Fix 502: retry Neo4j connection at startup, delay backend 50s to wait for Neo4j

backend/neo4j_setup.py

@@ -1,5 +1,6 @@
 from neo4j import GraphDatabase
 import os
+import time
 from dotenv import load_dotenv
 
 load_dotenv()
@@ -7,22 +8,48 @@ load_dotenv()
 
 class Neo4jConnection:
     def __init__(self, uri: str, user: str, password: str, database: str = "neo4j"):
-        self.driver = GraphDatabase.driver(uri, auth=(user, password))
+        self._uri = uri
+        self._auth = (user, password)
         self.database = database
+        self.driver = None
+        self._connect_with_retry()
+
+    def _connect_with_retry(self, retries: int = 10, delay: int = 6):
+        for attempt in range(retries):
+            try:
+                self.driver = GraphDatabase.driver(self._uri, auth=self._auth)
+                self.driver.verify_connectivity()
+                print(f"[neo4j] Connected on attempt {attempt + 1}")
+                return
+            except Exception as e:
+                print(f"[neo4j] Attempt {attempt + 1}/{retries} failed: {e}")
+                if attempt < retries - 1:
+                    time.sleep(delay)
+        print("[neo4j] WARNING: Could not connect — queries will fail until Neo4j is ready")
 
     def close(self):
-        self.driver.close()
+        if self.driver:
+            self.driver.close()
 
     def run_query(self, query: str, parameters: dict | None = None) -> list:
-        with self.driver.session(database=self.database) as session:
-            result = session.run(query, parameters or {})
-            return [record.data() for record in result]
+        if not self.driver:
+            self._connect_with_retry()
+        try:
+            with self.driver.session(database=self.database) as session:
+                result = session.run(query, parameters or {})
+                return [record.data() for record in result]
+        except Exception:
+            # Driver may have gone stale — reconnect once and retry
+            self._connect_with_retry(retries=3, delay=3)
+            with self.driver.session(database=self.database) as session:
+                result = session.run(query, parameters or {})
+                return [record.data() for record in result]
 
 
 neo4j_conn = Neo4jConnection(
-    uri=os.getenv("NEO4J_URI", ""),
+    uri=os.getenv("NEO4J_URI", "bolt://127.0.0.1:7687"),
     user=os.getenv("NEO4J_USERNAME", "neo4j"),
-    password=os.getenv("NEO4J_PASSWORD", ""),
+    password=os.getenv("NEO4J_PASSWORD", "clinicalmatch2024"),
     database=os.getenv("NEO4J_DATABASE", "neo4j"),
 )
 

docker/entrypoint.sh

@@ -58,4 +58,6 @@ fi
 mkdir -p /tmp/nginx-cache /tmp/nginx-body
 
 log "Starting all services via supervisord..."
+# nginx needs its tmp dirs writable
+mkdir -p /tmp/nginx-cache /tmp/nginx-body /tmp/nginx-run
 exec /usr/bin/supervisord -c /app/docker/supervisord.conf

docker/supervisord.conf

@@ -19,16 +19,16 @@ command=/opt/neo4j/bin/neo4j console
 environment=NEO4J_HOME=/opt/neo4j,JAVA_HOME=/usr/lib/jvm/java-17-openjdk-amd64
 autostart=true
 autorestart=true
-startsecs=30
+startsecs=45
 startretries=3
 stdout_logfile=/tmp/neo4j.log
 stderr_logfile=/tmp/neo4j.log
 redirect_stderr=true
 priority=10
 
-# ── FastAPI backend ────────────────────────────────────────────────────────────
+# ── FastAPI backend (starts after Neo4j is ready via retry logic in neo4j_setup.py) ──
 [program:backend]
-command=python3 -m uvicorn main:app --host 127.0.0.1 --port 8000 --workers 2
+command=bash -c "sleep 50 && python3 -m uvicorn main:app --host 127.0.0.1 --port 8000 --workers 1"
 directory=/app/backend
 environment=
     NEO4J_URI="bolt://127.0.0.1:7687",
@@ -40,7 +40,7 @@ environment=
     OPENAI_MODEL="%(ENV_OPENAI_MODEL)s"
 autostart=true
 autorestart=true
-startsecs=10
+startsecs=15
 startretries=5
 stdout_logfile=/tmp/backend.log
 stderr_logfile=/tmp/backend.log