Last modified: 2014-07-18 22:37:58 UTC
After the initial instance creation, a gmond process exists: | root@toolsbeta-gmond-test:~# ps $(cat /var/run/gmond.pid) | PID TTY STAT TIME COMMAND | 9080 ? Ssl 13:25 /usr/sbin/gmond --pid-file /var/run/gmond.pid | root@toolsbeta-gmond-test:~# However, this process isn't recognized by Puppet as a ganglia-monitor service so Puppet tries to restart it on every run: | root@toolsbeta-gmond-test:~# puppetd -tv | [...] | notice: /Stage[main]/Ganglia_new::Monitor::Service/Service[ganglia-monitor]/ensure: ensure changed 'stopped' to 'running' | [...] | root@toolsbeta-gmond-test:~# puppetd -tv | [...] | notice: /Stage[main]/Ganglia_new::Monitor::Service/Service[ganglia-monitor]/ensure: ensure changed 'stopped' to 'running' | [...] | root@toolsbeta-gmond-test:~# Also, this process doesn't feed information to http://ganglia.wmflabs.org/. Killing this process and rerunning Puppet ("kill -HUP $(cat /var/run/gmond.pid) && puppetd -tv") or rebooting the instance creates a gmond process that is recognized by Puppet and correctly feeds information to http://ganglia.wmflabs.org/.
Looks to me like this only happens on Precise instances. Since I'm resolved not to build new Precise images, I'm inclined to mark this as 'wontfix' as Trusty instances look right to me.
Makes sense (having to trust you for the situation with Trusty instances, though :-)).