The Importance of Uptime and All Those Nines
Posted by Margaret Carnohan on March 5, 2018 7:20 am
Data center pros talk a lot about uptime and availability to describe how robust their infrastructure is. At CoreSite, we hang our flag on our consistent track record of uptime, even reporting six nines uptime across our portfolio almost every year since our IPO, including last year. This begs the question: what do those nines mean to your business, and why should you care?
What are all those nines?
Uptime is, at its core, the calculation of how often a certain resource is available across all the minutes or seconds of a given year. It's a pretty simple concept in general, but in the data center realm it can become a convoluted one.
Typically, historical uptime is measured in "nines," which is industry speak for calculations that begin at 99%. That figure represents a ratio of how many minutes out of the total minutes in a year a particular system or platform is available and running properly, which is helpful in understanding the operational performance of a data center portfolio, an individual data center, or a system or system component at a very granular level.
Phrased in terms of downtime, on the low end, two nines-99%-equates to about 7 hours and 12 minutes of downtime monthly. As the uptime "nines" increase-to three (99.9%), four (99.99%), and five (99.999%)-downtime decreases. In general, five nines is considered to be a reasonably high level of reliability. And at six nines, or 99.9999%, an average customer in a six nines portfolio would experience about 2.6 seconds of downtime monthly, or less than 32 seconds per year.
As a data center provider with a mission that revolves around uptime, we routinely run operational excellence campaigns, akin to frequent internal audits and optimization plans, and put our facilities and field operations teams through rigorous training and testing to ensure we're always pushing the highest performance standards possible.
Why does uptime matter so much?
There's an old saying that a chain is only as strong as its weakest link. In the digital era, the same can be said for data centers.
As businesses continue to develop new applications and technology to stay competitive, it becomes increasingly critical that they are up and running at all times. Large e-tailers and financial brokers have reported that milliseconds of latency can result in millions of dollars in lost revenue. Even for the average enterprise, one minute of downtime now reportedly costs close to 9,000 dollars.1
In these terms, the difference between seven hours and seven seconds of downtime is evident. This is why organizations are increasingly handing over management of data center operations to providers that have a core business centered on keeping the power and cooling environment up and running all of the time, so that customer servers can likewise keep running, and why the global third-party data center market is expected to be worth $51.8 billion by 2020.2
Data center solutions will continue to play a pivotal role as business and organizational operations evolve in the digital era. Data security and access, virtualized infrastructure to replace outdated on-premises legacy systems, and connectivity to global companies will continue driving demands for ever-higher availability and reliability.