Telemetry events emitted by the driver.

Attach to these to get request rates, latency and error breakdowns without instrumenting call sites.

[:typedb, :request, :start | :stop | :exception]

One span per HTTP request, including each retry and each renew-and-retry attempt — so a single TypeDB.query/4 can produce several spans.

Measurements follow :telemetry.span/3: :system_time on start, :duration (native units) on stop and exception.

Metadata:

• :connection — the connection name
• :method — :get, :post, :put or :delete
• :path — the API path, e.g. "/transactions/open". Database and user names are not stripped, so treat it as potentially high-cardinality
• :attempt — 1-based attempt number within this request
• :status — HTTP status, on :stop only
• :error — the TypeDB.Error when one was produced, on :stop only

[:typedb, :sign_in, :start | :stop | :exception]

One span per sign-in. A healthy connection produces one on start-up and one per token lifetime; a spike means tokens are being rejected.

Metadata: :connection, and :error on failure.

Example

:telemetry.attach_many(
  "typedb-logger",
  [[:typedb, :request, :stop], [:typedb, :sign_in, :stop]],
  fn event, %{duration: duration}, metadata, _config ->
    Logger.info("#{inspect(event)} #{System.convert_time_unit(duration, :native, :millisecond)}ms " <>
                  "#{inspect(Map.take(metadata, [:method, :path, :status]))}")
  end,
  nil
)

Prometheus / telemetry_metrics

Telemetry.Metrics.distribution("typedb.request.stop.duration",
  unit: {:native, :millisecond},
  tags: [:method, :status]
)