Ok I know this error is well known to RAC People. Has anyone encountered this set of circumstances or some sort of derivaton thereof: > rac on win2k sp6 with oracle 92060 > communication to node fails - only way to resolve is to hard boot the machine > oracle instances come down ungracefully on node that is rebooted. > when node reboots we try and start oracle instances to join back in to the > cluster and whammo!! ORA-600 [KCCSBCK_FIRST] Thankfully Oracle Note:157536 says that I can resolve this buy rebooting the cluster....duhhh. why didn't I think of that. Can anyone shed any more light on this? Is it a bug @wip or is there another way around this. I have TAR open, awating the token ..hmm upgrade to 10g response Thanks all. Steve -- //www.freelists.org/webpage/oracle-l