Hi, i have a cluster with 342 CPUS under diferents OS (linux, tru64, hpux) all running ganglia gmond, al individual grafic are ok, the numer of CPUS are well but the general number of cpus are very irregular. the graph turn to 300 to 100 in minuts and recover to 200 an turn to 50 .... very strange.
my grid conf: CRID. Superdome Compaq SVG ---Cheaptb ---Svgd ---Bw Mete SUN i think the problem is on SVG but im not sure PD: atach a graph I need urgent help :(
<<attachment: graph.gif>>
