Sirs, I guess this is an FYI - This is the third time I've seen this problem. So, I thought I'd let you folks benifit from my experience...
See the two attached files to see waht the Cluster Summary char was doing. Yes, ugly. So, restarted gmetad and httpd service gmetad restart service httpd restart No, Joy. I looked at the Summary database: cd /var/lib/ganglia/rrds/__SummaryInfo__/ rrdtool dump cpu_num.rrd |more From this I saw the actual numbers in the database where jumping all around. Ok, bizarre. Now I was blowing up the source tarball, upping the debug messages in gmetad and configure, make'ing it. When I noticed: nobody 699 1 0 Oct08 ? 00:24:28 [gmetad2] (Today's the 13th) Was running on the ganglia host. WTF! What is gmetad2? Anyways, killed that bugger. Restarted gmetad - Happy, Happy Joy Joy. Don't know why I didn't notice that process sooner. I think I saw it.... But, assumed it was a sub-processes generated by gmetad that would exit with gmetad exited during a service restart.
<<inline: graphcpu.php.gif>>
<<inline: graph.php.gif>>
