benchmark.txt

Benchmark results:

  Date: 2016/03/20
  Version: asgi_redis 0.8.3 / daphne master / channels master / cpython 2.7.11
  Environment: AWS, 3x m3.large (1 daphne, 1 worker, 1 redis)
  Source: 1x m3.medium, different AZ, ab for HTTP, channels benchmarker for WS

  HTTP GET, Django View (concurrency 10, 10,000 total)
    Failures: 0.0%
    Rate: 160 req/s
    Mean: 60ms
    Median/95/99: 50ms / 140ms / 210ms

  HTTP GET, Django View (concurrency 20, 10,000 total)
    Failures: 0.0%
    Rate: 152 req/s
    Mean: 131ms
    Median/95/99: 96ms / 347ms / 538ms

  HTTP GET, Django View (concurrency 50, 10,000 total)
    Failures: 0.0%
    Rate: 160 req/s
    Mean: 310ms
    Median/95/99: 219ms / 887ms / 1360ms

  Websocket Echo, Django Consumer (concurrency 10, 10 msg/s, 200 sockets total)
    Failures: 0.0%
    Out of order: 0.0%
    Mean: 12ms
    Median/95/99: 11ms / 19ms / 39ms

  Websocket Echo, Django Consumer (concurrency 10, 50 msg/s, 1,000 sockets total)
    Failures: 0.0%
    Out of order: 0.0%
    Mean: 17ms
    Median/95/99: 14ms / 36ms / 60ms

  Websocket Echo, Django Consumer (concurrency 50, 250 msg/s, 1,000 sockets total)
    Failures: 0.0%
    Out of order: 0.0%
    Mean: 119ms
    Median/95/99: 77ms / 319ms / 458ms

  Websocket Echo, Django Consumer (concurrency 500, 50 msg/s, 1,000 sockets total)
    Failures: 0.0%
    Out of order: 0.0%
    Mean: 21ms
    Median/95/99: 20ms / 39ms / 54ms