When ssh is my "rsh_command" , it isn't terminating after executing "uptime" on the remote host. So the real server will not be removed from the cluster when the service (ex. apache) on it is down. When running "pulse -n" all goes normal. It occurs in piranha-0.4.14-1.i386.rpm
We believe this problem was already addressed in fixed in the current release of piranha (0.4.16). These RPMs are posted on http://people.redhat.com/kbarrett/HA/ The problem that was corrected was to deal with open pipes on the fork() call. By default, they also require the ipvsadm RPMs, and a kernel upgrade, in order to install. They "should" install on your existing system (without the ipvsadm upgrade) if you force it (i.e. specify "-f" on your RPM install), but if you experience any problems you'll have to upgrade before we can be of help. Can you try the updated software and tell us what happens?
No reply from user, but others that had experienced this same problem report that it is now resolved by the updated RPMs.