Search code examples
redissocket.ioamazon-ecsaws-fargateamazon-elasticache

Error Code 137 while using Elasticache Redis with ECS Fargate

I am running containers on ECS Fargate and everything is working as expected but there is one issue. When I have this redis code running in my server side, the server container is failing to reach a steady state and failing.

And when I expand the api container its showing Error Code 137

const REDIS_ENDPOINT = process.env.REDIS_ENDPOINT;

const pubClient = createClient({
  url: REDIS_ENDPOINT,
});

const subClient = pubClient.duplicate();

await Promise.all([pubClient.connect(), subClient.connect()]);

Plus there are no logs given by the server container in my cloudwatch. All other containers are working as expected such as client, nginx, mongo and redis.

  • After doing some more research, I read that its due to the memory and I tried increasing that to 15 gb in my ECS task definition which is more that enough (Just for checking). I am sure its not this issue since everything is working with 6 gb when I don't have the above mentioned code.

  • My elasticache redis endpoint is also in the correct format redis://mycompany-534sdf.serverless.use1.cache.amazonaws.com:6379

  • I have port 6379 open in my security group

  • Everything is working as expected locally when I am using docker compose

Really need some help here. I am pulling my hair out as I am typing this.

Solution

  • I was using Elasticache Redis serverless and found out that it does not support pub sub capability.

    You have to create your own cluster and everything will work fine.