Think like a Startup and you too can Deploy in the Clouds
popping..
A lot of people are still trying to figure out if cloud computing makes sense for big corporations with years of legacy baggage. At the same time, almost everybody understands that cloud computing is a no-brainer for startups due to reasons such as low cost to entry and speed to market. What cloud computing is doing is making startups a huge competitive threat for some enterprises because they can move from concept to production in a fraction of time and cost that monolithic corporations can.
My advice, start thinking like a startup!
Here are two examples of how to act like a startup:
Scenario #1 – Your company has asked you to build a new product or service from scratch.
Solution – Build a team of hard working and open minded individuals and create a startup atmosphere. Give them an aggressive timeline, a limited budget, and make them work offsite (another location or from home). Remove all of the corporate barriers and let them get creative. Give them the requirement that the software must be able to run at any location in case you switch cloud providers (this should also allow you to run it on premise if you find the clouds are not for you).
Advantages – (1) Reduces the politics and corporate obstacles that slow down innovation. (2) Trying this with something new rather than legacy is less risk. Besides, it can be built for the clouds as opposed to be rigged for the clouds. (3) The team should feel empowered and motivated which are two key elements in successful change initiatives.
Scenario #2 – Your company has asked you to reengineer an existing product and service.
Solution – Build the team like I mentioned in scenario #1. With most reengineering efforts, there will be some amount of time that the new system will have to run parallel with the existing system for validation purposes. What better way to run parallel than to run the new system in the clouds!
Advantages – (1) Lowers risk by leaving existing system untouched and not putting additional strains on existing infrastructure. (2) Great opportunity to gather key performance and financial metrics to build a business case for more cloud computing initiatives!!!
What are some scenarios that you can think of?
Did you enjoy this post? Why not leave a comment below and continue the conversation, or subscribe to my feed and get articles like this delivered automatically to your feed reader.
Comments
@Sandy,
They sort of do. On Amazon the cost for EC2 is extremely low. You can use a tool like Elastic Server On Demand (elasticserver.com) to install a standard configuration and quickly do some proof of concepts at a very low cost. I installed a WSO2 stack and a DJango stack on my EC2 account with just a few clicks.
I had a conversation today with someone who has quite a bit of practical experience with EC2, who claims that it is pretty slow — probably okay for a functional pilot, but not for production in general. Apparently, outbound web services calls from EC2-based apps are having to go through to many layers, which would tend to kill any BPM app with a lot of web services orchestration.
It would be interesting to see how they set it up. Which instance type did they use? They differ in memory, cores, storage, etc. How well were the web services architected? How was the database partitioned? I could see how a standard on-premise setup might not be optimal for EC2, but before I buy into that opinion that the cloud is too slow for BPM tools I would like to know more about the configuration.
Neither my acquaintance nor I said that the cloud was too slow for BPM, he just said EC2 was slow. There are other cloud platforms.
@Sandy,
Oh, I still wonder about how his app was designed and how his EC2 instance was configured. Now if he moved it to another vendor and it ran faster with the same config, then I would raise an eyebrow!
Would be great for vendors to offer pilot systems to enterprise customers: a standard image on EC2, for example, that the customer can work with themselves for a period of time. Much lower barrier to installation than an on-premise pilot.