Guessing alert threshold settings, coping with event storms, and scanning logs for root cause analysis at 2am is inefficient and soul-crushing. This session uses Marco’s experiences to give a reality-check on AI/ML, list what DevOps missed out on, and why he sees AIOps as the natural next step.
DevOps took us from SysAdmins to DeployAdmins to improve availability but came with a tidal wave of tools and environments, leaving detecting anomalies and finding root cause a task for the overcrowded war-room of siloed experts. Developers need to understand infrastructure, and operations needs to understand the SDLC. The good news is we can help each other! To understand why you need AIOps to survive DevOps, I need to level set on the myth and reality of machine learning today. Then we can examine what tends to get overlooked by DevOps approaches. Finally, let’s agree on what AIOps is and where it has helped best so far with use cases.