Holy Crap, I’m Locked Into This Cloud Hosting Service!

When a cloud host goes down, the sites hosted on it go down too. When Amazon EC2 went down earlier this month, TechCrunch wrote that the entire Internet world had stopped spinning, bringing down popular web services like Reddit, Foursquare, and Instagram. A similar outage occurred again this year April.

While cloud computing giants like Google, Amazon and Microsoft claim to have a 99.9 percent uptime, it still roughly equates to about nine hours per year of so-called accepted down-time. That could certainly be frustrating. And what can you do when your cloud host goes down? Nothing much, really. Once you choose a cloud hosting service, you’re pretty much “locked” in. You can’t switch over smoothly to another server because migrating requires time and every server runs on a different platform. Even now that Slicehost is under Rackspace’s ownership, switching is still quite a task to accomplish (and I speak from personal experience here).

Despite this problem, cloud hosting has provided many benefits. It saves you costs as you effectively throw the problems and responsibilities onto the cloud host. The folks at Media Temple or Linode are a lot more capable and experienced than your multi-tasking IT staff in handling the servers. You can depend on them to get the latest software upgrade and security patches. Cloud hosting also provides scalability and flexibility – you basically pay for what you ask for. As your business scales, you can simply upgrade your subscription to more RAM, storage, and bandwidth whenever needed. All this can be done without the need to incur more hardware costs on your end or staff to support your server expansion.

I like the flexibility that cloud hosts provide. But what if we could make cloud hosting even more flexible? Say Amazon EC2 goes down; what if we could automatically move our data to another server based on the last backup, thus keeping our businesses online? If we can’t eliminate downtime, then perhaps it’s only reasonable to have a backup plan for it. Businesses, after all, don’t really care if one server is up or down as long as their own sites are up and running.

But that’s tough because, as mentioned, every server runs on a different platform. Migrating isn’t a one-click process. It takes time… and…

Thankfully, there are folks who are rolling up their sleeves to solve the problem. A couple of weeks ago, I met up with Christopher Bousigues and Eric Meyer, co-founders of Nubefy, a Singapore-based cloud computing management platform which aims to minimize the problems in the cloud server market.

Here’s the idea: Think of Nubefy as a simple WYSIWYG interface that allows you to give commands to automatically move from cloud provider A (e.g. Amazon EC2) to cloud provider B (e.g. Rackspace) if A goes down. In this way you can control all of your cloud hosters via a single interface. This removes the need to sign-in and out of every server to manage them, thus saving time.

Also, because of agents installed in each of the different hoster’s servers, Nubefy is also able to effectively compare their performance and speed up your page/application load time directing data to the best performing hoster. To do that, Nubefy needs to make migrating from one server to another seamless.

Of course, there are also several other features that Chris and Eric aren’t quite ready to share since the product is still a work-in-progress. But we will definitely be in touch with them for further updates. And hopefully, one day we all can be free from cloud lock-in and maximize the true benefit of cloud computing.