| service: api |
| |
| runtime: python38 |
| entrypoint: gunicorn -b:$PORT dashboard.dispatcher:APP --worker-class gthread --threads 10 --timeout 60 |
| app_engine_apis: true |
| # 2048MB Memory, 2.4GHz CPU limit, support auto scaling. |
| instance_class: F4_1G |
| |
| automatic_scaling: |
| # We're setting the max concurrent request to 20, to allow AppEngine to scale |
| # the number of instances to handle API requests better. We're also going to |
| # keep around 10 instances at the ready to handle incoming requests better |
| # from a "cold start". |
| max_concurrent_requests: 20 |
| max_instances: 150 |
| max_pending_latency: automatic |
| min_instances: 3 |
| target_cpu_utilization: 0.8 |
| |
| env_variables: |
| GAE_USE_SOCKETS_HTTPLIB: 'true' |
| |
| inbound_services: |
| - warmup |
| |
| handlers: |
| - url: /.+ |
| script: dashboard.dispatcher.APP |
| secure: always |