Convert from Nomad to Consul service discovery

6min
|
Nomad
Consul

Nomad offers native service discovery but this may not be sufficient for some larger and more complex designs and implementations.

Integrating Consul into your application design allows you to continue to use service discovery and additional features not offered by Nomad’s service discovery, such as health checks, traffic shaping, and more.

Open the Consul UI

Consul is running alongside Nomad on the cluster you set up as part of the deployment in the previous tutorial.

Open the Consul UI by visiting the link in the Terraform output, logging in with the token from that output, and clicking on the Services page from the left navigation.

$ terraform output
# ...
The Consul UI can be accessed at http://ar-nomad-server-lb-2016482221.ca-central-1.elb.amazonaws.com:8500/ui
with the bootstrap token: 8f94ee77-bc50-4ba1-bf75-132ed6b9366e

This page shows the services currently registered with Consul. Notice that the Nomad and Consul agents running on the server and client machines are the only services present.

Consul UI Services page

Update the job spec to use Consul

In the previous tutorial, you set up the HashiCups application using native service discovery in Nomad.

Switching to Consul service discovery involves updating the provider value and the query function present in any template stanzas.

Nomad SD example

job "example-job" {
  group {
    task "simple-task" {
      service {
        name = "simple-service"
        port = "8080"
        provider = "nomad"
      }
    }
  }
  group "product-api" {
    task "product-api" {
    ## ...
        template {
            data        = <<EOH
{{ range nomadService "database" }}
DB_CONNECTION="host={{ .Address }} port={{ .Port }} user=user password=password dbname=db_name"
{{ end }}
EOH
            destination = "local/env.txt"
            env         = true
        }
    }
  }
}

Consul SD example

job "example-job" {
  group {
    task "simple-task" {
      service {
        name = "simple-service"
        port = "8080"
        provider = "consul"
      }
    }
  }
  group "product-api" {
    task "product-api" {
    ## ...
        template {
            data        = <<EOH
{{ range service "database" }}
DB_CONNECTION="host={{ .Address }} port={{ .Port }} user=user password=password dbname=db_name"
{{ end }}
EOH
            destination = "local/env.txt"
            env         = true
        }
    }
  }
}

Open the HashiCups job spec file, replace each instance of provider = "nomad" with provider = "consul", and save the file. There are six lines to update, one for each of the HashiCups services. This registers the services with Consul instead of Nomad.

hashicups.hcl

job "hashicups" {
  group "db" {
    # ...
    task "db" {
      # ...
      service {
        name = "database"
        port = "db"
        provider = "consul"
        # ...
      }
    }
  }
}

In the same file, replace each instance of nomadService with service, and save the file. There are six lines to update. This instructs Nomad to query Consul for the service address instead of its own internal service catalog.

hashicups.hcl

job "hashicups" {
  group "product-api" {
    # ...
    task "product-api" {
      # ...
      template {
        data        = <<EOH
{{ range service "database" }}
DB_CONNECTION="host={{ .Address }} port={{ .Port }} user=${var.postgres_user} password=${var.postgres_password} dbname=${var.postgres_db} sslmode=disable"
BIND_ADDRESS = "{{ env "NOMAD_IP_product-api" }}:${var.product_api_port}"
{{ end }}
EOH
        destination = "local/env.txt"
        env         = true
      }
    }
  }
}

Finally, be sure to update the load balancing configurations in the nginx task to match the Consul service example below.

hashicups.hcl

group "nginx" {
  # ...
  task "nginx" {
    # ...
    template {
      data =  <<EOF
proxy_cache_path /var/cache/nginx levels=1:2 keys_zone=STATIC:10m inactive=7d use_temp_path=off;
upstream frontend_upstream {
  {{ $allocID := env "NOMAD_ALLOC_ID" }}
  {{ range nomadService 1 $allocID "frontend" }}
  server {{ .Address }}:{{ .Port }};
{{ end }}
}
# ...
  }
}

hashicups.hcl

group "nginx" {
  # ...
  task "nginx" {
    # ...
    template {
      data =  <<EOF
proxy_cache_path /var/cache/nginx levels=1:2 keys_zone=STATIC:10m inactive=7d use_temp_path=off;
upstream frontend_upstream {
{{ range service "frontend" }}
  server {{ .Address }}:{{ .Port }};
{{ end }}
}
# ...
  }
}

Navigate back to your terminal and re-submit the HashiCups job spec to Nomad.

