Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 22 Apr 2001 11:48:36 -0400
From:      Seth <aberrant@superconductor.rush.net>
To:        freebsd-questions@freebsd.org
Subject:   ATA100 read/write perofrmance disparity
Message-ID:  <20010422114836.A6234@superconductor.rush.net>

next in thread | raw e-mail | index | archive | help
I have a Maxtor (Promise) Ultra ATA100 Controller and a Maxtor
60GB ATA100 drive, as follows:

(typing by hand here, please let me know if something obvious is wrong, could
be a typo)

atapci1: <Promise ATA100 controller> port 0xdc80-0xdcbf,0xdcf0-0xdcf3,0xdcf8-0xdcff,0xdcd8-0xdcdb,0xdce0-0xdce7 mem 0xfe000000-0xfe01ffff irq 11 at device 16.0 on pci0

ad4: 57241MB <Maxtor 36147H8> [116301/16/63] at ata2-master UDMA100

I also have an Adaptec 2940 SCSI controller that has my CD-ROM drive
attached (irq 10), and two Netgear FA311's (not enabled).  There are no
devices attached to the motherboard's IDE connectors, and they've been
disabled in the BIOS.

The machine is a dual-proc Pentium Pro 200 w/ 128 MB RAM, running 
4.3-RELEASE WITHOUT SMP enabled (right now).  I've confirmed that SMP
does not affect this test.

I've created a 2 GB /usr filesystem to test, and have enabled softupdates:

/dev/ad4s1e on /usr (ufs, local, softupdates)

My performance with this drive is disappointing relative to the SCSI-1
devices I used to have in this machine.  (The old disks were Seagate
4.3 GB SCSIs running at 10MHz, 10MB/sec transfer.)

using dd to test performance,with the new drive, I get
dd if=/dev/zero of=/usr/foo bs=16384 count=16384
16384+0 records in
16384+0 records out
268435456 bytes transferred in 34.839972 secs (7704813 bytes/sec)

dd of=/dev/null if=/usr/foo bs=16384
16384+0 records in
16384+0 records out
268435456 bytes transferred in 10.272343 secs (26131863 bytes/sec)

systat -iostat 1 shows ad4's tps around 55-60 for the write to /usr/foo,
and around 200-206 for the read from /usr/foo.

sysctl reports the following:

hw.ata.ata_dma: 1
hw.ata.wc: 0
hw.ata.tags: 0
hw.ata.atapi_dma: 0
hw.atamodes: ---,---,dma,---,---,---


Now, the Maxtors ship with a performance-limiting config: write verify is on
for the first 10 power cycles, and the acoustic management is on as well.
I've disabled both of these features and have also run the ATA66-ATA100
update.

Is this the expected performance for this controller/drive on this system?
Frankly, I was expecting better write performance -- this is about the same
as my old SCSI disks (although the reads look much better).

Any ideas would be much appreciated.  FYI, during the write test,
'top' in another window took 6 seconds to come up, and there were
significant (6-7 second) delays between "login" and "password" on the
other vtys.

Thanks,

Seth.

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20010422114836.A6234>