From 67e2cfd2f9ed64bb0f45b45a3ff70a83f3484dc9 Mon Sep 17 00:00:00 2001
From: Ben Zhang <ben.zhang@watonomous.ca>
Date: Sun, 21 Jan 2024 02:51:25 +0000
Subject: [PATCH] Attach build and environment info to Sentry

---
 .github/workflows/push-image.yml |  2 ++
 Dockerfile                       |  4 +++
 app.py                           | 55 ++++++++++++++++++++++++--------
 3 files changed, 48 insertions(+), 13 deletions(-)

diff --git a/.github/workflows/push-image.yml b/.github/workflows/push-image.yml
index 55366c8..72c9a48 100644
--- a/.github/workflows/push-image.yml
+++ b/.github/workflows/push-image.yml
@@ -40,5 +40,7 @@ jobs:
         with:
           context: .
           push: true
+          build-args: |
+            DOCKER_METADATA_OUTPUT_JSON
           tags: ${{ steps.meta.outputs.tags }}
           labels: ${{ steps.meta.outputs.labels }}
\ No newline at end of file
diff --git a/Dockerfile b/Dockerfile
index 5d38fe0..80f6d6e 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -1,5 +1,9 @@
 FROM python:3.11.5-slim
 
+# Pass information about the build to the container
+ARG DOCKER_METADATA_OUTPUT_JSON='{}'
+ENV DOCKER_METADATA_OUTPUT_JSON=${DOCKER_METADATA_OUTPUT_JSON}
+
 COPY requirements.txt /app/
 WORKDIR /app
 
diff --git a/app.py b/app.py
index 73afacb..59a2b70 100644
--- a/app.py
+++ b/app.py
@@ -17,28 +17,57 @@
 LOG_LEVEL = os.environ.get("LOG_LEVEL", "DEBUG")
 PORT = os.environ.get("PORT", 5000)
 HOST = os.environ.get("HOST", "0.0.0.0")
-SENTRY_DSN = os.environ.get("SENTRY_DSN", "")
 
 logging.basicConfig(level=LOG_LEVEL)
-if SENTRY_DSN:
-    logging.info("Sentry DSN provided, enabling Sentry SDK")
+
+# BUILD_INFO is generated by the build pipeline (e.g. docker/metadata-action).
+# It looks like:
+# {"tags":["ghcr.io/watonomous/repo-ingestion:main"],"labels":{"org.opencontainers.image.title":"repo-ingestion","org.opencontainers.image.description":"Simple server to receive file changes and open GitHub pull requests","org.opencontainers.image.url":"https://github.com/WATonomous/repo-ingestion","org.opencontainers.image.source":"https://github.com/WATonomous/repo-ingestion","org.opencontainers.image.version":"main","org.opencontainers.image.created":"2024-01-20T16:10:39.421Z","org.opencontainers.image.revision":"1d55b62b15c78251e0560af9e97927591e260a98","org.opencontainers.image.licenses":""}}
+BUILD_INFO=json.loads(os.getenv("DOCKER_METADATA_OUTPUT_JSON", "{}"))
+
+
+# Set up Sentry
+if os.getenv("SENTRY_DSN"):
+    build_labels = BUILD_INFO.get("labels", {})
+    image_title = build_labels.get("org.opencontainers.image.title", "unknown_image")
+    image_version = build_labels.get("org.opencontainers.image.version", "unknown_version")
+    image_rev = build_labels.get("org.opencontainers.image.revision", "unknown_rev")
+
+    sentry_config = {
+        "dsn": os.getenv("SENTRY_DSN"),
+        "environment": os.getenv("SENTRY_ENVIRONMENT", "unknown"),
+        "release": os.getenv("SENTRY_RELEASE", f'{image_title}:{image_version}@{image_rev}'),
+    }
+
+    logging.info(f"Sentry DSN found. Setting up Sentry with config: {sentry_config}")
+
     sentry_logging = LoggingIntegration(
         level=logging.INFO,        # Capture info and above as breadcrumbs
         event_level=logging.ERROR  # Send errors as events
     )
+
+    def sentry_traces_sampler(sampling_context):
+        # Inherit parent sampling decision
+        if sampling_context["parent_sampled"] is not None:
+            return sampling_context["parent_sampled"]
+
+        # Don't need to sample health checks
+        if sampling_context.get("asgi_scope", {}).get("path", "").startswith("/health"):
+            return 0
+        
+        # Sample everything else
+        return 1
+
     sentry_sdk.init(
-        dsn=SENTRY_DSN,
-        integrations=[
-            sentry_logging,
-        ],
-
-        # Set traces_sample_rate to 1.0 to capture 100%
-        # of transactions for performance monitoring.
-        # We recommend adjusting this value in production,
-        traces_sample_rate=1.0,
+        **sentry_config,
+        integrations=[sentry_logging],
+
+        traces_sampler=sentry_traces_sampler,
+
+        enable_tracing=True,
     )
 else:
-    logging.info("No Sentry DSN provided, Sentry SDK disabled")
+    logging.info("No Sentry DSN found. Skipping Sentry setup.")
 
 def split_dsn(dsn):
     """