$ nomad job run hashicups.hcl
==> 2022-08-10T15:13:38-04:00: Monitoring evaluation "1198ebaa"
    2022-08-10T15:13:38-04:00: Evaluation triggered by job "hashicups"
    2022-08-10T15:13:38-04:00: Allocation "2e4d9ce1" created: node "0c40d12b", group "frontend"
    2022-08-10T15:13:38-04:00: Allocation "789dfe9e" created: node "bacfe7b3", group "nginx"
    2022-08-10T15:13:38-04:00: Allocation "9a70e22b" created: node "bacfe7b3", group "product-api"
    2022-08-10T15:13:38-04:00: Allocation "d6092b2a" created: node "bacfe7b3", group "public-api"
    2022-08-10T15:13:38-04:00: Allocation "7dac8068" modified: node "d51c20d9", group "payments-api"
    2022-08-10T15:13:38-04:00: Allocation "fecf17cd" modified: node "0c40d12b", group "db"
    2022-08-10T15:13:39-04:00: Evaluation within deployment: "21e2f157"
    2022-08-10T15:13:39-04:00: Allocation "2e4d9ce1" status changed: "pending" -> "running" (Tasks are running)
    2022-08-10T15:13:39-04:00: Evaluation status changed: "pending" -> "complete"
==> 2022-08-10T15:13:39-04:00: Evaluation "1198ebaa" finished with status "complete"
==> 2022-08-10T15:13:39-04:00: Monitoring deployment "21e2f157"
  ✓ Deployment "21e2f157" successful
    
    2022-08-10T15:14:34-04:00
    ID          = 21e2f157
    Job ID      = hashicups
    Job Version = 1
    Status      = successful
    Description = Deployment completed successfully
    
    Deployed
    Task Group    Desired  Placed  Healthy  Unhealthy  Progress Deadline
    db            1        1       1        0          2022-08-10T19:23:49Z
    frontend      2        2       2        0          2022-08-10T19:24:18Z
    nginx         2        2       2        0          2022-08-10T19:24:32Z
    payments-api  1        1       1        0          2022-08-10T19:23:49Z
    product-api   1        1       1        0          2022-08-10T19:23:50Z
    public-api    1        1       1        0          2022-08-10T19:23:53Z

Navigate to the Services page in the Consul UI from earlier and note that each of the HashiCups services are now registered with Consul.

HashiCups services registered in the Consul UI

This job uses same the load balancer URL from the previous tutorial. The URL can also be found in the Consul UI by clicking on one of the nginx services and then the instance task name.

Note

Your browser may cache the HashiCups page resulting in an error when loading the products on the page. If this happens, try a hard refresh of the page or open it in a private browser window.

nginx service task in the Consul UI

You may need to scroll to the right in your browser window as the service URL is printed after the task name.

nginx service details in the Consul UI

Clean up

Run terraform destroy to clean up your provisioned infrastructure. Respond yes to the prompt to confirm the operation.

$ terraform destroy
Plan: 0 to add, 0 to change, 18 to destroy.
## ...
module.nomad-cluster.aws_instance.server[0]: Destruction complete after 30s
module.nomad-cluster.aws_instance.server[1]: Destruction complete after 30s
module.nomad-cluster.aws_instance.server[2]: Destruction complete after 30s
module.nomad-cluster.aws_iam_instance_profile.instance_profile: Destroying... [id=ar-nomad20220810185103967100000002]
module.nomad-cluster.aws_security_group.primary: Destroying... [id=sg-097e033fb088560d6]
module.nomad-cluster.aws_iam_instance_profile.instance_profile: Destruction complete after 0s
module.nomad-cluster.aws_iam_role.instance_role: Destroying... [id=ar-nomad20220810185103035400000001]
module.nomad-cluster.aws_security_group.primary: Destruction complete after 0s
module.nomad-cluster.aws_security_group.server_lb: Destroying... [id=sg-01a751f96df44445e]
module.nomad-cluster.aws_iam_role.instance_role: Destruction complete after 0s
module.nomad-cluster.aws_security_group.server_lb: Destruction complete after 1s

Destroy complete! Resources: 18 destroyed.

Be sure to follow the additional cleanup steps around AMIs outlined in the AWS cluster setup tutorial so you don't incur any additional charges for image storage.

Next steps

In this collection of tutorials, you learned about the native service discovery present in Nomad, how to enable services to use it, and how to transition from native service discovery to Consul service discovery.

Check out the running Nomad workloads on edge nodes tutorial for another example that uses Nomad service discovery.

Deploy an app

Next Collection

Templates