Waypont deploy fails: ! transport is closing

Waypoint
#1

Hi,
I am new to waypoint and I am using it with gitlab on aws to deploy a simple web-application. I followed the instruction (Integrating Waypoint with GitLab CI/CD | Waypoint by HashiCorp) and everything worked fine in the beginning. After developing the application further, I am facing now the issue that every running pipeline fails.

$ waypoint deploy
Deploying...
Found existing ECS cluster: waypoint
Found existing IAM role to use: ecr-xyz-core
! transport is closing
Cleaning up file based variables
ERROR: Job failed: exit code 1

Does anyone have a hint for me what exactly goes wrong and how I can fix that?

#2

Welcome! If you’re running 0.2.2 and using aws-ecs, this is likely a known bug that we’ll be doing a release today to correct. In the meantime, you can use v0.2.1 or try adding an empty logging {} block to your deploy, for example:
use “aws-ecs” {
region = “us-east-1”
memory = “512”
logging {}
}

If this is doesn’t fix it, please share your waypoint hcl for more troubleshooting.

#3

Thanks for the feedback, the workaround logging {} worked.
thank you! :+1:

1 Like
#4

We just released 0.2.3 which includes this fix. Glad the workaround did the trick as well!
https://discuss.hashicorp.com/t/waypoint-0-2-3-released/21323/2

#5

Hi
I just update to 0.2.3 and still get the error, even with the logging {} trick.

Deploying…

  • Found existing ECS cluster: waypoint
  • Found existing IAM role to use: ecr-example-nodejs
  • Using default subnets for Service networking
    ! transport is closing

as you can see I am just trying the AWS tutorial.

#6

@fmiche did you update both your local CLI and remote server versions?

#7

I was able to successfully install a 0.2.3 server both locally on Docker and remotely on EKS, and successfully run waypoint up on each installation with the below hcl file. Let me know if you are able to verify your server version and we’ll go from there!

project = "example-nodejs"

app "example-nodejs" {
  labels = {
    "service" = "example-nodejs",
    "env" = "dev"
  }

  build {
    use "pack" {}
    registry {
      use "aws-ecr" {
        region = "us-east-1"
        repository = "waypoint-example"
        tag = "latest"
      }
    }
  }

  deploy {
    use "aws-ecs" {
      region = "us-east-1"
      memory = "512"
    }
  }
}
#8

@krantzinator I checked local CLI and docker Waypoint server. they are both 0.2.3.
© 2021 HashiCorp

  • Waypoint v0.2.3
  • API: v1
  • Entrypoint: v1

Still getting

» Deploying…

  • Found existing ECS cluster: waypoint
  • Found existing IAM role to use: ecr-example-nodejs
  • Using default subnets for Service networking
    ! transport is closing
#9

@fmiche are you able to run the up command with -vvv and see if any useful ERROR messages come through?

#10

@krantzinator : Here it is:
» Deploying…

  • Found existing ECS cluster: waypoint
  • Found existing IAM role to use: ecr-example-nodejs
  • Using default subnets for Service networking
    ! transport is closing
    2021-03-05T09:34:12.841+0100 [ERROR] waypoint.runner.config_recv: error receiving configuration, exiting: err=“rpc error: code = Unknown desc = context canceled”
    2021-03-05T09:34:12.841+0100 [TRACE] waypoint.runner.config_recv: exiting receive goroutine
    2021-03-05T09:34:12.841+0100 [TRACE] waypoint: stopping signal listeners and cancelling the context
    2021-03-05T09:34:12.841+0100 [WARN] waypoint: context cancelled, stopping interrupt listener loop
#11

We need to figure out if the error is bubbling up as part of attempting to create the default subnets, or as part of creating the ALB after the subnet step completes.
Do you have useful logs on your AWS side?