Required

Configure your application name. May contain up to 3 names seperated by ;

Application name can be configured in two ways:

• Environment variable: NEW_RELIC_APP_NAME=MyApp
• Application config: config :new_relic_agent, app_name: "MyApp"

An optional list of key/value pairs that will be automatic custom attributes on all event types reported (Transactions, etc). Values are determined at Agent start.

Options:

• {:system, "ENV_NAME"} Read a System ENV variable
• {module, function, args} Call a function.
• "foo" A direct value

This feature is only configurable with Application config.

Example:

config :new_relic_agent,
  automatic_attributes: [
    environment: {:system, "APP_ENV"},
    node_name: {Node, :self, []},
    team_name: "Afterlife"
  ]

Some Agent features can be controlled via configuration.

Logs In Context

This feature can be run in multiple "modes":

• forwarder The recommended mode which formats outgoing logs as JSON objects ready to be picked up by a Log Forwarder
• direct Logs are buffered in the agent and shipped directly to New Relic. Your logs will continue being output to their normal destination.
• disabled (default)

Logs In Context can be configured in two ways:

• Environment variable NEW_RELIC_LOGS_IN_CONTEXT=forwarder
• Application config config :new_relic_agent, logs_in_context: :forwarder

Infinite Tracing

Infinite Tracing gives you more control of sampling by collecting 100% of Spans and sending them to a Trace Observer for processing.

You can configure your Trace Observer in two ways:

• Environment variable NEW_RELIC_INFINITE_TRACING_TRACE_OBSERVER_HOST=trace-observer.host
• Application config config :new_relic_agent, infinite_tracing_trace_observer_host: "trace-observer.host"

Some Agent features can be toggled via configuration.

Security

• :error_collector_enabled (default true)
  • Toggles collection of any Error traces or metrics
• :db_query_collection_enabled (default true)
  • Toggles collection of Database query strings
• function_argument_collection_enabled (default true)
  • Toggles collection of traced function arguments

Instrumentation

Opting out of Instrumentation means that :telemetry handlers will not be attached, reducing the performance impact to zero.

• :plug_instrumentation_enabled (default true)
  • Controls all Plug instrumentation
• :ecto_instrumentation_enabled (default true)
  • Controls all Ecto instrumentation
• :redix_instrumentation_enabled (default true)
  • Controls all Redix instrumentation
• :request_queuing_metrics_enabled
  • Controls collection of request queuing metrics

Configuration

Each of these features can be configured in two ways, for example:

• Environment variables: NEW_RELIC_ERROR_COLLECTOR_ENABLED=false
• Application config: config :new_relic_agent, error_collector_enabled: false

An optional list of labels that will be applied to the application.

Configured with a single string containing a list of key-value pairs:

key1:value1;key2:value2

The delimiting characters ; and : are not allowed in the key or value

Labels can be configured in two ways:

• Environment variables: NEW_RELIC_LABELS=region:west;env:prod
• Application config: config :new_relic_agent, labels: "region:west;env:prod"

Required

Configure your New Relic License Key.

License Key can be configured in two ways, though using Environment Variables is strongly recommended to keep secrets out of source code:

• Environment variables: NEW_RELIC_LICENSE_KEY=abc123
• Application config: config :new_relic_agent, license_key: "abc123"

Configure the Agent logging mechanism.

This controls how the Agent logs it's own behavior, and doesn't impact your applications own logging at all.

Defaults to the File "tmp/new_relic.log".

Options:

• "stdout" Write directly to Standard Out
• "Logger" Send Agent logs to Elixir's Logger
• "file_name.log" Write to a chosen file

Agent logging can be configured in two ways:

• Environment variable: NEW_RELIC_LOG=stdout
• Application config: config :new_relic_agent, log: "stdout"