Skip to content

Gateway runtime with two CPUs

The table below displays the resource allocations for the gateway-related components used in the performance tests.

Component CPU Memory Router Concurrency GOMAXPROCS
Gateway Controller 1 2 GB — —
Gateway Runtime 2 2 GB 2 2

Throughput (requests/sec) vs. concurrent users

The graph below shows how gateway throughput changes as concurrent users increase for the API without policies and the API with mediation policies.

Line chart of gateway throughput in requests per second versus concurrent users at 100, 500, 800, and 1000 users, comparing the API without policies and the API with mediation policies on a two-CPU gateway runtime

Key observations:

  • Maximum throughput for both APIs occurs at 100 concurrent users on this two-CPU configuration.
  • Throughput decreases as concurrent users increase beyond 100 due to resource contention on the gateway runtime.

Average response time (ms) vs. concurrent users

The graph below shows how average response time changes for both APIs as concurrent users increase. The backend delay was configured to 0 ms for these tests.

Line chart of average response time in milliseconds versus concurrent users at 100, 500, 800, and 1000 users, comparing the API without policies and the API with mediation policies on a two-CPU gateway runtime

Key observations:

  • Average response time increases as concurrent users grow due to resource contention on the gateway runtime.

Response time percentiles vs. concurrent users

The graphs below show the 90th, 95th, and 99th percentile response times at 0 ms backend delay. Percentile values indicate the response time below which that percentage of requests completed, for example, the 99th percentile is the response time exceeded by only 1% of requests.

Line chart of 90th, 95th, and 99th percentile response times in milliseconds versus concurrent users for the API without policies on a two-CPU gateway runtime

Key observations:

  • 90th, 95th, and 99th percentile response times increase as concurrent users grow.
  • Percentile values represent the response time below which that percentage of requests completed.
  • Higher concurrency widens the spread between lower and upper percentiles.

Line chart of 90th, 95th, and 99th percentile response times in milliseconds versus concurrent users for the API with mediation policies on a two-CPU gateway runtime

Key observations:

  • Percentile measurements include Set Header policy execution on each request and response.
  • Percentile trends follow the same pattern as concurrent users increase across the test range.

Throughput (requests/sec) vs. concurrent users for varying payloads

The graph below shows how throughput changes with concurrent users for payload sizes of 50 B, 1 KiB, 10 KiB, and 100 KiB.

Line chart of gateway throughput in requests per second versus concurrent users for payload sizes of 50 B, 1 KiB, 10 KiB, and 100 KiB on a two-CPU gateway runtime

Key observations:

  • Throughput varies with payload size; smaller payloads support higher request rates.
  • For each payload size, throughput decreases as concurrent users increase.

Test scenario results in CSV format are available here.