Hang on! Didn’t we just finish other races recently or still racing SMAC, Cloud, Big Data etc. It is always Old Wine in New Bottle or may be time has come now for that old idea, which never took off the ground.
‘Neighbours Envy, Owners Pride’ said one famous commercial for selling their TV sets. I see many companies are in the rush to implement DevOps just to be either first or beat the competition or in their sincere interest to roll-out the product faster. The result being, you implemented DevOps but don’t get the desired output yet. We know that the chain is only as strong as the weakest link. Hence, choosing the individual tool is as critical as making all the tools function as one. The sum total performance of all the tools put together being more than the performance of individual tools. There could be many such ‘weak links’ that may pull down your objective.
Can we compare DevOps to Lego? We try to put the pieces to get the ‘right’ output. Even if you don’t put the right pieces together, still you get to see something in Lego i.e a building with 2 levels instead of 3. You still got a building! Today, DevOps is not far behind in comparison.
With various permutation & combination of tools & technologies that you put together, you certainly get a model, but is this what you desired as a goal?
Have a blueprint ready for DevOps even before you start inducting tools. Understand the following minimum before you venture:
- Your business model
- Software development methodology
- Product roll-out plans
- Target customer profile
- Your budget
These are highly critical inputs for developing a DevOps framework which I am referring to as Blueprint. Once you have this ready, it is only a question of evaluating various tools and mapping it across the life cycle. The true success of DevOps is when software is developed, tested and deployed in a fully automated factory shop-floor-like setup.