<HTML dir=ltr><HEAD><TITLE>[ofa-general] Re: why is CPU util/service demand so much higher withSDP than TCP?</TITLE>
<META http-equiv=Content-Type content="text/html; charset=unicode">
<META content="MSHTML 6.00.2900.3059" name=GENERATOR></HEAD>
<BODY>
<DIV id=idOWAReplyText23145 dir=ltr>
<DIV dir=ltr><FONT face=Arial color=#000000 size=2>Can you give more information on what this recv_poll module does? We are not interested in latency, but in throughput. Do we need to change this module for this setup?</FONT></DIV>
<DIV dir=ltr><FONT face=Arial size=2></FONT> </DIV>
<DIV dir=ltr><FONT face=Arial size=2>Regards,</FONT></DIV>
<DIV dir=ltr><FONT face=Arial size=2></FONT> </DIV>
<DIV dir=ltr><FONT face=Arial size=2>Koen</FONT></DIV></DIV>
<DIV dir=ltr><BR>
<HR tabIndex=-1>
<FONT face=Tahoma size=2><B>Van:</B> general-bounces@lists.openfabrics.org namens Michael S. Tsirkin<BR><B>Verzonden:</B> vr 27/04/2007 6:23<BR><B>Aan:</B> Rick Jones<BR><B>CC:</B> general@lists.openfabrics.org<BR><B>Onderwerp:</B> [ofa-general] Re: why is CPU util/service demand so much higher withSDP than TCP?<BR></FONT><BR></DIV>
<DIV>
<P><FONT size=2>> Quoting Rick Jones <rick.jones2@hp.com>:<BR>> Subject: why is CPU util/service demand so much higher with SDP than TCP?<BR>><BR>> So, while playing around with my new netperf SDP_RR test I've noticed that<BR>> a single-byte _RR test over SDP has a much higher transactions per second<BR>> (ie lower latency) than over TCP over the same HCA, but the CPU utilization<BR>> is _very_ much higher and the service demand (cpu per transaction) as well.<BR>> CPU util being higher makes sense with a higher transaction rate, but not<BR>> the increased service demand - well at least not to my experience thusfar.<BR><BR>That's expected.<BR>SDP by default uses polling aggressively to trade off service demand for latency.<BR>You can play with recv_poll module parameter to tune that.<BR><BR>--<BR>MST<BR>_______________________________________________<BR>general mailing list<BR>general@lists.openfabrics.org<BR><A href="http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general">http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general</A><BR><BR>To unsubscribe, please visit <A href="http://openib.org/mailman/listinfo/openib-general">http://openib.org/mailman/listinfo/openib-general</A><BR></FONT></P></DIV>*** Disclaimer ***<br><br>Vlaamse Radio- en Televisieomroep<br>Auguste Reyerslaan 52, 1043 Brussel<br><br>nv van publiek recht<br>BTW BE 0244.142.664<br>RPR Brussel<br>http://www.vrt.be/disclaimer<br> <br></BODY></HTML>