Adding telemetry to an application

Jupyter Telemetry enables you to log events from your running application. (It’s designed to work best with traitlet’s Application object for simple configuration.) To use telemetry, begin by creating an instance of EventLog:

from jupyter_telemetry import EventLog

class MyApplication:

    def __init__(self):
        ...
        # The arguments
        self.eventlog = EventLog(
            ...
            # Either pass the traits (see below) here,
            # or enable users of your application to configure
            # the EventLog's traits.
        )

EventLog has two configurable traits:

- `handlers`: a list of Python's logging handlers that handle the recording of incoming events.
- `allowed_schemas`: a dictionary of options for each schema describing what data should be collected.

Next, you’ll need to register event schemas for your application. You can register schemas using the register_schema_file (JSON or YAML format) or register_schema methods.

Once your have an instance of EventLog and your registered schemas, you can use the record_event method to log that event.

# Record an example event.
event = {'name': 'example event'}
self.eventlog.record_event(
    schema_id='url.to.event.schema',
    version=1,
    event=event
)

previous

Writing a schema for telemetry