When we talk about cloud computing, those of us “in the biz” often go on and on about clouds themselves. But ultimately, IaaS and PaaS clouds don’t do anything themselves. They are merely a substrate on which to build other things, the infrastructure plumbing for the application development and deployment that we really want to do. At ServiceMesh, spend a lot of time with our clients talking about using clouds to change the solution delivery lifecycle (SDLC), enabling enterprises to use cloud-IT to deliver business value faster. (Yes, I know that “SDLC” is often referred to as the software delivery lifecycle, but that frequently implies just homegrown applications development, which “solution delivery lifecycle” seeks to include as a subset, also comprising roll-out of pre-built applications.)
There are a couple of considerations that enterprises need to think about when it comes to cloud-base solutions delivery:
- How do I develop cloud-native applications? What are my tools going to look like and how can I leverage the cloud to do continuous delivery of those applications? Ultimately, how fast can I make the development cycle? How can I make the time from the conception of the application to its delivery into customer hands (where it produces the planned business value) as short as possible?
- Then, what about all those existing non-cloud-native apps deployed in my sprawling data centers? As a large enterprise, you might have 10,000 or more deployed applications (sometimes much more). What’s the best way to work through that portfolio, prioritize which applications you concentrate on first, and determine which applications are destined to stay in the existing environment?
Recently, the folks at Focus.com held a roundtable of experts to discuss application delivery in the cloud, and I was lucky enough to be invited. The panel discussed some of these considerations and many others. You can hear the recorded roundtable here.
If your enterprise is struggling with these issues, contact ServiceMesh and we can help you understand the issues, create a strategy, and then a series of concrete implementation steps. Remember, it’s all about the apps!
Latest Tweets
Interesting post. I would like to think Cloud as all about APIs at the fundamental level. Apps could be layered on top of them. And looking at the application development trend rapidly moving towards portable device platforms like iPads and tablets, it makes even more sense.
Is there more to “Cloud native apps” vs “Non-Cloud native apps” ? Didn’t understand what do you mean with “native” there ?
Ramesh, on cloud-native vs. non-cloud-native… Most apps deployed in data centers today, even if virtualized, were not written with the cloud in mind. The expectation is that they are deployed on single servers for long periods of times. Cloud-native apps are written with the cloud in mind from the get-go. Typically, functionality is broken into smaller units and the application can handle new nodes being brought up dynamically to handle auto-scaling, or the migration of data from place to place, or cloud to cloud. Those cloud-native applications will generally be more resilient to cloud failure (what happened to Amazon EC2 a couple months ago, for instance) as well as perform better.