feat: TCP health check integration + ECS health checks on staging (#5377)

-----

Co-authored-by: Rodrigo López Dato <rodrigo.lopezdato@flagsmith.com>

deps: bump flagsmith-common from 1.10.0 to 1.11.0

Author: khvn26

Dockerfile

@@ -119,7 +119,7 @@ FROM wolfi-base AS api-runtime
 
 # Install Python and make it available to venv entrypoints
 ARG PYTHON_VERSION
-RUN apk add curl python-${PYTHON_VERSION} && \
+RUN apk add python-${PYTHON_VERSION} && \
   mkdir /build/ && ln -s /usr/local/ /build/.venv
 
 WORKDIR /app
@@ -140,7 +140,7 @@ ENTRYPOINT ["/app/scripts/run-docker.sh"]
 CMD ["migrate-and-serve"]
 
 HEALTHCHECK --interval=2s --timeout=2s --retries=3 --start-period=20s \
-  CMD curl -f http://localhost:8000/health/liveness || exit 1
+  CMD flagsmith healthcheck tcp
 
 # * api-runtime-private [api-runtime]
 FROM api-runtime AS api-runtime-private

api/manage.py

@@ -1,17 +1,9 @@
 #!/usr/bin/env python
 import os
-import sys
 
 from common.core.main import main
 
 if __name__ == "__main__":
-    # Backwards compatibility for task-processor health checks
-    # See https://github.com/Flagsmith/flagsmith-task-processor/issues/24
-    if "checktaskprocessorthreadhealth" in sys.argv:
-        import scripts.healthcheck
-
-        scripts.healthcheck.main()
-
     os.environ.setdefault("DJANGO_SETTINGS_MODULE", "app.settings.local")
 
     main()

api/poetry.lock

@@ -154,7 +154,7 @@ pygithub = "2.1.1"
 hubspot-api-client = "^8.2.1"
 djangorestframework-dataclasses = "^1.3.1"
 pyotp = "^2.9.0"
-flagsmith-common = "^1.10.0"
+flagsmith-common = "^1.11.0"
 django-stubs = "^5.1.3"
 tzdata = "^2024.1"
 djangorestframework-simplejwt = "^5.3.1"

api/pyproject.toml

@@ -1,26 +1,11 @@
 import logging
 import sys
 
-import requests
-
-HEALTH_LIVENESS_URL = "http://localhost:8000/health/liveness"
-
-
-logger = logging.getLogger(__name__)
-
-
-def main() -> None:
-    logger.warning(
-        f"This healthcheck, invoked by {' '.join(sys.argv)}, is deprecated. "
-        f"Use the `{HEALTH_LIVENESS_URL}` endpoint instead."
-    )
-    status_code = requests.get(HEALTH_LIVENESS_URL).status_code
-
-    if status_code != 200:
-        logger.error(f"Health check failed with status {status_code}")
-
-    sys.exit(0 if 200 >= status_code < 300 else 1)
-
+from common.core.main import main
 
 if __name__ == "__main__":
-    main()
+    logging.getLogger(__name__).warning(
+        f"This healthcheck, invoked by `{' '.join(sys.argv)}``, is deprecated. "
+        "Please use one of the `flagsmith healthcheck` commands instead."
+    )
+    main(["flagsmith", "healthcheck", "http"])

api/scripts/__init__.py

@@ -6,7 +6,9 @@
     "containerDefinitions": [
         {
             "name": "flagsmith-task-processor",
-            "command": ["run-task-processor"],
+            "command": [
+                "run-task-processor"
+            ],
             "essential": true,
             "environment": [
                 {
@@ -21,6 +23,10 @@
                     "name": "AWS_DEFAULT_REGION",
                     "value": "eu-west-2"
                 },
+                {
+                    "name": "DJANGO_ALLOWED_HOSTS",
+                    "value": "*"
+                },
                 {
                     "name": "DJANGO_SETTINGS_MODULE",
                     "value": "app.settings.production"
@@ -255,4 +261,4 @@
     ],
     "cpu": "1024",
     "memory": "2048"
-}
+}

api/scripts/healthcheck.py

@@ -6,7 +6,26 @@
     "containerDefinitions": [
         {
             "name": "flagsmith-task-processor",
-            "command": ["run-task-processor"],
+            "command": [
+                "run-task-processor"
+            ],
+            "portMappings": [
+                {
+                    "containerPort": 8000,
+                    "hostPort": 8000,
+                    "protocol": "tcp"
+                }
+            ],
+            "healthCheck": {
+                "command": [
+                    "CMD-SHELL",
+                    "flagsmith healthcheck tcp"
+                ],
+                "interval": 10,
+                "timeout": 2,
+                "retries": 5,
+                "startPeriod": 5
+            },
             "essential": true,
             "environment": [
                 {
@@ -21,6 +40,10 @@
                     "name": "AWS_DEFAULT_REGION",
                     "value": "eu-west-2"
                 },
+                {
+                    "name": "DJANGO_ALLOWED_HOSTS",
+                    "value": "*"
+                },
                 {
                     "name": "DJANGO_SETTINGS_MODULE",
                     "value": "app.settings.production"

infrastructure/aws/production/ecs-task-definition-task-processor.json

@@ -17,6 +17,16 @@
                     "protocol": "tcp"
                 }
             ],
+            "healthCheck": {
+                "command": [
+                    "CMD-SHELL",
+                    "flagsmith healthcheck tcp"
+                ],
+                "interval": 10,
+                "timeout": 2,
+                "retries": 5,
+                "startPeriod": 5
+            },
             "essential": true,
             "environment": [
                 {