Implement database initialization with retries (#86)

* Implement database initialization with retries

Added additional logging & retry mechanism for database initialization in run.sh.

* Update run.sh

Co-authored-by: Copilot <[email protected]>

---------

Co-authored-by: Copilot <[email protected]>

run.sh

@@ -6,10 +6,32 @@ WORKERS_VALUE="${WORKERS:-1}"
 TIMEOUT_VALUE="${TIMEOUT:-120}"
 PORT_VALUE="${PORT:-10000}"
 
-# Initialize the database
+# Initialize the database with retries (handle DB cold start)
 echo "Initializing database..."
-python scripts/init_pgvector.py
+MAX_RETRIES="${DB_INIT_RETRIES:-10}"
+SLEEP_BASE="${DB_INIT_SLEEP_BASE:-5}"
+
+i=1
+until python scripts/init_pgvector.py; do
+  if [ "$i" -ge "$MAX_RETRIES" ]; then
+    echo "Database initialization failed after $i attempts; exiting."
+    exit 1
+  fi
+  echo "init_pgvector.py failed (attempt $i/$MAX_RETRIES). Sleeping $((SLEEP_BASE ** i))s before retry..."
+  sleep $((SLEEP_BASE ** i))
+  i=$((i + 1))
+done
 
 echo "Starting gunicorn on port ${PORT_VALUE} with ${WORKERS_VALUE} workers and timeout ${TIMEOUT_VALUE}s"
 export PYTHONPATH="/app${PYTHONPATH:+:$PYTHONPATH}"
-exec gunicorn --bind 0.0.0.0:${PORT_VALUE} --workers "${WORKERS_VALUE}" --timeout "${TIMEOUT_VALUE}" app:app
+
+# Use explicit logging options so any worker crash or import error is visible
+exec gunicorn \
+  --bind 0.0.0.0:${PORT_VALUE} \
+  --workers "${WORKERS_VALUE}" \
+  --timeout "${TIMEOUT_VALUE}" \
+  --log-level debug \
+  --access-logfile - \
+  --error-logfile - \
+  --capture-output \
+  app:app