I think Amazon has finally released what can be considered one of the most interesting enhancements to its cloud in the last few years. Not a price cut, not the next Uber-super feature for the DevOps, but a basic functionality which has been available in on-premeises virtual infrastructures for many years now: VM high availability!
After the long debate about “Pets and cattle”, now your cattle look more like pets… and this will allow to migrate legacy apps to the public cloud more quickly and smoothly than before.
About Auto Recovery for EC2
The AWS official blog describes it much better than I could do. But, long story short, an instance can be continuously monitored through CloudWatch and, if a problem arises, instead of just having the alarm go off, now the instance can also be automatically restarted. A VMware fan could easily say that it sounds very much like reinventing the wheel… and so it is, indeed! But this is new for Amazon EC2.
It’s just an addition to CloudWatch but it’s very important for all those applications written without having the cloud in mind (or before the cloud). We are pretty much talking about all those critical applications which take for granted a high availability cluster of some sort underneath.
Why it is important
Not all the applications are born with cloud in mind, nor are all the applications rewritten to be run on the cloud! And, in most cases, these applications can still be found at the core of many IT infrastructures. With this simple addition to CloudWatch we are getting much closer to having a high availability system for these old applications.
In practice, up to now there were doubts about moving some old, but critical, applications onto Amazon AWS, because of this problem. it will be much easier now.
In fact, up until just recently, If you had to grant a certain level of HA on Amazon (or any other cloud with similar characteristics), you had to design it yourself at the application level. This was not always possibile, especially if you are a small shop and you don’t have full control of the application or you can’t afford to redesign it from scratch. On the contrary, from now on you can have both your brand spanking new (web) front end alongside of your legacy clunky back-end on the cloud… All the pieces of your infrastructure next to each other and, in certain cases, avoiding some complexity.
Closing the circle
Cloud war isn’t only about price, but about features too. Amazon needs to become more appealing to traditional IT if it wants to maintain the leading role in the future of enterprise clouds.
On the other hand, vendors like IBM and VMware are not as competitive as Amazon at the moment (both in terms of prices and features), but they know very well how legacy IT works and they are actively working to meet traditional enterprise needs first!
Based on what has already happened in the last two years, I think we will be seeing similar features on Microsoft Azure and Google Cloud Platform very soon.