slow performance on MSA1000

[Luan Trieu Xa]

hi all,
I have got a small cluster with 2 ML570G3 with QLA HBA (281541-B21); MSA1000 (201723-B22); MSA Fabric Switch 2/8(Includes 4 2Gb SFF-SW Trnsvr (288247-B21) with 5 HDD 72GB 15K U320 UNI HDD (286778-B22)

I run RH3.0, this is the boot message
qla2x00_detect num_hosts=0
scsi2 : QLogic QLA2312 PCI to Fibre Channel Host Adapter: bus 27 device 1 irq 97
Firmware version: 3.03.11, Driver version 7.05.00p8

blk: queue f64efe18, I/O limit 4294967295Mb (mask 0xffffffffffffffff)
scsi: unknown type 12
Vendor: COMPAQ Model: MSA1000 Rev: 4.48
Type: Unknown ANSI SCSI revision: 04
blk: queue f64efc18, I/O limit 4294967295Mb (mask 0xffffffffffffffff)
Vendor: COMPAQ Model: MSA1000 VOLUME Rev: 4.48
Type: Direct-Access ANSI SCSI revision: 04
blk: queue f64efa18, I/O limit 4294967295Mb (mask 0xffffffffffffffff)
scsi(2:0:0:0): Enabled tagged queuing, queue depth 16.
scsi(2:0:0:1): Enabled tagged queuing, queue depth 16.
Attached scsi disk sda at scsi2, channel 0, id 0, lun 1
resize_dma_pool: unknown device type 12

the system have a very bad IO performance. the system run Oracle EBS 11.5.10. when a request run, the io is about 60-70 % which lead to a very low performance application.

please help me
thanks in advance

[MStoumba]

Just a thought. What cluster size did you choose when you setup the array ?

[Luan Trieu Xa]

thank Mike,
I am using ocfs1, I do not know how to find information cluster size and block size. but I think the default cluster size is 4K and block size is 128 bytes.
the fie attached is the ocfstool show
and this is blockdev ouput
root@oracle1 root]# blockdev --report
RO RA SSZ BSZ StartSec Size Device
rw 120 512 1024 0 569051648 /dev/sda
rw 120 512 512 63 569038302 /dev/sda1