Skip to main content
Version: 16.x

Application Access High Availability (HA)

You can deploy the Application Service in a Highly Available (HA) configuration in a couple of common ways: combined instances and separate instances. Both of those revolve around pointing multiple Application Services to the same application.

Combined instances​

The most common approach is to assign the same name to each Application Service proxying the application.

If you have two Application Services proxying the same application, configuration for both agents would be identical:

# Same config for both agents.
app_service:
  enabled: "yes"
  apps:
  - name: "elastic"
    uri: https://elasticsearch.example.com:9200

With this configuration, you will see only a single entry for the application in tsh apps ls:

$ tsh apps ls
Application Description Type Public Address            Labels
----------- ----------- ---- ------------------------- -------------------
elastic                 HTTP elastic.proxy.example.com teleport.dev/origin

When connecting, Teleport will randomly pick the Application Service instance to connect through to provide some load balancing. If the selected instance is down (e.g. in case of AZ outage), Teleport will try to connect via other instances.

Separate instances​

With separate instances, each Application Service instance proxying the application assigns it a different name. This allows you to explicitly pick the agent you want to connect to the application over:

# Application service instance #1.
app_service:
  enabled: "yes"
  apps:
  # Note the name is different than instance #2 but the URI is the same.
  - name: "elastic-us-east-1a"
    uri: https://elasticsearch.example.com:9200
# Application service instance #2.
app_service:
  enabled: "yes"
  apps:
  # Note the name is different than instance #2 but the URI is the same.
  - name: "elastic-us-east-1b"
    uri: https://elasticsearch.example.com:9200

With this configuration, both services will appear as two separate entries in tsh apps ls output and you will have to pick one explicitly when connecting:

$ tsh apps ls
Application         Description Type Public Address            Labels
------------------- ----------- ---- ------------------------- -------------------
elastic-us-east-1a              HTTP elastic.proxy.example.com teleport.dev/origin
elastic-us-east-1b              HTTP elastic.proxy.example.com teleport.dev/origin

This approach is useful when you want to have control over which instance you wish to connect to.