> What concerns me is that this is forcing the system into swap instead
> of reclaiming some of the memory. I suspect this is a Redhat'ism so
> to speak, it is not behaviour I normally see, it seems to be hard to
> get things onto the swap device unless you really pound on the system.
The last time the machine was up long enough to exhibit this behavior, it had
about 100M in swap which could not manage to unswap. Summing over the SIZE
of all processes gave about 150M or so (the machine has 384M of real ram).
summing over the RSS of all processes gave about 80M, and starting anything
new forced more and more out of real ram and into swap.
> Could you try the 2.4.5 based rpm instead and see if this one exhibits
> the same behaviour?
I'm going to leave the system up for a day or two more to see if i can get
the behavior to exhibit itself again, i'm already starting to experience
slowdown, but it hasn't become severe. If the behavior does occur again, I'll
post some stats then and try the 2.4.5 based rpms.
-- Gerald
|