With the advent of the public cloud, we saw new types of infrastructure management patterns, which included autoscaling as well as self-healing-based techniques, making the application scaling and fault recovery process as automated as possible. During the next wave of automation, we saw DevOps become more mainstream, which was also a result of blurring lines between true developer and system operations roles due to abstracted cloud services. As a result, we saw faster and frequent deployments, easy rollback mechanisms in the case of issues, and better tools and services, which made the entire process easier and integrated. However, even with these type of advanced automation techniques, there's a need for operations and updates at certain operating system and application level aspects, which is still mostly a manual effort. Newer services like Amazon Lambda and Amazon serverless Aurora are taking that kind of heavy lifting away from the customers' hands, but still, there's a long way to go until the systems are truly smarter and can handle most of the ops-related aspects on their own. This may also mean that systems will have to be smarter, where instead of reactively responding to a situation they will proactively predict possible failures scenarios and the need for any changes, thereby reducing the need to manually correct anything or even build sophisticated automations, which may not even be needed!
To enable the aforementioned transition, technologies like Artificial Intelligence and Machine Learning will play a big role. In fact, in part, that has already started to happen, where many cloud providers' services are powered by predictive modelling techniques behind the scenes. As an example, Amazon Macie uses Machine Learning to automatically discover, classify, and protect sensitive data in AWS. Likewise, Amazon GuardDuty identifies suspected attackers through integrated threat intelligence feeds and uses Machine Learning to detect anomalies in account and workload activity. These are great examples of Machine Learning being used in the space of security, however, as we progress, the same principles and mechanisms will become more mainstream for application deployment and management models, where typical operations will change greatly.
Future trend #3
Cloud services and systems will be smarter, where typical requirements around infrastructure and application operations will diminish, thereby resulting in newer principles of NoOps.