I have a workstation with a Gigbyte board and an Intel Corporation 82801EB (ICH5) SATA Controller (rev 02) that's apparently getting bitten by https://bugzilla.redhat.com/show_bug.cgi?id=502499
It's running CentOS 5.4 32 bit, and the user is suffering from high load and sluggish response whenever a disk write of more than a few MB is made. iostat shows the disk at 100%.
[root@marge ~]# uname -a Linux marge.gizmopartners.com 2.6.18-164.6.1.el5 #1 SMP Tue Nov 3 16:18:27 EST 2009 i686 i686 i386 GNU/Linux
[root@marge ~]# hdparm -i /dev/hda
/dev/hda:
Model=WDC WD2000JB-00GVC0, FwRev=08.02D08, SerialNo=WD-WCAL83129186 Config={ HardSect NotMFM HdSw>15uSec SpinMotCtl Fixed DTR>5Mbs FmtGapReq } RawCHS=16383/16/63, TrkSize=57600, SectSize=600, ECCbytes=74 BuffType=DualPortCache, BuffSize=8192kB, MaxMultSect=16, MultSect=16 CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=268435455 IORDY=on/off, tPIO={min:120,w/IORDY:120}, tDMA={min:120,rec:120} PIO modes: pio0 pio1 pio2 pio3 pio4 DMA modes: mdma0 mdma1 mdma2 UDMA modes: udma0 udma1 udma2 udma3 udma4 *udma5 AdvancedPM=no WriteCache=enabled Drive conforms to: Unspecified: ATA/ATAPI-1 ATA/ATAPI-2 ATA/ATAPI-3 ATA/ATAPI-4 ATA/ATAPI-5 ATA/ATAPI-6
* signifies the current active mode
Any suggestions welcomed.
Thanks!
--Chris
On 17/11/09 17:12, Chris Boyd wrote:
I have a workstation with a Gigbyte board and an Intel Corporation 82801EB (ICH5) SATA Controller (rev 02) that's apparently getting bitten by https://bugzilla.redhat.com/show_bug.cgi?id=502499
It's running CentOS 5.4 32 bit, and the user is suffering from high load and sluggish response whenever a disk write of more than a few MB is made. iostat shows the disk at 100%.
I've seen vibration impact disk i/o on sata disks as well. Make sure this is not your major problem ( eg. run the disks outside the machine for a few days - depending on your environment, that might be possible ).
-KB
On Nov 17, 2009, at 11:28 AM, Karanbir Singh wrote:
I've seen vibration impact disk i/o on sata disks as well. Make sure this is not your major problem ( eg. run the disks outside the machine for a few days - depending on your environment, that might be possible ).
I saw that page too :-)
There's another WD disk in the box on a different controller that does not suffer from the problem.
--Chris
On 17/11/09 17:31, Chris Boyd wrote:
I've seen vibration impact disk i/o on sata disks as well.
There's another WD disk in the box on a different controller that does not suffer from the problem.
Are they identical ? swap them around in the chasis and see if that makes any difference.
I had a 4 disk raid10 built on samsung fd2 1.5TiB disks, getting ~ 123 to 128MiB/sec. Moving them from the ich5 machine to a fanless Via Esther ( 1200Mhz ) based box took me upto 240MiB/sec and its the CPU here that cant go faster. Same rig/kit/rack point/psu - just a diff mobo with all the fans from around the mobo removed. Just the ones in the psu churn enough air to keep the disk <= 36 Deg C.
- KB
On 11/17/2009 06:12 PM, Chris Boyd wrote:
I have a workstation with a Gigbyte board and an Intel Corporation 82801EB (ICH5) SATA Controller (rev 02) that's apparently getting bitten by https://bugzilla.redhat.com/show_bug.cgi?id=502499
...
[root@marge ~]# hdparm -i /dev/hda
SATA and hda?
Can you boot with the additional option:
hda=noprobe
and see if you get the disk as sda and better performance?
Mogens
-
Chris Boyd -
Karanbir Singh -
Mogens Kjaer