Hi, On 09/07/2010 06:45 PM, Matt Weinstein wrote: > I've just put enough instrumentation in to get some numbers.
> My device thread is running at SCHED_RR pri=66, I have approximately > 256 client threads, the time from last packet send() on REQ to last > device recv() on XREP is averaging around 40ms with a stddev of about > 36-40 (10,000 samples) > > There's limited or no variation reordering poll upcalls (reactors) > > The other direction (device XREP >> user REQ) is almost always nil. > > I haven't declared any HWMs so presumably they are infinite. > > > Linux dev2201.corp.xplusone.com 2.6.18-194.8.1.el5xen #1 SMP Thu Jul 1 > 19:41:05 EDT 2010 x86_64 x86_64 x86_64 GNU/Linux How many CPU's on the guest and how many on the host? Are you overcommiting CPU's? Overcommiting RAM? What does vmstat 1 10 look like on the guest when it's behaving this way? How does it look like on the host? regards, Solvi
signature.asc
Description: OpenPGP digital signature
_______________________________________________ zeromq-dev mailing list [email protected] http://lists.zeromq.org/mailman/listinfo/zeromq-dev
