Normally the adoption of something as game changing as AIOps takes a lot of time and conversation with a customer because their environments grow at a rapid pace, their toolsets are many, their infrastructure teams still operate in silos and they must support a complex mix of legacy and cloud-native applications. Recently, however,
we’ve witnessed something unprecedented and remarkable; our customers are rushing to have this conversation. Covid-19 is speeding digital transformation because it is now a critical imperative. Layoffs have dramatically reduced human resources in every department, including IT. This means that humans in IT Ops can’t continue to pour through logs and emails to figure out and resolve security or operational issues. They have no choice but to spend their time on better things.
Capabilities brought by the AIOps space were already contributing to the morphing of the DevOps movement into the NoOps movement. DevOps combined the functions of Developers and IT Operations to work closely together to enable continuous integration and continuous delivery of software function and features. This game changing school of thought caused organizations to go from a few releases a year to a few releases a day. But the pressure is even higher for continuous improvement and the next logical area to attack is IT Ops. Enter the “NoOps” concept, wherein the operational environment is automated to the point that a team is not needed to manage it. The term NoOps was first expressed by Mike Gualtieri of Forrester Research, who proclaimed that he didn’t want application developers to have to waste time with Operations Engineers anymore. He suggested that it was now possible to automate the operational environment to the point that a team is not needed to manage it.
In a NoOps scenario, IT processes are completely overhauled such that repetitive tasks are removed and basic, intelligent tasks are handled: event correlation is used to automate the handling of situations that result in downtime. The hope is that automating the deployment, monitoring and management of applications ultimately leads to a world where environments no longer require humans to hover over critical business services in a command center. NoOps can also help combat the top reasons for downtime: infrastructure configurations and application changes. Delivering infrastructure-as-code enables the environment to be more easily stabilized and helps reduce some of the root causes.
NoOps requires a single-pane-of-glass for monitoring the full stack, applications and their dependencies, but to date this has not been possible. The reality is that we have a single-glass-of-pain consisting of siloed monitoring across domains and application teams: a myriad of tools exist across a dizzying number of ITOM categories. The answer is to think of the monitoring and operational stack as a platform, rather than a set of tools. A platform, versus a specific set of tools, can ingest data from diverse data sources, provide a real time view of all the ever-changing components and provide an eventing engine that deploys machine learning to develop intelligence that can enable self-healing. Download our reference architecture which will help you plot and modularize your tools and start pulling them together to orchestrate the ultimate AIOps platform.