Hi everyone, I using OCFS2 as NFS Server. When a large file(10GB) copy is performed simultaneously , performance deteriorates remarkably. For example, when the NFS client of machine1 and machine2 exists, the file of 10 GB of each of test1, test2, and test3 is created to /work by which the NFS share was carried out. The following two commands are simultaneously executed in this state machine1, cp /work/test1 /work/test1-2 cp /work/test2 /work/test2-2 in machine2, the following command is executed further. cp /work/test3 /work/test3-3 OCFS2 Although 1.4.7 has taken about the time which the copy of test3 takes about 10 minutes from 5, it may require for the copy of test1 and test2 which is carrying out simultaneous execution by machine1 40 minutes or more. Moreover, kernel-2.6.32-100.0.19.el5 is used and it is OCFS2 1.6.3 is used, it is improved a little and the copy of test1 and test2 is ended in 20 to 25 minutes. however -- the case where XFS and ext3 are used -- test1, test2, and test3 -- it ends within 10 minutes, respectively. Is there any method of something improving?