Date: Mon, 31 May 1999 09:34:42 -0400 (EDT) From: David Gilbert <dgilbert@velocet.ca> To: Luigi Rizzo <luigi@labinfo.iet.unipi.it> Cc: bmah@CA.Sandia.GOV, dgilbert@velocet.ca, wollman@khavrinen.lcs.mit.edu, freebsd-net@FreeBSD.ORG Subject: Re: Linux and Solaris tcp beating FreeBSD 4:1 Message-ID: <14162.36850.611680.632521@trooper.velocet.ca> In-Reply-To: <199905310309.FAA18066@labinfo.iet.unipi.it> References: <199905310017.RAA14008@stennis.ca.sandia.gov> <199905310309.FAA18066@labinfo.iet.unipi.it>
next in thread | previous in thread | raw e-mail | index | archive | help
>>>>> "Luigi" == Luigi Rizzo <luigi@labinfo.iet.unipi.it> writes: [trying to improve my packet-loss performance] Luigi> The trivial way to implement this is by encapsulating pkts, Luigi> with the disadvantage that in some cases one might hit against Luigi> the MTU of the path. As an alternative one could periodically One problem with the MTU-type solutions is that the other machine in question is only 12ms away. 36 packets transmitted, 35 packets received, 2% packet loss round-trip min/avg/max/stddev = 12.418/12.798/16.146/0.732 ms ... so there's not a lot of room for several packets to be in transit. I had a look at the code mentioned, but it's beyond me (at least at my current amount of business) to modify that code. Dave. -- ============================================================================ |David Gilbert, Velocet Communications. | Two things can only be | |Mail: dgilbert@velocet.net | equal if and only if they | |http://www.velocet.net/~dgilbert | are precisely opposite. | =========================================================GLO================ 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?14162.36850.611680.632521>