neigh_periodic_timer (neigh_table_init_no_netlink) excessive timer wakeups
Andreas Mohr
andi at lisas.de
Wed Mar 26 02:44:33 PDT 2008
Hi,
On Wed, Mar 26, 2008 at 10:09:12AM +0100, Andreas Mohr wrote:
> I'll do a HZ=100 test run now and possibly further code investigation
> as time permits, however one thing is obvious: compared to other
> timer system activity this one is very visible and should be tackled
> one way or another.
OK, with HZ=100 I now always get _exactly_ 50,0 neigh table wakeups:
38,6% ( 50,2) lisa : neigh_add_timer (neigh_timer_handler)
38,5% ( 50,0) <kernel core> : neigh_table_init_no_netlink (neigh_periodic_ti
(and I conveniently managed to catch the rare lisa activity here as
well)
Not sure what this different number should tell me, though
(after all it's not directly proportional to the 100 <-> 300 change).
The only way for me to gain sufficient insight about this would be to dive
into the inner handling and interval calculation, tree size etc. again.
Andreas Mohr
More information about the Power
mailing list