The only gmond I am restarting is the one on the management node. I check the management node problem two ways, one is with the web frontend running on a system completely outside the clusters here, and the second is with the /opt/ganglia/bin/ganglia tool that is part of the ganglia python client installed on the management node. I run this one the management node with a "--clustersize" argument. Instead of returning 135 it returns 1
Testing this with the ganglia python client on the problem management node eliminates the effects of any issues with the gmetad part of the system. -----Original Message----- >From: Bernard Li <[EMAIL PROTECTED]> >Sent: Mar 21, 2006 4:53 PM >To: "Steven A. DuChene" <[EMAIL PROTECTED]>, >[email protected] >Subject: RE: [Ganglia-general] gmond stops recognizing the rest of the cluster > >> BTW, I have added a line to the /etc/init.d/gmond script to >> add a host route on the >> system with the dual network interfaces to point 239.2.11.71 >> to the network interface >> that faces to the internal network of the cluster. > >Why don't you just add that as a static route? > >BTW, when gmond on the node starts marking itself as offline, is gmond >still running? i.e. does /etc/init.d/gmond status say it's "running"? >It might be helpful to un-daemonize gmond and show more debugging >messages with that gmond process. > >Cheers, > >Bernard
