SLO & Error Budget Lab

Synthetic data

Explore how Payments Hub service level objectives and error budgets drive alerting and release decisions.

Service

SLI type

SLO target: 99.90%

Window

Synthetic incident severity: 0.5% error rate

Error budget

43.2 min

(1 − SLO) × 30 days

Burn rate

5×

observed error rate ÷ (1 − SLO)

Budget consumed

100%

if sustained across the window

Budget remaining

exhausted in 6 days at this rate

Multi-window burn-rate alerts

Action	Budget consumed	Long window	Short window	Burn rate threshold	Error rate threshold	Status
Page	2%	1h	5m	14.4	1.44%	Quiet
Page	5%	6h	30m	6	0.6%	Quiet
Ticket	10%	3 days	6h	1	0.1%	Would fire

Latency: Time to serve a request, tracking failed requests separately from successful ones.
Traffic: Demand on the service, such as requests or payments per second.
Errors: Rate of requests that fail, explicitly or by breaching a policy.
Saturation: How full the service is: the most constrained resource, before it degrades.