This content is also available at learn.palantir.com ↗ and is presented here for accessibility purposes.
Data timeliness is a chief concern for pipeline maintainers and data consumers alike. The Time Since Last Updated (TSLU) check helps ensure your pipeline is updating at the required cadence. For variety, this task will guide the installation of the TSLU check on the output of your Flight Alerts schedule build using the Data Health application rather than in Data Lineage as you did in the previous task. Either method is acceptable for adding health checks.
flight_alerts_clean
node and click on the word Open (not on the ↗
) in the fly-out menu to open the dataset in the Dataset application.1
next to it. Click the menu item to open the Data Health app as a tab inside the Dataset app.flight_alerts_raw
has new logic). If we had an actual source system with its own schedule, we'd have a better idea about how frequently we should be expecting new data to flow in from the external source. Since we don't, let's set a variable expectation based on deviations from the median TSLU.1
deviation above the median (median calculated based on the last 10 check results).By default, check creators are automatically subscribed to alerts on check failure and will receive email and/or in-app notifications depending on the configured preferences. Read this documentation entry to learn more about these settings. You should also read this page, which further unpacks the nuances of “watching” a check. Also, read this documentation page about time-based checks.
Note in the image below the ability to pause, snooze, edit, or delete each check individually.