Data warehouse ingestion
Pre-requisite
You first need to declare your data warehouse as a data source in mediarithmics as detailed in the data warehouse management section.
Synchronization strategies
When trying to ingest data from a data warehouse to mediarithmics, you will need to configure a service account with specific rights on the data you want to expose : the service account will need to at least have read access on the data.
Working dataset/schema
If you want to use synchronization strategies that can handle updates and deletes on existing data, you will need to create a working dataset/schema. On this dataset/schema, the service account should have editing rights : this will be used to create temporary tables and track changes on your tables.
Available strategies
Profile data
Timestamp based : all updates in the source table (insert, update, delete) are impacted in mediarithmics. This requires a timestamp column tracking the last update of each line of the table.
Hash table : all updates in the source table (insert, update, delete) are impacted in mediarithmics. A temporary table is created based on a hash of columns of the source table.
Full import : every synchronization will import the full table. It does not handle deletes or updates, this is a append-only strategy.
Event data
Timestamp based : import all events that occurred since the last import. This requires a timestamp column tracking when each event of the table has occurred. It does not handle deletes or updates, this is a append-only strategy.
Summary
Without a working dataset/schema only append-only strategies are available : timestamp based for event data, and full import for profile data.
Timestamp based
Events : No Profile : Yes
Full import
No
Hash table
Yes
Last updated
Was this helpful?