[MPICH] any way to ask nemesis to turn-off and turn of active polling ?
Darius Buntinas
buntinas at mcs.anl.gov
Fri Dec 21 15:27:24 CST 2007
If you have P processors, and you're running P slaves and 1 master, I'm
not sure how you could have P+1 processes running at 100%. Are you
running one slave per processor, then adding an additional master?
If you have a process waiting in a blocking receive, it will show that
it's using 100% of the CPU if it has its own CPU to run on, but if that
process has to share a CPU with another process that's doing some work,
only then will you see the CPU usage of the waiting process go down.
-d
On 12/17/2007 04:48 PM, chong tan wrote:
> I am running RedHat enterprise 5. sysctl complains that
> sched_compat_yield is not known for kernel.
>
> BTW, I run the test, and both master and slaves utilize 100% of CPU.
>
> any suggestion ?
>
> thanks
> tan
>
>
>
> ----- Original Message ----
> From: Darius Buntinas <buntinas at mcs.anl.gov>
> To: chong tan <chong_guan_tan at yahoo.com>
> Cc: mpich-discuss at mcs.anl.gov
> Sent: Monday, December 17, 2007 1:23:46 PM
> Subject: Re: [MPICH] any way to ask nemesis to turn-off and turn of
> active polling ?
>
>
>
> On 12/17/2007 01:00 PM, chong tan wrote:
> > Thanks,
> > I don;t have root access to the box. I will see if I can ask sys-admin
> > to do it. I am running
> > Linux snowwhite 2.6.18-8.el5 #1 SMP
> >
> > DO you know if the broken yield got into this version ?
>
> I don't know, but you can try the master/slave programs from the
> discussion we had on sched_yield a few months ago:
>
> Master:
>
> int main() {
> while (1) {
> sched_yield();
> }
> return 0;
> }
>
> Slave:
>
> int main() {
> while (1);
> return 0;
> }
>
> Start 4 slaves first, THEN one master, and check 'top'. If it shows
> that the master is taking more than 1% or so, you have a kernel with the
> 'broken' yield.
>
> > FYI : the 'yield' people said it is not 'broken', it is in fact the
> > 'right yield'.
>
> Maybe, but Linus is on my side :-)
>
> -d
>
> >
> > tan
> >
> >
> >
> > ----- Original Message ----
> > From: Darius Buntinas <buntinas at mcs.anl.gov
> <mailto:buntinas at mcs.anl.gov>>
> > To: chong tan <chong_guan_tan at yahoo.com
> <mailto:chong_guan_tan at yahoo.com>>
> > Cc: mpich-discuss at mcs.anl.gov <mailto:mpich-discuss at mcs.anl.gov>
> > Sent: Monday, December 17, 2007 10:50:12 AM
> > Subject: Re: [MPICH] any way to ask nemesis to turn-off and turn of
> > active polling ?
> >
> >
> > Try setting the processor affinity for the "average" processes (map each
> > one to its own processor). If you have a kernel with the "broken"
> > sched_yield implementations, that may not help.
> >
> > If you have a "broken" sched_yield implementation, you can try doing
> > this as root:
> > sysctl kernel.sched_compat_yield=1
> > or
> > echo "1">/proc/sys/kernel/sched_compat_yield
> >
> > -d
> >
> >
> > On 12/17/2007 11:35 AM, chong tan wrote:
> > > Yes, in a very subtle way which has major impact on performance.
> I will
> > > try to decribe it a litle here:
> > >
> > > system has 32G, total image 35G. Load is a litle offbalance
> > > mathematically, 4X dual core, running 5 processes.
> > > 4 processes are the same size, each runs on a CPU. the last
> process is
> > > very small, about10% of others, run
> > > on a core of one of the CPU. SO 1 CPU runs 2 procs: average
> (P1)one and
> > > light one (P2).
> > >
> > > All proc do first MPI comm in a fixed algorithmic point. The 'useful'
> > > image is about 29G at that point, and should
> > > fit into the physical memory. P2 get there in a heart beat, then
> > > others., followed by P1 which took another 60+ minutes
> > > to get there. If I combine P1 and P2 into 1 process, then I don;t no
> > > see this extra delay.
> > >
> > > tan
> > >
> > >
> > >
> > > ----- Original Message ----
> > > From: Darius Buntinas <buntinas at mcs.anl.gov
> <mailto:buntinas at mcs.anl.gov>
> > <mailto:buntinas at mcs.anl.gov <mailto:buntinas at mcs.anl.gov>>>
> > > To: chong tan <chong_guan_tan at yahoo.com
> <mailto:chong_guan_tan at yahoo.com>
> > <mailto:chong_guan_tan at yahoo.com <mailto:chong_guan_tan at yahoo.com>>>
> > > Cc: mpich-discuss at mcs.anl.gov <mailto:mpich-discuss at mcs.anl.gov>
> <mailto:mpich-discuss at mcs.anl.gov <mailto:mpich-discuss at mcs.anl.gov>>
> > > Sent: Monday, December 17, 2007 8:02:17 AM
> > > Subject: Re: [MPICH] any way to ask nemesis to turn-off and turn of
> > > active polling ?
> > >
> > >
> > > No, there's no way to do that. Even MPI_Barrier will do active
> polling.
> > >
> > > Are you having issues where an MPI process that is waiting in a
> blocking
> > > call is taking CPU time away from other processes?
> > >
> > > -d
> > >
> > > On 12/14/2007 04:53 PM, chong tan wrote:
> > > > My issue is like this :
> > > >
> > > > among all the processess, some will get to the point of first MPI
> > > > communication points faster than
> > > > than other. Is there a way that I tell nemesis to start
> without doing
> > > > active polling, and then turn
> > > > on active polling with some function ?
> > > >
> > > > Or should I just use MPI_Barrier() on that ?
> > > >
> > > > thanks
> > > > tan
> > > >
> > > >
> > > >
> > ------------------------------------------------------------------------
> > > > Be a better friend, newshound, and know-it-all with Yahoo!
> Mobile. Try
> > > > it now.
> > > >
> > >
> >
> <http://us.rd.yahoo.com/evt=51733/*http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ
>
> >
> > >
> > > > >
> > >
> > >
> > >
> ------------------------------------------------------------------------
> > > Be a better friend, newshound, and know-it-all with Yahoo! Mobile. Try
> > > it now.
> > >
> >
> <http://us.rd.yahoo.com/evt=51733/*http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ
>
> >
> > > >
> >
> >
> > ------------------------------------------------------------------------
> > Be a better friend, newshound, and know-it-all with Yahoo! Mobile. Try
> > it now.
> >
> <http://us.rd.yahoo.com/evt=51733/*http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ
>
> > >
>
>
> ------------------------------------------------------------------------
> Be a better friend, newshound, and know-it-all with Yahoo! Mobile. Try
> it now.
> <http://us.rd.yahoo.com/evt=51733/*http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ
> >
More information about the mpich-discuss
mailing list