Date: Mon, 2 Oct 2000 07:08:53 -0700 (PDT) From: Julian Elischer <julian@elischer.org> To: Jonathan Lemon <jlemon@flugsvamp.com> Cc: net@freebsd.org Subject: Re: SACK in FreeBSD TCP. Message-ID: <Pine.BSF.4.10.10010020703300.36257-100000@InterJet.elischer.org> In-Reply-To: <200010021358.e92DwbA14157@prism.flugsvamp.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Well I just look at the data flowing... we lose bursts of packets out of a single window, and of course then we get teh entire window retransmitted... because of the long RT delay, the entire retransmitted window arrives before the ack we send to say "hey we got it all now" gets to the server in the US.. (I have both ends of this transmittion so I can see that it is happenning....) (tcpdump is your friend). Of course this etra stuff queues up on the far end of my modem link and forces packets from teh NEXT window to get dropped..... Once the error rate drops below 1 packet lost for 20 or so transmitted, I start to get some data through but below that ( most of the time) I get about 500 bytes/sec throughput.. (FBSD<-> FBSD) Julian On Mon, 2 Oct 2000, Jonathan Lemon wrote: > In article <local.mail.freebsd-net/Pine.BSF.4.10.10010020644350.36257-100000@InterJet.elischer.org> you write: > >Is anyone working on a SACK (Selective acknowlegement) implementation > >for FreeBSD? > > I believe that Jayanth was working on it at one time, you could ask him. > > > >It would make a huge difference to performance out here at the edge of the > >univertse (Perth, Western Australia) > > That's what all the Australians I know claim. But is there any hard > data to back that up as well? (E.g.: does Linux perform better out > there?) > -- > Jonathan > > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-net" in the body of the message > 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?Pine.BSF.4.10.10010020703300.36257-100000>