Wow. That is really amazing. It's really cool that you not only see uptime, but the graph also measures performance. --thanks

Hi Nelson, We measure from different locations in Europe and North America (7 in total). What you should keep in mind is that what we mean by the response time changes slightly depending on what kind of check type you are using. A Ping check will always show just the roundtrip (like Ping should do). The HTTP check type includes the loading of the HTML in the response time of the page (but no loading of images, etc). For example, the graph above is most likely an HTTP check, and is the recommended way for monitoring web pages since it shows that the web server actually delivers a page properly (instead of just checking network connectivity, etc). All the other error cases become implicit in this, i.e. you will find out about HTTP errors, network problems, DNS issues, etc, since they are all connected to your site working and delivering pages.

From where do you measure response time? In the past it looked like my Pingdom-measured response times were dominated by a long round trip to your monitoring hosts, but that seems to have changed recently.

That's just amazing! Not only was the responses time significantly improved, the wild fluctuations also went away. Could this user have also switched from shared hosting to dedicated hosting...