Containers seem to be the default strategy for most methods migrating to the cloud or being constructed there, and for good reasons. They provide portability and scalability (using orchestration) that is harder to realize with different enabling know-how. Furthermore, there is a healthy ecosystem round containers, and an answer is simpler to outline.
Nevertheless, very similar to other hyped applied sciences lately, akin to AI, serverless, and so forth., we’re seeing many situations the place containers are misapplied. Corporations are selecting containers when other enabling applied sciences can be better, extra value-environment friendly solutions.
Certainly, I feel we’re leaving hundreds of thousands of dollars on the table by selecting a know-how that’s not the appropriate fit. We’re going after points for hype and another trending know-how on the CV.
The core draw back of containers immediately is the overapplication of container improvement and the migration of present purposes to containers in “software modernization” tasks. It’s not that containers don’t work—in fact they do. But many things “work” which might be massively inefficient in comparison with different applied sciences.
Most corporations are chasing the good thing about portability for a workload that is unlikely to ever move from its goal host platform. Also, and most significantly, they do not perceive that to really reap the benefits of what containers supply requires an entire re-structure of the appliance in most situations, which they sometimes didn’t do.
Internet-new improvement has this drawback as nicely. Enterprises are spending as much as 4 occasions the cash to build the same software utilizing container-based mostly improvement and deployment versus more conventional strategies. Also at problem, the container-based mostly software might value extra to function through the use of extra cloud-based mostly assets, similar to storage and compute. It additionally costs extra to safe and extra to control.
When evaluating containers, listed here are a couple of core points to think about:
- Give attention to returning worth back to the business. It’s the previous story of builders and engineers who don’t look out for the enterprise as much as they need to. Don’t comply with the hype.
- Don’t overstate advantages, akin to portability, which will by no means be used. If it costs twice or even four occasions the money to get there, what are the probabilities you’ll ever move an software?
- Perceive operational costs. Containers might value extra to function in the long run. I’m not saying don’t ever use containers, but perceive the true value of maintaining them through the years.
- Use architectural greatest practices. Purposes typically have to be redesigned for containers to be efficient. “Wrapping” one thing doesn’t offer you efficiency by default.
This can be a cautionary story to level out the necessity for a healthy skepticism about any know-how. I’m using containers for instance, however it might really be any know-how. Control the value returned to the business, and you’ll probably make the fitting calls.