I have a cluster with 2 NIC card. eth1 is the inside interface talking to the other gmond running nodes.
eth1 is correctly configured in gmond.conf and I added the multicast route to eth1. gmetad + the web front-end are running on a separate machine outside the cluster Everything is running and all the nodes are correctly reported. Except that the system runs fine for 8-10 minutes. After 8 or 10 minutes the gmond running on the front node only reports itself and I have no long messages from the other nodes. I run gmond in debug level 10 on the front node and after 8 or 10 minutes I do not see any message from the other node. As anyone seen something like this before ? All my nodes are running RH 9.0 Thanks - Didier
