
Four Steps to Start Chaos Engineering

#42 - Chaos Engineering - Mikołaj Pawlikowski
Topic 6 of 11
Published Jun 14, 2021 · Clip 25:42–28:15 (2:34)
Miko recaps the four steps of a chaos experiment. First, observability: being able to reliably measure something, such as server up/down or requests per second. Second, the steady state: the normal range of that metric. Third, forming a hypothesis, e.g., if disk becomes 50% slower the system still handles a target rate. Fourth, implementing the experiment with available tools and observing the results.