Diagnostika

Správa runů
- Diagnostika
  - Diagnostika a inspekce output akcí - run jede ale nedělá to co chci
  - Diagnostika komponent runu - run nejede nebo potřebuji větší vhled do runu
  - Notifikace

Na obecných postupech se opět asi nic moc nezměnilo, jen to chce aktualizovat jak se k věcem dostat. Možná i víc rozpracovat co znamenají. A přidat notifikace.

This section helps you to diagnose problems and see if the Run works as expected.

Input data

Sometimes it may be useful to inspect the data the Scenario processes. Often it’s unclear whether a situation is a problem in the input data or the implementation, and it’s crucial to identify the component which misbehaves.

For details about inspecting Runtime topics see the chapter for Normalizer Diagnostics. Note the Scenario only works with the main Runtime topic.

Side effects / Output actions

Inspecting the side effects is useful to see if Run behaves as it should. You can use the following APIs in the developer console to inspect output actions.

Runtime.list_run_output_actions(run_id) - returns a list of all output actions. Note the list of all output actions might get big for Runs that run for a longer period of time. Use with care ideally only if you know there isn’t many output actions.
Runtime.stream_run_output_actions(run_id) - returns a stream of all output actions. The Stream is implemented efficiently, it only reads the data if it is requested. It is recommended to read up on Streams to be most effective with them.

Built-in Run diagnostics

The built-in Run diagnostics provides information about the distribution of output actions in time. It is similar to Normalizer Diagnostics and provides the number of output actions (successful and failed, separately) in the last minutes, hours, days and weeks.

Run diagnostic is available in the Admin UI - locate the Run in the Scenarios section and you will see a Diagnostics button in the top right corner. Developer console provides Runtime.runs_diagnostics() which returns diagnostic information for all running runs.

Run diagnostics only inspects the last 10000 output actions by default, if you see at least X in the output, it means the interval wasn’t read in its entirety and therefore, we only know there’s at least X output actions in the interval, there could be more.