When your cloud service goes down, will your company be able to continue doing business as usual?
When you move part of your small business to the cloud, you’re giving the cloud provider control over some of your company’s data. You’re sending your data over the public Internet and storing it on a third-party server. You trust that the data you store on your provider’s network will be safe and remain accessible. At some point, though, your provider’s network will suffer an outage and you’ll be unable to access your data in the cloud for at least a short time.
Outages don’t make cloud computing unreliable or risky; you just need to be prepared. Here are some tips to keep your business running when an outage does occur.
First, be on the offensive. Don’t move anything to the cloud that your company can’t survive without, whether that’s a customer relationship management (CRM) application your salespeople use every day, a financial operation that’s run weekly, or an email service that’s used constantly throughout the workday. Most cloud providers’ outages are brief; but some, like Amazon’s Web Services outage in April 2011, can separate a company from its cloud service for several days. So, it’s a good idea to keep your business-critical applications and data on your own network. Of course, you should have a solid disaster recovery plan in place for that server, that data, and those applications.
Second, decide which apps are best suited for moving to the cloud and find out how your cloud provider supports them. Ask a lot of questions about your cloud provider’s data center architecture and disaster recovery plans. Does the provider have a redundant power source it can use in case of a power outage? Does the cloud service have a failover server should the original server fails? How quickly will you be notified in case there is an outage and how will the provider alert you?
Third, prepare for the inevitable service failure. Here are three steps you can take to get your small business ready for the cloud outage that you know will happen someday:
1. Know your service provider’s SLA. Understand the service level agreement (SLA) governing your cloud service so you know what you can expect from the provider in case of an outage. The SLA states how quickly the provider promises to recover from an outage or disaster. It also outlines what kind of compensation you can expect when you lose access to your cloud service.
2. Have a plan B. This will be different for each cloud service, depending on how important it is to your business operations. Whatever it is, your plan B must provide a temporary way to work around the outage. For a company that absolutely relies on the service, plan B might require duplicating efforts across two cloud services. For example, if you’re performing off-site hosted backups, you might decide to back up key data on two separate off-site servers; if one service fails, you’ll have the same backup preserved on the second service.
For other types of cloud service outages, plan B might simply mean notifying users that they can’t access a particular application for the time being. A salesperson might need to save a CRM file on his or her desktop and then upload it to the cloud service when it’s up and running again, for example.
3. Practice your plan B. Everyone who uses the cloud service be familiar with your plan B and how it works. Practice it. Pretend your cloud provider has an unplanned outage and see how well plan B works in real life. Try notifying your users impacted by the outage and check to make sure they were able to follow your work-around.
Despite every cloud provider’s best efforts and state-of-the-art data centers, their networks will go dark at some point, whether this is due to planned maintenance or an unforeseen disaster. This doesn’t mean you should be scared, because there are risks with non-cloud offerings as well; you just need to be prepared to keep your business running when it does happen.
Has your company weathered a cloud service outage? What advice can you share?