As a tech journalist who's covered enterprise IT for more than a decade, I've heard quite a few variations on the old "no one ever got fired for buying IBM" tech industry maxim. Over the years, IT managers have told me they dodged the pink slip by aligning their companies with EMC, VMware and Microsoft. So I wasn't that surprised when an enterprise IT manager recently told me, "no one ever got fired for going with AWS."
Entrusting Amazon Web Services (AWS) with all of your IT business seems a bit premature at this stage in the game, but I'll let that stand. It is, after all, the 800-pound gorilla in the cloud computing space. And if you're committed to a single cloud vendor, it's probably as good a choice as any.
Just because an organization standardizes on AWS, however, doesn't mean this 800-pound gorilla shouldn't be held accountable -- and kept on its toes.
One way is to lay the groundwork for a multi-cloud strategy.
Cedexis is the company behind the Radar Community, a free service that monitors major cloud and content delivery network providers including AWS, but also Joyent, Google Compute Engine, Microsoft Azure, Rackspace and IBM/SoftLayer, among others. Cedexis also offers the paid OpenMix load-balancing service that keys off Radar performance data so workloads can be balanced or moved across cloud regions, data centers and even between clouds -- depending on performance. Maintaining the flexibility to move a workload to a different cloud if needed sounds like a good idea, but something few do. OpenMix's primary competition comes from AWS Route 53, a DNS service that connects user requests to other resources … in AWS, said Rob Malnati, vice president of marketing and business development at Cedexis.
So much for managing risk.
I can understand that an organization would choose not to go down a multi-cloud route. It can be complicated, time consuming, expensive and error-prone. But that doesn't mean you should let your primary cloud provider off the hook. At the very least, cloud consumers need to monitor the performance of resources running in AWS, as well as the performance that consumers of those applications experience. In a perfect world, that data would feed into a common portal such that AWS resources can be managed side-by-side with workloads that are running (gasp!) on-premises.
Luckily, there's no shortage of companies that provide some level of cloud monitoring, including CopperEgg, NewRelic, Keynote, Site24x7 and CA Nimsoft, to name a few. The key is to pick one. Because no one ever got fired for knowing what the hell is going on.
Alex Barrett is editor in chief of Modern Infrastructure. Write to her at email@example.com.
Dig deeper on Amazon EC2 (Elastic Compute Cloud) management