[email protected] said: > One thought I had was to make a small cluster of several VM servers and > several bare metal ones and add all of them to some clients and simply > observe deltas in ntpq stats.
I've been pleasantly surprised at how well ntpd works on VM. loopstats/peerstats/rawstats would probably work better than ntpq. I would set up the servers you described, and one more monitoring system. On the monitoring system, I'd use the normal servers that the test boxes are using, and then add all the test servers with noselect and turn on loopstats, peerstats and rawstats. Then collect data for a while and graph the offset you observe from rawstats on the monitoring server. If all the offsets are the same, that probably means the clock on the monitoring server is off. The eye is pretty good at analyzing that sort of data, at least after you get the graphs setup right. -- These are my opinions. I hate spam. _______________________________________________ pool mailing list [email protected] http://lists.ntp.org/listinfo/pool
