The three fact tables, comprising the signup and plan facts encapsulated in Psyberg's stateless mode, along with the cancel fact in stateful mode, serve as inputs for the stateful sequential load ETL pipeline. This data pipeline monitors the various stages in the customer lifecycle.
By focusing solely on updates and avoiding reprocessing of data based on a fixed lookback window, both Stateless and Stateful Data Processing maintain a minimal change footprint. This approach ensures data processing is both efficient and accurate.
The Psyberg wrapper for this stateful ETL looks at the updates to the upstream Psyberg powered fact tables to determine the date-hour range to reprocess.
[
Collection
]
[
|
...
]