Throughput measures how many requests a service handles per second. It complements response speed: a service may respond in 50ms yet bottleneck if it handles only 10 requests per second. section 3.2.2
Use throughput to turn service capacity into an operational target:
- Measure requests per second alongside response-time percentiles and error rates rather than optimizing speed alone. section 3.2.2
- Set targets from peak traffic patterns and preserve headroom, typically two to three times normal peak traffic, for unexpected surges. section 3.2.2
- Connect the metric to user-facing outcomes: insufficient capacity can make pages, applications, dashboards, or checkout flows feel unreliable or unresponsive. section 3.2.2 section 3.6
Capacity under peak load
A service handling 500 requests per second at peak has a concrete throughput baseline. Compare that baseline with expected peak traffic and planned headroom, then monitor it with latency and error-rate data. section 3.2.2
Table 3.4 Backend performance metrics
|
Metric
|
Measures
|
Example
|
|
Response time percentiles (e.g. p50, p95, p99)
|
How fast do requests complete across the distribution of requests
|
“p95 of 200ms” means 95% of requests complete within 200ms
|
|
Throughput
|
Requests handled per second
|
Service handles 500 requests per second at peak
|
|
Error rate
|
Percentage of requests that fail
|
0.05% of payment requests return errors
|
Use the backend performance metrics table to compare throughput with response-time percentiles and error rate in a single capacity view. section 3.2.2































