Here's my not as pretty graph of the last day. http://imgur.com/p5hbD

Some spikes but nothing too bad. Did the www.google.com hostname resolve to the same host over that whole period ?

Our nagios monitors to our ec2 servers regularly run into periods where they will timeout due to ec2 latency.

I don't get what he mean with 'sporadic', it looks to me more like 'systemic' would be the applicable word.

the cloudkick guy sohuld either use a consistent scale for the graph or make the difference more obvious

the cloudkick guy sohuld either use a consistent scale for the graph or make the difference more obvious