Dipankar Sarma a écrit :
On Tue, Mar 21, 2006 at 04:37:19PM +0100, Eric Dumazet wrote:
Jesper Dangaard Brouer a écrit :
There is definitly high memory pressure on this machine!
Slab memory usage, range from 39Mb to 205Mb (at the moment on the
production servers).
Did you tried 2.6.16 ?
It contains changes in kernel/rcupdate.c so that not too many RCU elems are
queued (force_quiescent_state()). So in the case a rt_cache_flush is done,
you have the guarantee all entries are not pushed into rcu at once.
Well, memory pressure or not, the oopses shouldn't be happening :)
Perhaps we should look at them before we work around memory
pressure through the rcu batch tuning stuff in 2.6.16 ?
One of the oopses looked like the rcu callback function pointer
getting corrupted indicating that it was double freed or
problem with RCU itself.
Yep, but as this is a production server, I believe its owner might want a fast
way to have it back to life :)
And the RCU change in 2.6.16 is definitly a big improvement when a dump of
million entries is done :)
If RCU quiescent state is forced, maybe the bug (in the route cache code) will
trigger faster ?
Eric
-
: send the line "unsubscribe linux-net" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
