Netra SPARC T4-1 disk problem?

hi,experts. i found my application response time is more longer than before. this application is running on two NETRA T4-1 node by vcs cluster. when i switch to the other node and the problem disappear. i checked the network and some os relation log file did not find helpful information. who can give me some advice. thanks.

before i did the dd test i already switch to the other node(ahost)

oot@ahost # time dd if=/dev/zero of=/ddtest bs=1024000 count=2048

2048+0 记录进入

2048+0 记录调出

real    0m17.969s

user    0m0.014s

sys     0m3.435s

root@ahost # 

root@ahost # time dd if=/ddtest of=/dev/null bs=1024000 count=2048