XP10000 Performance Problems

[ICF]

Hi,

we want to migrate our main database from XP128 SANs to XP10000. Our developers has written a test pl/sql script, which tests a tablespace which resides on only one RAID Group (4 HDDs, Raid 5 3D+1P). On XP128, I need about 44sec, on XP10000 I need 53sec, which means, the new XP10000 is about 15% slower.

What are we doing wrong, why is XP10000 slower than XP128.

Both tests are made with no clients connected except the test machine.

Another mystery is that testtool says a striped lvol residing on two raidgroups has the same throughput than a lvol residing on only one Raidgroup (raw devices measured).

Can anybody please help me ?

Thanks

bk-ffm.de