From 1a574c2f7f3c2ebf6a54dd59ef2e63e574130736 Mon Sep 17 00:00:00 2001
From: Nicolas Will <nicolaswill@github.com>
Date: Sat, 15 Jun 2024 00:39:21 +0200
Subject: [PATCH] Add RabbitMQ connect retry and healthcheck

---
 cmd/agent/main.go  | 23 ++++++++++++++++++++++-
 docker-compose.yml |  6 ++++--
 2 files changed, 26 insertions(+), 3 deletions(-)

diff --git a/cmd/agent/main.go b/cmd/agent/main.go
index 83ae80a..6326323 100644
--- a/cmd/agent/main.go
+++ b/cmd/agent/main.go
@@ -100,11 +100,32 @@ func InitializeQueue(jobsQueueName, resultsQueueName string) (*RabbitMQQueue, er
 
 	rabbitMQURL := fmt.Sprintf("amqp://%s:%s@%s:%s/", rabbitMQUser, rabbitMQPassword, rabbitMQHost, rabbitMQPort)
 
-	conn, err := amqp.Dial(rabbitMQURL)
+	const (
+		tryCount      = 5
+		retryDelaySec = 3
+	)
+
+	var conn *amqp.Connection
+	var err error
+
+	for i := 0; i < tryCount; i++ {
+		slog.Info("Attempting to connect to RabbitMQ", slog.Int("attempt", i+1))
+		conn, err = amqp.Dial(rabbitMQURL)
+		if err != nil {
+			slog.Warn("Failed to connect to RabbitMQ: %w", err)
+			if i < tryCount-1 {
+				slog.Info("Retrying in %d seconds", retryDelaySec)
+				time.Sleep(retryDelaySec * time.Second)
+			}
+		}
+	}
+
 	if err != nil {
 		return nil, fmt.Errorf("failed to connect to RabbitMQ: %w", err)
 	}
 
+	slog.Info("Connected to RabbitMQ")
+
 	ch, err := conn.Channel()
 	if err != nil {
 		conn.Close()
diff --git a/docker-compose.yml b/docker-compose.yml
index 97e3a8e..a74cbac 100644
--- a/docker-compose.yml
+++ b/docker-compose.yml
@@ -32,8 +32,10 @@ services:
     networks:
       - backend
     healthcheck:
-      test: [ "CMD", "rabbitmqctl", "status" ]
-      interval: 1s
+      test: [ "CMD", "nc", "-z", "localhost", "5672" ]
+      interval: 5s
+      timeout: 15s
+      retries: 1
 
   server:
     build: