The cloud experiment is over and the debate in enterprises about its benefits and risks is settled. We know it works, it’s more flexible and cheaper, and it makes it easier for IT to align with business needs. So should buyers put their applications into a cloud environment?
My advice: Don’t rearchitect your legacy applications that were designed and implemented in a legacy environment and port them over to the cloud. Organization of all sizes have been waiting for providers’ porting solutions. Unfortunately, that’s sort of like the Samuel Beckett tragicomedy play, “Waiting for Godot,” in which two characters wait days for Godot even though they don’t know where or when he might arrive. Buyers wait, thinking cloud porting solutions will arrive in the market, but it just doesn’t happen. That’s because porting is really expensive and really risky.
I’ve blogged in the past about CSS Corp Cloud Services and Redwood Software platforms for easily migrating legacy apps to the cloud. But as we get further into the cloud story, it looks like replatforming offerings will be far rarer than we anticipated. I’m not saying they won’t exist; I’m just saying they won’t be the dominant model.
As the smoke clears from cloud experimenting and pilots, the best-practice dominant model for moving into the cloud is shaping up as follows:
- Look for opportunities to make incremental improvements to your legacy environment. Rework legacy by increasing the level of virtualization and automation in your data center.
- When you develop new applications, architect them for the cloud environment.
This strategy of adding virtualization and automation may get your legacy environment into a private cloud, but it doesn’t get you into the agile low-cost public cloud environment. However, it allows you to improve the efficiency and resiliency of the existing legacy environment without the huge cost and risk of rearchitecting.
The strategy also helps CIO organizations regain some of the influence and credibility they’ve lost with business units as they’ve addressed new functionalities enabling where the business is moving. It enables the organization to be more agile, better aligned and do so with lower cost, which significantly relieves the tension of having to get a huge amount of funding for a set of high-risk legacy projects.
The fact is for many legacy applications the best you can do is make incremental progress. You can move them out of dedicated hardware into virtualized hardware. And other than some potential cost savings, there is little to no business benefit from taking on the risk of reengineering them for a public infrastructure or shared environment.
We saw this same best-practice model happen with distributed computing; new applications went into distributed computing and eventually we reached a tipping point where we needed to move legacy apps. I anticipate the new functionalities, new work will similarly drive the shift from legacy to cloud.
Going forward until the tipping point occurs, put all your efforts into standing up your organization’s new environment to take full advantage of the business alignment, flexibility and cost that the cloud family offers and just make incremental changes to your legacy environment. If you wait for a huge re-platforming surge of cloud porting solutions, I believe you’ll be waiting for Godot.