Error Budget Exhausted — Mid-Sprint
Your service hits zero error budget with 2 weeks left in the sprint. Walk through the policy activation, stakeholder conversation, and recovery steps.
Overview
When the budget hits zero
The moment the error budget reaches zero is the highest-stakes SRE conversation. The policy is clear on paper; the hard parts are the product conversation, attributing correctly what burned the budget, and not restarting the clock prematurely.
> [!IMPORTANT] The error-budget policy is pre-agreed. Exhaustion triggers it; it is not a new negotiation.
The three hard parts
- The product conversation — Pressure to keep shipping features anyway
- Attribution — Blaming the wrong team or double-counting burn events
- The clock — Declaring recovery by calendar date instead of the rolling window
Interview questions
- Your checkout service SLO is 99.95% over 30 days. On day 16, a 45-minute partial outage exhausts the remaining error budget. What do you do?
- The product team pushes back on the feature freeze. They have a high-priority customer delivery due in 3 days. How do you handle it?
- How do you prevent the same budget exhaustion from happening next quarter?
Sign up free to read the full topic
More SRE topics
Start free on Cariara · Coding problems · System design studio