Hi All,

Am looking for some assistance in localising and resolving a problem I am having with a new OES server we are looking at putting into production. Specs are:

IBM SystemX 3650 M3
ServeRAID M1015 Raid Controller
8 x 300GB SAS drives in a RAID 5 Array

100MB /boot (std reiser)
15GB / (evms container)
2GB Swap (evms container)
1.9TB NSS pool (remaining disk space)

OES2 SP1 on SLES10SP3 (32 Bit)

My problem is this:

Writing:
(Compressed NSS Volume)
dd if=/dev/zero bs=1024 count=1000000 of=/media/nss/<CompressedVolumeName>/write.txt
1000000+0 records in
1000000+0 records out
1024000000 bytes (1.0 GB) copied, 45.5099 seconds, 22.5 MB/s

(Uncompressed NSS Volume)
dd if=/dev/zero bs=1024 count=1000000 of=/media/nss/<UnCompressedVolumeName>/write.txt
1000000+0 records in
1000000+0 records out
1024000000 bytes (1.0 GB) copied, 42.3887 seconds, 24.2 MB/s

(Standard Reiser Partition)
dd if=/dev/zero bs=1024 count=1000000 of=/tmp/write.txt
1000000+0 records in
1000000+0 records out
1024000000 bytes (1.0 GB) copied, 4.79056 seconds, 214 MB/s

Reading:
dd if=/tmp/write.txt bs=64k | dd of=/dev/null
15625+0 records in
15625+0 records out
1024000000 bytes (1.0 GB) copied, 2.21742 seconds, 462 MB/s
2000000+0 records in
2000000+0 records out
1024000000 bytes (1.0 GB) copied, 2.21765 seconds, 462 MB/s

dd if=/media/nss/DATA/write.txt bs=64k | dd of=/dev/null
15625+0 records in
15625+0 records out
1024000000 bytes (1.0 GB) copied, 2.52663 seconds, 405 MB/s
2000000+0 records in
2000000+0 records out
1024000000 bytes (1.0 GB) copied, 2.52691 seconds, 405 MB/s

Block sizes, for both the standard Linux partitions and the NSS pool, are set to 4k.

It doesn't appear to be hardware related, given the performance of the std Reiser partitions.

Having said that the RAID Array is set to write through.

Anyone have any suggestions?

Many thanks.