Availability in the Cloud

July 28th, 2008 by - 15,500 views

Today I came across a very interesting post, written by Mukul Kumar, which poses an intriguing topic of Cloud availability. Spawned most likely because of some “lack” of availability on Amazon’s S3 recently, Kumar discusses how companies might want to look towards creating redundancy across Cloud providers (such as GoGrid) in order to increase the availability of a company’s online presence.

All too frequently, companies look to a single solution for their hosting options, whether it be within the Cloud or using more traditional hosting methods. These single solutions can easily translate into single sources of failure, “don’t put your eggs in one basket” being the phrase the comes to mind.

Kumar illustrates some ways to make a company’s Cloud reliance a bit more redundant by using some traditional methods of choosing various providers.


Using a fairly straight-forward rsync (an open source utility that provides fast incremental file transfer) methodology, Kumar shows how servers can be configured as hot-standbys using different Cloud providers, allowing for redundancy to take place.


For those looking to create “bullet proof” installations of their hosted presence, I definitely recommend giving Kumar’s post a good read.

The following two tabs change content below.

Michael Sheehan

Michael Sheehan, formerly the Technology Evangelist for GoGrid, is a recognized technology, social media, and cloud computing pundit and blogger who writes regularly about technology news and trends.

4 Responses to “Availability in the Cloud”

  1. Raj says:

    This is a good suggestion. But this is kind of applying the principles of traditional load balancing architecture/design to could or grid computing infrastructures. Doing this is good, but then it completely kills the purpose of using redundant servers on a grid/cloud anyways. 100% uptime is a necessity for Grids and Coulds ….

    My 2 cents!

  2. Randy Bias says:

    Since 100% uptime is impossible, this is more than a valid strategy, it’s sound business sense. It’s no different than a disaster recovery site. Honestly, the key problem is that you need a way to build your application in a cloud provider agnostic way.

    That’s where CloudScale comes in.

  3. Michael Sheehan says:


    Since Cloud Computing is pretty much an evolution of traditional computing, I think it is important to look at how things were resolved traditionally and then apply it to the Cloud. There will be more and more services that will offer these types of solutions. Redundancy within a specific grid is fine but the idea is that you need redundancy across the cloud as well.

    Agree with that. I’m working on further defining Cloud Computing (and the Cloud Pyramid I talk about in previous posts) to include Aggregators and “Extenders” (for lack of a better term). CloudScale is probably one of many that will emerge, capitalizing on Cloud Infrastructure providers.

    Thanks for your comments!

  4. [...] into this category, all they need to do now is improve application availability and make sure the Cloud is always available and I’ll be ready to hook up my [...]

Leave a reply