Thanks very much for posting your data.
Have you tried Andrea's patch
ftp://ftp.suse.com/pub/people/andrea/kernel/2.2.9_andrea3.bz2 ?
I suspect it would help your SMP performance.
See http://www.kegel.com/mindcraft_redux.html
Please do keep us informed of your progress.
(Er, you did mean 512MBytes RAM, not 512KBytes, RAM, right?)
Thanks,
Dan
> My apologies if this does not belong here. I'm not sure if this is an
> Apache problem, or a kernel problem. I'm having some big problems in which
> it appears that a single PII 400Mhz or a single AMD 400 will outrun a dual
> PII 450 at http requests from Apache. These machines were running 2.2.2,
> with a small patch from Alan Cox to make them work under smp with our bios.
> A kernel upgrade to 2.2.9 made the problem WORSE. At least we don't need
> the patch to get it to boot smp though... :) Anyway, here's some data we've
> accumulated:
>
> Testing the Linux/Apache HTTP Servers has given the following data
>
> AMD K6 400Mghz 512KRam Linux 2.0.36 Apache 1.3.6. 208.32.10.88
> PII 400 MHz 128K Ram Linux 2.0.36 " 208.32.10.90
> Dual PII/450Mghz 512KRam Linux 2.2.8 " 208.32.10.67
> Dual PII/450Mghz " Linux 2.0.36 " "
>
> Data for HTTP Server Tests: 100 1MByte mpeg files stored on local disks.
>
> Results
>
> AMD 400Mghz K6 accepts and delivers mpeg test files to 1000 simultaneous clients @ 57.6Kbits/sec.
> PII 400Mghz accepts and delivers mpeg test files to 1000 simultaneous clients @ 57.6Kbits/sec.
>
> Dual PII/450Mghz Linux 2.0.36 and 2.2.8 delivers mpeg files far less than 300 simultaneous @57.6Kbits/sec
> with associated TCP/IP errors at the client simulators 10054 (reset by peer) and 10060 (connection time out).
> Connection time in genearl from simulator to Server were tremendous,
> averaged about 5 seconds, normally viewed as error.
>
> I'm going to post this to the appropriate web server newsgroup as well.
> Sorry if this is a bit off-topic, but I think it might be kernel related,
> or at least an interaction between the kernel and Apache. Apache is version
> 1.3.3. Thanks in advance...
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/