
> I see; so the memory that once was used, wouldn't be returned back to
> the OS, right?
AFAIK the OS takes care of all/most of the resources - file are closed
(but not saved), memory is released etc... if you see frozen netscape,
before killing it check the free memory and check the memory after you
killed -9 netscape, lot of memory is freed (not sure if all)
> > This is a case where you may have to shut down. Sometimes you can get
> > the buggers if you shoot at 'em enough ways though.
>
> Whoa, I tried many times, kill -9, killall -9 <progname>, to no avail.
> BTW, if I unload the NIC driver (along with lo), would the daemon exit?
> I was thinking about it, but since I was remote logging in to the
> machine, rebooting was the only option.
check if it's a zombie - if it's a zombie it was already killed, just
waits for something to be really unloaded from memory (usually it
doesn't take long). IIRC one reason an application is zombie is that its
parent waits for return value (which is sort of held by zombie, waitin
for parent to process the info or something like that).
generally, if you see a zombie, just wait. sooner or later they should
go away...
erik
Reply to: