netdev
[Top] [All Lists]

Re: [RFC] TCP congestion schedulers

To: John Heffner <jheffner@xxxxxxx>
Subject: Re: [RFC] TCP congestion schedulers
From: Rick Jones <rick.jones2@xxxxxx>
Date: Fri, 08 Apr 2005 13:20:33 -0700
Cc: netdev@xxxxxxxxxxx
In-reply-to: <Pine.LNX.4.58.0504081512180.20627@xxxxxxxxxxxxxx>
References: <421CF5E5.1060606@xxxxxxxxx> <20050223135732.39e62c6c.davem@xxxxxxxxxxxxx> <421D1E66.5090301@xxxxxxxx> <421D30FA.1060900@xxxxxxxxx> <20050225120814.5fa77b13@xxxxxxxxxxxxxxxxx> <20050309210442.3e9786a6.davem@xxxxxxxxxxxxx> <4230288F.1030202@xxxxxxxxx> <20050310182629.1eab09ec.davem@xxxxxxxxxxxxx> <20050311120054.4bbf675a@xxxxxxxxxxxxxxxxx> <20050311201011.360c00da.davem@xxxxxxxxxxxxx> <20050314151726.532af90d@xxxxxxxxxxxxxxxxx> <m13bur5qyo.fsf@xxxxxx> <Pine.LNX.4.58.0504081512180.20627@xxxxxxxxxxxxxx>
Sender: netdev-bounce@xxxxxxxxxxx
User-agent: Mozilla/5.0 (X11; U; HP-UX 9000/785; en-US; rv:1.6) Gecko/20040304

For the record, here are some benchmarks from an ia64 over GigE.  I
set the MTU to 564 so it actually stressed the CPU.  Numbers are
throughput (10^6 bits/sec).

Command line used: netperf -H 192.168.1.3 -l -1000000000 -c -C -v 2.

I so rarely see anyone use the byte count limits for -l - nice to know they still work :)

FWIW, the argument to -l is passed through netperf's "convert()" routine which means you can use K|M|G for powers-of-two kilo, mega and giga; or k|m|g for powers of ten:

    netperf -H 192.168.1.3 -l -1g -c -C -v 2

The sender was a 1-CPU 900 MHz Itanium2.  The receiver was a 1-CPU 2.4 GHz
Pentium 4.  The sender reported over 99% utilization; the receiver
reported about 50%.  The NICs were both fiber SysKonnect 9843's connected
back to back.

   Normal reno    Modular reno
        392.77          392.59
        393.96          393.66
        393.22          393.72
        393.12          393.81
        392.09          393.37
        393.3           393.58
        391.81          393.22
        393.11          394.1
        391.32          393.77
        392.94          393.03

average 392.76          393.49
stdev   0.79            0.44

Looks like noise.  Fair enough.

rick jones

back to trying to figure-out why netperf IPv6 tests and ping6 won't work with local scope addresses but will with global...


<Prev in Thread] Current Thread [Next in Thread>