Hi everyone, I'm getting a lot of

Could not reach user code server

errors. I'm using docker+ECS to run one "main" Dagster instance and one daemon. The daemon is logging the error and I get an

ERROR: closing handshake failed

on the executing instance. It occurs like 1-2 times a day. Any ideas how to debug the issue?

hi max - are you running the code server in its own task like in the example here? https://github.com/dagster-io/dagster/tree/0.14.5/examples/deploy_ecs (The

user_code

service in the docker-compose file) That may be a good idea so that if something brings the code server down you have more visibility into what's going on, and ECS can restart it

Well almost. I'm not using a dedicated user code server, I just have 2 tasks running. I'll give that a try though and add one more task. Still have to figure out how to integrate that into my CI/CD pipeline, currently I'm just replacing the image and restarting the two tasks (daemon + dagit)

Another nice thing about using a separate task for your code is that then that's the only task that you need to restart - you won't need to restart the daemon and the dagit on each change

Thanks I'll try that