Chapter 7. Autonomous Chaos Engineering
Most teams that run chaos engineering well today do it quarterly. Game days, scheduled exercises, and deliberate disruptions are on dates chosen weeks in advance. The discipline produced enormous value for an entire generation of SRE practice, and the value has not gone away. What has gone away is the assumption underneath: that the system being tested in March is similar enough to the system being tested in June that quarterly snapshots are sufficient evidence of resilience. In an agentic environment, the system you tested in March is not the system you are running in June, and the reason is not that anyone shipped a major change. It is that the system itself has evolved continuously. Services have come and gone. Dependencies have shifted implicitly. Traffic patterns have moved with the seasons. Agents have learned new behaviors. Decision paths exist now that did not exist at the last game day, formed by accumulated reasoning rather than by any deliberate engineering decision.
The core question chaos engineering ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access