On Sat, 17 Jun 2000, RJ Atkinson wrote:
I've been able to get DAMN NEAR line rate with sendfile(2) zero-copy transfer, and some optimizations for the entire packet output path, using FreeBSD and some AMD K7 800s, without using jumbo frames.
Curious. sendfile(2) is using TCP ? Any TCP extensions present ? This was off-the-shelf FreeBSD ? Which version ? Which GigE card (NetGear GA-620 ?) ?
Based off 4.0-STABLE, using some work I'm doing on the FreeBSD TCP/IP stack (mainly cleaner code and a few trivial optimizations at this point, nothing earth shattering), some additional optimizations and shortcuts through the stack based on IP Flow which I'm writing, using back to back NetGear GA620s, 512k send/recvspace buffers and a 1MB socket buffer, and a really quick in-kernel ack & discard in the receiving end.
Ah, so this was neither an off-the-shelf software configuration nor a normal user application. I see.
The last time I tried it with standard userland transfers was on back to back p3 500s which pulled about 450Mbps between a GA620 and an Intel GE.
This is consistent with what I would have expected from a normal user application on a modern UNIX over BSD Socket API for that speed of Intel CPU. Folks using CPUs with correct byte-ordering (e.g. non-DEC MIPS, SPARC) get about 1 Mbps/MHz under UNIX. Folks seem to get about 0.8 Mbps/MHz under NT/Intel (assuming a ~1518 byte MTU).
sendfile(2) is a current off-the-shelf method to improve things dramatically. The other question is what you do with the data once you receive it (in the example of an ftp, there is no equivilent recvfile(2) to do zero copy transfers). And of course, mbufs and checksums are the rest of the problem (for BSD derivatives, linux has its own very special brand of problems :P). -- Richard A Steenbergen <ras@e-gerbil.net> http://www.e-gerbil.net/humble PGP Key ID: 0x138EA177 (67 29 D7 BC E8 18 3E DA B2 46 B3 D8 14 36 FE B6)