Understanding Uptime
Friday, June 19, 2009 at 6:04PM | Print Article In the age of "on-demand services" uptime is everything! This is true whether you are the provider or the customer. For example, if you are an e-commerce retailer (such as Play) your online infrastructure is critical, if your service goes down, even for a few minutes, it could result in a huge loss of custom. The same applies to being the customer, Software as a Service (SaaS) or cloud computing is becoming more and more attractive to enterprise customers, whether it be e-mail and calendaring through providers such as Google or outsourcing your entire server farm to large corporate players such as Amazon or AT&T.
Therefore it pays any budding IT/Business analyst to understand the real world difference between 99% uptime and 99.999% (also known as Five Nines) uptime. The table below shows the breakdown of uptime and what that means in regards to downtime over a one year period.
| Uptime | Downtime (over one year) |
|---|---|
| 90% | 876 hours (36.5 days) |
| 95% | 438 hours (18.25 days) |
| 99% | 87.6 hours (3.65 days) |
| 99.9% | 8.76 hours |
| 99.99% | 52.56 minutes |
| 99.999% | 5.256 minutes |
| 99.9999% | 31.536 seconds |
As you can see from the table, the difference between 99% (which sounds great in principle) and 99.999% is in fact 87.5 hours of downtime over one year. That means that .999% has just cost you three days of service. As a result whenever I am dealing with a company that is going to host a service on my behalf, I always take note of how many nines they are offering and aim to negotiate as close to five nines (the sweet spot) as possible.