uptime
Uptime is difficult to measure, and we've put a lot of effort into deciding how best to track it for our customers. The most useful tools from an administrative perspective are applications like Nagios and Zabbix that compile statistics we can use to tune servers and preemptively diagnose problems. The problem is that the uptime measurements these tools provide tracks availability from the monitoring server to the monitored server (which means within our network), and that's it.
This probably is not the best way to compile data that's meaningful to our clients.
In the end, we decided to use third party measurements to track our availability from outside our network. The following measurements should give you a good feel for our performance.
Summary:
This graphic summarizes the measurements of our client servers and averages them to compute a single uptime measurement. This is a free service that measures availability from worldwide locations (including Paris and Moscow, which tend to run slower than checks from Texas), and performs checks every minute.
Another View:
This graphic is provided by another uptime tracker, using a proprietary tracking method.
content on this site © 2002-2006 WELLBUILTNETWORKS LLC