how do i interpret these stats? good, bad, ugly, what to change to improve? 9i RAC, RH Linux, (2) nodes, OCFS on a SAN w/RAID5.

gcs forward cr to pinged instance
gcs immediate (compatible)
gcs immediate (null) converts
gcs immediate cr (compatible)
gcs immediate cr (null) converts
ksxr poll remote instances
CGS wait for IPC msg