Date: Tue, 17 Oct 2000 15:33:10 -0400 From: Mike Tancsa <mike@sentex.net> To: Gerhard Sittig <Gerhard.Sittig@gmx.net>, freebsd-stable@FreeBSD.ORG Subject: Re: PERC2 RAID support in 4.1-STABLE Message-ID: <5.0.0.25.0.20001017151619.06bfbb10@marble.sentex.ca> In-Reply-To: <20001017200318.E25237@speedy.gsinet> References: <5.0.0.25.0.20001017090048.06f1eb60@marble.sentex.ca> <5.0.0.25.0.20001012111508.05f2c0b0@marble.sentex.ca> <Pine.BSF.4.21.0010171339330.44225-100000@plato.salford.ac. uk> <5.0.0.25.0.20001017090048.06f1eb60@marble.sentex.ca>
next in thread | previous in thread | raw e-mail | index | archive | help
At 08:03 PM 10/17/00 +0200, Gerhard Sittig wrote:
>Where can I learn more about the pitfalls? I have a model 466
>waiting to get employed (RAID5 config with three IBM drives) and
>get very poor throughput in the 3MB/s range. I'm aware of the
>expensive directory manipulations but even "dd if=/dev/zero
>of=$FILE bs=8388608 count=..." is this slow and I don't know
>where to start searching. This has been with 4.0-R and 4.1-R,
>cvsup to -STABLE has just finished and results will be available
>soon.
>
>Is there a checklist for the things to do / to make sure? Could
>it be bad cabling rather than a software configuration thing?
I found that RAID 5 on 3 drives just sucks plain and simple. RAID 0+1 with
4 drives gave me good throughput as well as the nice comfort level that the
particular project wanted. i.e. cold swap hard drives, decent random IO etc.
The tests below are all on the same PIII 533, 256MB of RAM, with only the
drives and controllers changing (Mylex DAC960PB, 466, Promise ATA 66
controller)
Random read Sequential read Random write Sequential write
ID K/sec /sec K/sec /sec K/sec /sec K/sec /sec
ad7e 1573.7 98 1348.4 82 1804.0 112 2845.7 174
da0e 1889.2 117 1935.6 118 1611.4 99 1736.5 106
ad4c 1548.5 96 1339.3 82 1800.4 111 2745.5 168
amrd0c 1806.4 112 13499.7 824 507.7 31 6790.2 414
vinum0 1369.0 85 7067.9 431 3254.2 199 6234.9 381 0
vinum0 2934.8 183 3863.0 236 2599.9 161 4421.8 270 1
vinum0 2525.8 157 10657.7 650 3278.1 204 14671.6 895 3
amrd0e 1240.8 77 9915.2 605 412.3 26 5869.0 358 4
mlxd0c 4552.7 283 6541.1 399 2139.7 133 8128.7 496 5
466 3730.3 230 23152.7 1413 2184.6 135 2418.9 148 6
atlas466 4907.3 306 21745.5 1327 2189.5 137 2165.9 132
0=ata 66 on the same bus (master slave)
1=quantum atlast IV da0 and da1 stripe
3=ata 66 two masters
4=IBM SCSI drives
5=Atlast IV in RAID7 on Mylex
6=amrd 466, 4 drives, raid1+0
---Mike
------------------------------------------------------------------------
Mike Tancsa, tel +1 519 651 3400
Sentex Communications mike@sentex.net
Cambridge, Ontario Canada www.sentex.net
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?5.0.0.25.0.20001017151619.06bfbb10>