AWS DevOps Guru: Automated Operational Issue Detection

During today’s AWS re:Invent event, Andy Jassy unveiled DevOps Guru, a novel solution designed to assist DevOps teams in pinpointing operational challenges that may be affecting application performance. Think of it as complementing CodeGuru, the service introduced last year focused on identifying problems within your code prior to deployment.

The tool functions using a comparable methodology, leveraging machine learning to detect issues concerning the operational aspects of applications. Jassy stated today, “I’m pleased to introduce Amazon DevOps Guru, a new service that employs machine learning to detect operational problems before they affect users.”

The system functions by gathering and examining data from application metrics, logs, and events to “detect anomalies from typical operational behavior,” as detailed in the company’s announcement post.

This offering positions AWS to compete with providers such as Sumo Logic, DataDog, and Splunk by delivering comprehensive operational insights into potential application issues, including incorrect settings or resource limitations.

Upon detecting a problem, the service is capable of sending notifications via SMS, Slack, or other channels to the relevant team, along with suggested solutions for swift resolution.

Notably, billing is based on the volume of data analyzed by the service, eliminating the need for a fixed monthly subscription. The company emphasizes that this approach requires no initial investment or long-term commitment.