Great to be part of Electric Cloud's final Continuous Discussions (#c9d9) discussion of 2017 with Jim Lyman of 451 Research, Mike Kavis of Cloud Technology Partners, Torsten Volk of Enterprise Management Associates, and Anders Wallgren and Sam Fell of Electric Cloud. We talked about our predictions for DevOps in 2018 as well as our resolutions.
Following are the key takeaways:
There will be a greater prevalence of open source software in DevOps. Research shows the more open source software used to implement a DevOps methodology positively correlates with full DevOps deployment. It's difficult to out-innovate a passionate crowd - something open source and DevOps has in common.
More top-down adoption of DevOps as the c-suite see the benefits of the methodology.
Push to AI/ML while leveraging expectations since real-world applications are very specific. As such, it's hard to know how quickly organizations will benefit.
A greater movement to cloud-native applications and building on top exposing data to data scientists so they can begin to provide valuable insights for business.
Serverless containers - the intermediate step between containers to consuming services directly.
DBAs become data architects and any IT administrative role should be wary of their job going away in the next few years, as such there's an opportunity for current administrators to look for ways to expand their expertise to provide greater value to their organizations.
More automated testing of code, including its security, and the evolution of DevSecOps. Also more automation throughout the entire DevOps methodology rather than fragmented automation.
Vision into the entire DevOps process so there's transparency and no "finger pointing" among the different members of the team.
Cross-platform, omni-platform so different solutions can plug and play with each other which is what end-users want - driving communities and process. Standardization, innovation, and a passion for open source and DevOps.
Reduction in the delays between stages that become obvious when doing value-stream mapping. The process needs to be streamlined so automation can accelerate DevOps. The companies that are able to respond quickly will be the ones that win. Responsiveness has always been part of DevOps. It will be interesting to see how AI/ML drives the process.
More women in DevOps leadership positions. The more women in DevOps, the more successful organizations will be - we're already seeing this happen across the industry.
DevOps delivers value that helps to deliver outstanding customer experience (CX). Listen and respond to the customer quickly.
If DevOps is failing within an organization, the organization needs to fix the cultural issues that haven't been addressed.
Get rid of the gaps in the value stream. Find the gaps and put in artifacts that can be automated.
DevOps focuses on people and process. Look at the process and optimize it before automating it so you're not automating unnecessary processes.
Big data, AI/ML must deliver value. Improve the quality of data to get more value from it.
Stop using the word "legacy" for software that's already written as opposed to code which is not tested or testable. If the code still works, it can be moved forward. You can do DevOps with legacy software that's been tested and is testable.