chore: dump Library metadata from /binaries/metadata.txt if available

utils/_context/containers.py

@@ -1114,6 +1114,11 @@ def post_start(self):
 
         logger.stdout(f"Library: {self.library}")
 
+        if self._container is not None:
+            exit_code, output = self.exec_run("cat /binaries/metadata.txt")
+            if exit_code == 0 and output:
+                logger.stdout(f"Library metadata:\n{output.decode('utf-8', errors='replace').strip()}")
+
         if self.appsec_rules_file:
             logger.stdout("Using a custom appsec rules file")
 

utils/build/docker/python/install_ddtrace.sh

@@ -18,6 +18,12 @@ elif [ "$(ls *.whl | wc -l)" = "1" ]; then
 elif [ $(ls python-load-from-s3 | wc -l) = 1 ]; then
     GIT_REF=$(cat python-load-from-s3)
     echo "Install ddtrace from S3, git ref: ${GIT_REF}"
+    # Try to fetch `metadata.txt` and save it so test output can display it, but don't fail if we cannot
+    # This includes commit sha, pipeline id, etc that was used to build the artifact
+    curl -sf https://dd-trace-py-builds.s3.amazonaws.com/${GIT_REF}/metadata.txt > metadata.txt || true
+    if [ -s metadata.txt ]; then
+        cat metadata.txt
+    fi
     # Install from S3 bucket
     # NOTE: Artifacts age out after 2 weeks, if this fails then you need to first run the dd-trace-py GitLab CI for the desired commit again
     # NOTE: Must have `--no-index` otherwise `pip` will look for the highest available version between S3 and PyPI