The High Cost Of Cloud Service Provider Downtime
With so many businesses moving over the cloud, even just a few minutes of service provider downtime can have a massive effect not only on individual businesses, but on the economy as a whole. Over the last five years, there has been an estimated 568 hours of provider downtime, resulting in an economic impact of $71.7 million. The average business loses an average of $300,000 per hour in an outage, with brokerage firms missing out on approximately $6.5 million per hour. Clearly, it’s time to take a good look at your provider and decide if they are giving you the best possible service.
Of course, most providers issue Service-Level Agreement (SLAs) to their customers, but these agreements all too often give little more than an illusion of safety. Most providers state they have 99.95% availability and will pay out if they fall below that level, but as stated in this whitepaper, that number can be misleading. On one hand, this percentage could either be averaged out monthly or over the entire year, allowing a four hour outage once a year without payout. As stated earlier, that much downtime would cost the average business over a million dollars.
Clearly, relying on SLAs simply isn’t enough. Instead, the prudent end-user needs to do some in depth research before signing a contract with a provider. First steps naturally involve diving into the provider’s history of outages and how they were handled. Failure to follow best practices when responding to an outage is a good indication that problems in the future are likely to reoccur. One provider simply failed to switch over to backup generators in a power outage, prolonging the outage considerably. Such a failure suggests the provider has failed to have emergency procedures in place or has left improperly trained staff on duty, both of which don’t bode well for the future.
In an age of increasing frequency in natural disasters, many SLAs state that they are not responsible for any outages caused by factors outside their control. Hurricanes, floods, tornadoes, and earthquakes could shut data centers for an unacceptable amount of time with no payout from the SLA. If your SLA does not cover you for natural disasters, it’s important to make sure the provider has multiple data centers spread out geographically and can store backups of all your information.
And so it’s time to take a good look at our providers and decide if they really are the right match for us. While migrating to a different provider can be costly, in the end uptime is king. If someone else can guarantee fewer outages, higher levels of redundancy, and a faster response time when outages inevitably happen, the cost of migration pales in comparison. If you’re unsure how to proceed from here, be sure to read this paper, which breaks down uptime standards into six basic characteristics.
By Nick Kleeman