STAGE 1: Define a user-centered indicator
Concept: An SLI measures a user-relevant service behavior.
Task: Define successful requests, eligible traffic and the measurement window.
Acceptance: Exclude only traffic justified by the documented definition.

My assumptions:

My work / file link:

My checks and results:

What I changed:


STAGE 2: Calculate the budget
Concept: An error budget follows from a target and a denominator.
Task: Calculate allowed unsuccessful requests under the sample objective.
Acceptance: State that the chosen target is a case assumption, not a universal standard.

My assumptions:

My work / file link:

My checks and results:

What I changed:


STAGE 3: Practice incident response
Concept: Incident work needs coordination and a clear timeline.
Task: Draft roles, updates and a mitigation decision from the synthetic incident.
Acceptance: Separate confirmed facts from hypotheses in every update.

My assumptions:

My work / file link:

My checks and results:

What I changed:


STAGE 4: Learn without blame
Concept: A review should improve the system rather than assign motives.
Task: Write contributing factors, follow-up actions and verification dates.
Acceptance: Do not claim a root cause merely because an event preceded the outage.

My assumptions:

My work / file link:

My checks and results:

What I changed:
