 | | | Disconnecting idle users | Disconnecting idle users 2005-06-16 - By John Haxby
Back
Lee Whatley, Contractor wrote:
> Actually the TMOUT value won't work for us anyway. According to the
> man page it only keeps track of time since input was given to the
> terminal (i.e. it does "idle timeout"). We need something that, in
> addition to tracking idle time, sets a hard limit on how long a
> session can be open all together (regardless of whether its idle or not).
It's not quite what you want, but "ulimit -t" will set a limit on the
number of CPU seconds a process can have. Combined with TMOUT you might
get close to what you want. A good alternative would be to hack the
source of bash and pop in a hard time limit there: just have the shell
monitor the difference between now and the start time just before it's
main select(). That might be enough to satisfy people, though it's
still amenable to being broken. (You've got to be careful going for
more complex schemes to make sure you don't terminate processes early,
sometimes much too early.)
Personally, I don't believe in having idle timeouts -- the problem
they're trying to solve is either mythical or can be better solved in a
different way.
jch
--
Taroon-list mailing list
Taroon-list@(protected)
http://www.redhat.com/mailman/listinfo/taroon-list
Earn $52 per hosting referral at Lunarpages.
|
|
|