View Source EventStreamex.Operators.Logger.ErrorLoggerAdapter behaviour (EventStreamex v1.3.1)

Behaviour to handle errors with operators (EventStreamex.Operators.Operator).

When implementing this behaviour, you must put use a GenServer.

There is currently one adapter available and used by default: EventStreamex.Operators.Logger.LoggerAdapter. This adapter simply logs the error to the console.

The executor (EventStreamex.Operators.Executor) who catches the operator failure and calls the error logger will also execute a telemetry event so that you can see errors over time in a chart.

There are 2 kinds of errors:

Retry: The operator failed but will retry soon
Failed: The operator retried several times but failed again.

For more information about the retry strategy, see EventStreamex.Operators.Executor.

Summary

Types

retry_status()

The retry status

Callbacks

log_failed(atom, any, retry_status)

Same as log_retry/3 but when no more retry will be attempted

log_retry(atom, any, retry_status)

Logs an error with the operator, that will be retried again soon.

start_link(any)

This function will be called at startup to spawn the error logger adapter.

Functions

child_spec(init_arg)

Returns a specification to start this module under a supervisor.

Types

retry_status()

(since 1.0.0)

@type retry_status() :: %{
  max_retries: integer(),
  curr_retries: integer(),
  max_restart_time: integer(),
  backoff_multiplicator: float(),
  curr_time_to_wait: integer()
}

The retry status:

max_retries: Maximum number of retries attempted
curr_retries: The current retries count already attempted (starting at 1 after the first failure)
max_restart_time: The maximum number of milliseconds to wait for the next retry
backoff_multiplicator: A decimal number multiplied to the curr_time_to_wait to define the next amount of time to wait for the next retry (maxed by max_restart_time)
curr_time_to_wait: The current time to wait in milliseconds before the next retry

Callbacks

log_failed(atom, any, retry_status)

(since 1.0.0)

@callback log_failed(atom(), any(), retry_status()) :: :ok

Same as log_retry/3 but when no more retry will be attempted

Params

module: The module of the operator
reason: The failure reason
retry_status: The current retry count, the time to wait for the next retry, etc...

log_retry(atom, any, retry_status)

(since 1.0.0)

@callback log_retry(atom(), any(), retry_status()) :: :ok

Logs an error with the operator, that will be retried again soon.

Params

module: The module of the operator
reason: The failure reason
retry_status: The current retry count, the time to wait for the next retry, etc...

start_link(any)

(since 1.0.0)

@callback start_link(any()) :: {:ok, pid()}

This function will be called at startup to spawn the error logger adapter.

The param comes from the configuration of the adapter:

error_logger_adapter: {EventStreamex.Operators.Logger.LoggerAdapter, []},

The return value is a tuple with the pid of the spawned process

Functions

child_spec(init_arg)

(since 1.0.0)

Returns a specification to start this module under a supervisor.

See Supervisor.