P10000 interesting behavior with remote copy

We have two 3par V400. We create a remote copy replication group for one disk.

Tests for replication links are attached.

If we test the disk with Crystal Disk Mark with Windows 2003 we obtain:

-----------------------------------------------------------------------

* MB/s = 1,000,000 byte/s [SATA/300 = 300,000,000 byte/s]

Sequential Read : 250.976 MB/s

Sequential Write : 41.326 MB/s

Random Read 512KB : 119.629 MB/s

Random Write 512KB : 55.892 MB/s

Random Read 4KB (QD=1) : 3.719 MB/s [ 907.9 IOPS]

Random Write 4KB (QD=1) : 1.128 MB/s [ 275.4 IOPS]

Random Read 4KB (QD=32) : 204.851 MB/s [ 50012.3 IOPS]

Random Write 4KB (QD=32) : 34.534 MB/s [ 8431.1 IOPS]

Test : 1000 MB [C: 76.8% (23.0/30.0 GB)] (x1)

Date : 2012/03/14 9:15:29

OS : Windows Server 2003 R2 Enterprise Edition SP2 [5.2 Build 3790] (x86)

If we stop the replication and do the test again we obtain:

-----------------------------------------------------------------------

* MB/s = 1,000,000 byte/s [SATA/300 = 300,000,000 byte/s]

Sequential Read : 338.687 MB/s

Sequential Write : 190.494 MB/s

Random Read 512KB : 187.277 MB/s

Random Write 512KB : 156.946 MB/s

Random Read 4KB (QD=1) : 6.564 MB/s [ 1602.6 IOPS]

Random Write 4KB (QD=1) : 4.204 MB/s [ 1026.3 IOPS]

Random Read 4KB (QD=32) : 216.928 MB/s [ 52960.8 IOPS]

Random Write 4KB (QD=32) : 57.501 MB/s [ 14038.4 IOPS]

Test : 1000 MB [C: 76.8% (23.0/30.0 GB)] (x1)

Date : 2012/03/14 9:07:43

OS : Windows Server 2003 R2 Enterprise Edition SP2 [5.2 Build 3790] (x86)

So, I am are very concern about all the writes if replication is started.

Has anybody an idea about this behavior?

Thanks,

Dragos

P.S. This subject has been edited. - Hp Forum Moderator