Google and uptime

For a relatively small outage that Google had a few months ago, they sure did get a lot of press.  As Google responded, “Over 1 million business customers” have come to expect 100% up-time and flawless service from Google, so I guess it should not have been a surprise that blogosphere lit up during this interruption.  It certainly set the stage to for a great discussion around up-time and reliability.  

Let’s take a deeper look at the reality of what Google provides; the chart below from the Radicati Group documents average downtime as experienced by business for several enterprise messaging systems on a monthly basis.  It’s clear that the traditional on-premise approach is at least an order of magnitude less reliable than Gmail.   

I am not one to be swayed by reports & averages, and real world experience trumps any marketing hype or consultants research, in my opinion.  In this light, we setup our own enterprise monitoring solution that tracks Google on a regular basis.  

The charts below are from monitoring our own internal Google Apps Gmail over the last few months. Keep in mind, this is actual production data, not just a simple “ping” of gmail.com.  These results are based on successful log-in actions via an automated script, performed every minute from no less than 12 internationally dispersed servers that intelligently check for full Gmail availability.  

 

If you look at things in even more detail, you can see that not only is Google delivering on the SLA, they provide a great user experience.  See the response time graph below; this is the amount of time that it takes for the login process to occur.  Google is providing sub-second response time over this entire same period.  How long does it take you to login to your current on-premise mail system?