Re: RAID5 question.

The answer is more difficult!

I use RAID-s in 3 layers!

Layer 1 : (RAID5 in 11+1 raw disk in 1 pc) x4 (4 disk nodes)
Layer 2 : RAID1 4x 1/2 nodes (only for ability to backup complete nodes)
Layer 3 : RAID0 to 4 nodes -> 1 big 8TB disk. :-)

I use gnbd for this.
The gnbd sends only small packets, and I think, the too much readahead is
the problem, because the whole 8TB array is phisically stripped to 44 raw
disks, but NOT!
When I disable the readahead, the performace is more worse.
Now  I grow the readahead in the raw disk to 1MB, the raid5 to 10MB, the
raid1 to 1MB, and the raid0 to 8MB, and the performance is great now! :-)

(web+ftp  server for ~200-300 people at one time, downloading a lot of
small(1-60KB) and big (~15-50MB) files)

Thansk for you all!

----- Original Message -----
From: "Raz Ben Jehuda" <redacted>
To: <redacted>
Cc: <redacted>
Sent: Thursday, August 04, 2005 6:25 PM
Subject: Re: RAID5 question.

how is the disks performance ? is it OK ?

On Thu, 2005-08-04 at 18:13 +0200, djani22@dynamicweb.hu wrote:

quoted

Yes, it is there, I know it.
But its only for resync or not? :-)

----- Original Message -----
From: "Raz Ben Jehuda" <redacted>
To: <redacted>
Cc: "David Greaves" <redacted>; <redacted>
Sent: Thursday, August 04, 2005 5:56 PM
Subject: Re: RAID5 question.

quoted

take a look at /proc/sys/dev/raid/speed_limit_max . it is kilobytes if

i

quoted

quoted

recall correctly.

On Thu, 2005-08-04 at 17:45 +0200, djani22@dynamicweb.hu wrote:

quoted

Thanks a lot for you, and Raz!

The raw devices readahead I already set with the hdparm, and
/sys/block/*/queue/read_ahead_kb, but the md device is only

changeable

quoted

with

quoted

quoted

blockdev. :-)
I dont use lvm, because it is too slow for me....

But unfortunately the problem is still here for me. :(
I continue searching the bottleneck...

Thanks!

Janos

----- Original Message -----
From: "David Greaves" <redacted>
To: "Raz Ben Jehuda" <redacted>
Cc: <redacted>; <redacted>
Sent: Thursday, August 04, 2005 4:56 PM
Subject: Re: RAID5 question.

quoted

And notice you can apply different readahead to:
The raw devices (/dev/sda)
The md device (/dev/mdX)
Any lvm device (/dev/lvm_name/lvm_device)

David

Raz Ben Jehuda wrote:

quoted

read the blockdev man page

On Thu, 2005-08-04 at 16:06 +0200, djani22@dynamicweb.hu wrote:

quoted

Hi list, Neil!

I have a little question because I'v got some performance

problem...

quoted

quoted

quoted

quoted

quoted

quoted

Is the RAID5 do any type of headahead with the default config?
It is possible to disable the readahead? (if it is.)
Or just the io_sched do that on the disks?

How can I disable the whole headahead on top of the RAID5 array?

The RAID5 can do multiple small (<4kb) reads of disks for the

multithreaded

quoted

quoted

quoted

read requests?
(chunk size is 32k)

Thanks for helping!

Janos

-
To unsubscribe from this list: send the line "unsubscribe

linux-raid" in

quoted

quoted

quoted

quoted

quoted

the body of a message to majordomo@vger.kernel.org
More majordomo info at

http://vger.kernel.org/majordomo-info.html

quoted

quoted

quoted

quoted

quoted

quoted

--

-
To unsubscribe from this list: send the line "unsubscribe

linux-raid"

quoted

in

quoted

quoted

quoted

the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

-
To unsubscribe from this list: send the line "unsubscribe

linux-raid" in

quoted

quoted

quoted

the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

--
Raz
Long live the penguin

-
To unsubscribe from this list: send the line "unsubscribe linux-raid"

in

quoted

quoted

the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

--
Raz
Long live the penguin

-
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html