Yes, but that's really no different than without limits,
as if you did not set a limit but just wait a little longer
it will reach the real system limits, and have the same
problem, but there'll be less you can do about it.

The difference is how much this affects other processes.


There's actually a couple of cases here -- the physical
memory resource managed via rcapd and virtual memory
which includes swap space -- if you limit physical memory
it will page out unused pages earlier than it would otherwise,
but keep allocating new memory until it also runs out of
virtual memory -- if you also limit virtual memory the unlimited
process growth will be stopped before the system as a whole
runs out (the process will most likely die unless it takes special
action). If the system runs out of virtual memory other processes
will start to die, you will get fork failed messages, etc.

The best solution is to fix the memory leak, and use the resource
controls to manage interapplication competition rather than
expecting them to stop the application from breaking itself.
Of course, you often don't have access to the code to let you
do that, so you end up doing whatever you can :-)
 
 
This message posted from opensolaris.org
_______________________________________________
opensolaris-discuss mailing list
[email protected]

Reply via email to