Re: oltp ~10% regression with 2.6.27-rc5 on stoakley machine

[Lin Ming <ming.m.lin@intel.com>]

> > > Thats bizarre... that just indicates the better clock, which should give
> > > better (read fairer) scheduling hurts your workload.
> > > 
> > > Is there anything I can run to see if we can fix the scheduler perhaps?
> > 
> > I observed schedstats of sysbench, there's more
> > "nr_failed_migrations_hot"
> > 
> > 2.6.27-rc4: se.nr_failed_migrations_hot 11
> > 2.6.27-rc5: se.nr_failed_migrations_hot 95
> > 
> > task migration failed because of task_hot, the system is un-balanced?
> 
> Ah, that makes sense, a more accurate clock could indeed make more tasks
> hot.
> 
> Can you try fiddling with: /proc/sys/kernel/sched_migration_cost ?

sched_migration_cost		regression
----------------------          -------------
50000                           ~6%
0				~8%
500000 (default)		~10%
5000000                         ~14%
-1				~19%
			
> 
> Also, we used to have some auto-tuning in there, which dissapeared some
> time ago, gregory brought it back to live recently, perhaps he likes to
> share? :-)
>