I haven't been watching it too closely, but I just did some bonnie
comparisons on my 2.2.15 kernel, and my 2.4.0-test2 kernel, and here's
what I got for results:
-------Sequential Output-------- ---Sequential Input--
--Random--
-Per Char- --Block--- -Rewrite-- -Per Char- --Block---
--Seeks---
Machine MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec
%CPU
2.2.15 512 5960 96.0 18136 25.5 6849 19.4 6906 99.0 14918 11.7 234.9
3.2
-Per Char- --Block--- -Rewrite-- -Per Char- --Block---
--Seeks---
Machine MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec
%CPU
2.4.0-te 512 5750 96.5 12156 20.4 5284 9.2 5530 80.8 10905 9.3 180.5
2.8
I appologize for the crappy formatting...
Notice however that for block I/O, the new kernel only does about 2/3 the
I/O the 2.2.15 kernel does on the same hardware. Is this a known
performance issue with the new kernel, or are the bonnie tests not
accurate for some reason, or what?
My kernel configuration is about the same, but I compiled all the block
device support into the kernel on the 2.4.0-test2, whereas it was
modularized on the 2.2.15 kernel (except for SCSI support, the Tekram
driver, and SCSI disk support, since my root partition is on the tekram).
Hardware the tests ran on:
Dual CPU celery 400MHz
Tekram 390U2B using symbios driver
256MB RAM
IBM 9GB 10,000RPM LVD
Adaptec 2940 (for CD writer and DVD player)
8GB WDC IDE drive (for Winblows -- gotta play Halflife...)
The filesystem I ran it on was my /opt partition; df output shown below:
Filesystem 1k-blocks Used Available Use% Mounted on
/dev/sda1 149712 44582 97400 31% /
/dev/sda3 248911 164121 71939 70% /home
/dev/sda7 4987112 1629668 3104108 34% /opt
/dev/sda6 3023744 1919536 950612 67% /usr
/dev/sda5 298455 44341 238704 16% /var
/dev/hda1 665348 417928 247420 63% /dos/c
/dev/hda5 5619748 2136492 3483256 38% /dos/d
I was not expecting such a drastic difference in performance, and
definitely not expecting it to go DOWN...
-- PGP/GPG Public key at http://cerberus.ne.mediaone.net/~derek/pubkey.txt ------------------------------------------------------ Derek D. Martin | Unix/Linux Geek derekm@mediaone.net | ddm@MissionCriticalLinux.com ------------------------------------------------------- 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/
This archive was generated by hypermail 2b29 : Mon Jun 26 2000 - 21:00:06 EST