Date: Sat, 8 Feb 2003 10:46:04 -0800 (PST) From: John Polstra <jdp@polstra.com> To: net@freebsd.org Cc: wes@softweyr.com Subject: Re: Preferred Gigabit interfaces for -CURRENT Message-ID: <200302081846.h18Ik48O090124@vashon.polstra.com> In-Reply-To: <200302080049.00472.wes@softweyr.com> References: <15939.2823.45299.471388@canoe.velocet.net> <200302080049.00472.wes@softweyr.com>
next in thread | previous in thread | raw e-mail | index | archive | help
In article <200302080049.00472.wes@softweyr.com>, Wes Peters <wes@softweyr.com> wrote: > On Friday 07 February 2003 01:25, David Gilbert wrote: > > > > So... I need to know what GigE chipsets I should test. I recently > > tested Intel GigE cards ... with dismal results... less than half the > > packets-per-second on the (otherwise) same hardware. Small packets > > (as in DOS attacks) are a real concern here. > > Wow, this wasn't my experience at all. At my previous employer we > used Intel EEPro 1000 Server cards with the em(4) driver on FreeBSD > 4.5 with nary a hitch and excellent performance. This was on > ServerWorks chipset motherboards with P-III and P4 processors. I think I may know why the two of you formed different impressions of the device's performance. Not too long ago, the em driver was updated. One of the changes made was to completely and unconditionally disable the chip's receive interrupt coalescing logic. I performed some very cursory performance tests using small packets and found that this change caused a 25% decrease in the maximum packet rate of the device on the hardware I was using (which was pretty high-end stuff with plenty of PCI bus bandwidth). I don't remember exactly when this change was made, but you can find out with "cvs annotate". Search for "RDTR". I don't know the official reason why this change was made, but I would guess it was a work-around for a chip bug. I can't think of any other possible motivation for taking such a performance hit. John -- John Polstra John D. Polstra & Co., Inc. Seattle, Washington USA "Disappointment is a good sign of basic intelligence." -- Chögyam Trungpa To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200302081846.h18Ik48O090124>