This document discusses models of availability and complexity in large systems. It covers several services (SERVICEA-SERVICEG) and different priorities like performance, fault tolerance, and availability. It discusses how complex systems are difficult to comprehend and provides an example of a supply chain. It introduces the concepts of accidental and essential complexity. The document advocates for chaos engineering practices like experimenting in production by varying real-world events and automating experiments. Finally, it discusses intuition engineering and provides contact information.
18. Complex systems, by their nature, are unreasonable
systems. They exceed the capacity of a human mind
to comprehend the interaction of parts in any
predictive capacity.
UNREASONABLE SYSTEMS
54. The facilitation of experiments
to uncover systemic weaknesses
E N G I N E E R I N G
principlesofchaos.org
55.
56.
57.
58.
59. Tests:
Define the valence of a known, asserted property
Experiments:
Create new knowledge by nullifying a hunch
TESTS != EXPERIMENTS
60. ● Build a hypothesis about steady-state
behavior
principlesofchaos.org
61. ● Build a hypothesis about steady-state
behavior.
● Vary real-world events
principlesofchaos.org
62. ● Build a hypothesis about steady-state
behavior.
● Vary real-world events.
● Experiment in production
principlesofchaos.org
63. ● Build a hypothesis about steady-state
behavior.
● Vary real-world events.
● Experiment in production.
● Automate experiments to run
continuously
principlesofchaos.org
64. ● Build a hypothesis about steady-state
behavior.
● Vary real-world events.
● Experiment in production.
● Automate experiments to run
continuously.
● Minimize the blast radius
principlesofchaos.org
65. ● Build a hypothesis about steady-state
behavior
● Vary real-world events
● Experiment in production
● Automate experiments to run
continuously
● Minimize the blast radius
principlesofchaos.org