I have to backport the '9.2' release of 3w-9xxx to FC3 and RHEL* for our customers, and I'm curious as to if this driver set made it into mainline kernel and included into the FC4 kernel tree. Can anybody help me with this? Here is modinfo for the 9.2 kernel module:
filename: ./3w-9xxx.ko version: 2.26.03.015
On Wed, 2005-04-27 at 14:00 -0700, Jesse Keating wrote:
I have to backport the '9.2' release of 3w-9xxx to FC3 and RHEL* for our customers, and I'm curious as to if this driver set made it into mainline kernel and included into the FC4 kernel tree. Can anybody help me with this? Here is modinfo for the 9.2 kernel module:
filename: ./3w-9xxx.ko version: 2.26.03.015
Checking the 2.6.12-rc3 sources show that the current 3w-9xxx driver is 2.26.02.002.
Looks like the authors need to send some patches upstream so that it gets included in mainline.
David Hollis wrote:
On Wed, 2005-04-27 at 14:00 -0700, Jesse Keating wrote:
I have to backport the '9.2' release of 3w-9xxx to FC3 and RHEL* for our customers, and I'm curious as to if this driver set made it into mainline kernel and included into the FC4 kernel tree. Can anybody help me with this? Here is modinfo for the 9.2 kernel module:
filename: ./3w-9xxx.ko version: 2.26.03.015
Checking the 2.6.12-rc3 sources show that the current 3w-9xxx driver is 2.26.02.002.
Looks like the authors need to send some patches upstream so that it gets included in mainline.
Do either of these versions address the "Extremely high iowait with 3Ware array and moderate disk activity" bug? It's been open for just over a year & has 259 comments.
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=121434
I can confirm this bug with v2.26.02.001, fwiw.
-Jeff
On Thu, 2005-04-28 at 10:49 -0600, jeff wrote:
Do either of these versions address the "Extremely high iowait with 3Ware array and moderate disk activity" bug? It's been open for just over a year & has 259 comments.
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=121434
I can confirm this bug with v2.26.02.001, fwiw.
Most likely not as much as you'd like. That unfortunately is an artifact of 3ware's design. We never use these systems for high usage scenarios like a database server or sometimes even a home directory server. Nearline backup and slow storage is what we consider them useful for. The '9.2' release does have significant performance increases (firmware and driver need to be updated) but I don't think they'll solve the IO Wait state any time soon.
Jesse Keating wrote:
On Thu, 2005-04-28 at 10:49 -0600, jeff wrote:
Do either of these versions address the "Extremely high iowait with 3Ware array and moderate disk activity" bug? It's been open for just over a year & has 259 comments.
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=121434
I can confirm this bug with v2.26.02.001, fwiw.
Most likely not as much as you'd like. That unfortunately is an artifact of 3ware's design. We never use these systems for high usage scenarios like a database server or sometimes even a home directory server. Nearline backup and slow storage is what we consider them useful for. The '9.2' release does have significant performance increases (firmware and driver need to be updated) but I don't think they'll solve the IO Wait state any time soon.
I can't benchmark it since it's a live system, but the following helped with the iowait problem under FC2:
echo 512 > /sys/block/sda/queue/nr_requests
It was suggested on LKML that nr_requests be double queue_depth. I get much more even performance now. It no longer bogs down & goes out to lunch when I throw a DVD ISO at it, for example.
I also use: blockdev --setra 16384 /dev/sda
-Jeff
-
David Hollis -
jeff -
Jesse Keating