Yet even the mighty Amazon — or, more accurately, its cloud services unit, Amazon Elastic Compute Cloud — collapsed in April for about 24 hours, taking with it customers like Reddit, foursquare, Hootsuite and hundreds of other websites, some of them run by small businesses.
The outage shows that, no matter the size and brand name, cloud services aren’t 100-percent trustworthy. If your business relies on a cloud service, particularly if you’re in the type of industry that can’t tolerate any outages, it’s imperative to come up with a “Plan B.”
Mitigate the Risks of Cloud Computing
Patrick Gray, president of Prevoyance Group, an IT consulting firm, says there’s an obvious appeal to working with a known cloud provider, but if you’re a small company, don’t expect great customer service. “They might gently giggle under their breath if you call to complain,” says Gray. The alternative is to work with a smaller, hungrier provider, but then your chances of a blackout are probably greater.
In other words, there’s no safe haven in cloud computing, so the best you can do is prepare yourself. You can’t ignore the advantages the cloud offers, but you can learn to mitigate the risks. Gray and other experts suggest taking the following precautions:
- Carefully review cloud service contracts. Dazzled by cloud technology, most small to midsize businesses accept boilerplate contracts that protect the cloud purveyors from any liability. “Instead of accepting those, push for protections against service outages,” says Gray.
- Do your homework about the cloud service vendor. Get a third party to check them out. Ask about their past incidents, outages and downtime. “Make sure their levels of service match your business requirements,” says Charles King, principal analyst of Pund-IT.
- Simplify your use of cloud services. Evan Cooke, co-founder and CTO of Twilio, an API provider, noted in a blog post after Amazon’s cloudpocalypse that Twilio wasn’t affected. One major reason for this is because Twilio built simple services run by a single host rather than ones using multiple hosts.
- Build redundancies into your system. Cooke recommends having software quickly identify failures and retry requests. “By running multiple redundant copies of each service, one can use quick timeouts and retries to route around failed or unreachable services,” he writes.
Be a Smart Cloud Shopper
Of course, not every small business has the luxury of rejiggering their software in case of a cloud service outage. Until now, most people have expected services from big cloud purveyors like Amazon to work all the time. Since that will probably never be a reality, the best approach is to try to wring as many concessions as possible out of your cloud partner before you sign a contract. As King says, “This is just like any other kind of shopping.”