Limping Roadrunner - What's next in DevOps ?

"DevOps" - one of the most talked word in today's IT world. DevOps is one of the software development process combines both development phase and Operations phase (support phase). You can call it as an evolution from Agile and CICD models.

Automation plays a major role in DevOps world and to have continuous delivery in place, all the inter connected components of software development life cycle should be connected and automated.

It should have the ability to check in code, ready the environment, populate data, functional test the checked in code, test the regression suite and deploy to production environment. In fact, many service industry offers shift left DevOps as well where they automate requirement analysis, tearing the requirements into test case using model based approach.

What is missing or rather in other terms not matured and foreseen as next wave in DevOps?

• Learning from the Ops (common incidents, defects) and self-healing codes to fix the issues or suggesting code / business changes with cognitive calculation
• Plug in micro services to report on health of system based on studies on system logs, performance logs, network connections, interfaces etc
• Infra structure validation to make sure the connected hardware, servers, networks are working fine and connected in DevOps life cycle
• Add performance testing to the mix, in a fully automated fashion that includes scalability through the enterprise, and scalability of user traffic is extremely important
• Feedback mechanisms - Ops to Development needs to be accelerated. Instead of reporting the issues, intelligent methods to be built to provide feedback on fixing the issue rather than reporting the issue
• Traceability from Requirements till the incidents or defects in operations layer that includes environment, data, infra, performance scenarios etc.
• Audit the automation process. You cannot have single or handful of automation tools to run the DevOps story. A car runs on multiple parts but all these parts has great handshake mechanism, which will talk to each other for smoother operation. Similarly you need to connect all the missing dots in the entire automation lifecycle
• Companies will favor container-based solutions that give them flexibility and portability in a hybrid environment that may contain services from multiple cloud vendors
• Data Security adhering to strict federal compliance and Industry standards
• DevOps are no longer be technology centric approach, it is leaning towards business requirements approach
• Cloud implementations of DevOps to meet the needs for an on-demand model.

We have different technologies and automation solutions to support the DevOps life cycle implementation. In fact, the deployments are getting faster and faster with as many as six releases happening in single day.  It might take less time to do a complete implementation including code check, validation, testing, deployment, but if there is any break / issues in the deployment cycle it might take more to understand and fix the issue.

Next phase of DevOps will be more on how the system identifies the broken chain, understands, learns and heal them automatically. Need to wait and watch.