Notes
This sprint we were mostly focused on bugfixes and internally facing improvements. We fixed a bug in the journal entry lines endpoint that was causing duplication among the lines in the output. There’s now only one row in the results for each journal entry line, and we’ve added the columns we use for deduplication (journal line Workday ID and a raw updated timestamp) to the API output. We also fixed a bug related to filtering catalog items on their catalog name.
A couple of sprints ago we talked about using a MySQL JSON datafield for certain related records instead of a parent-child table relationship. That’s been successful, so this sprint we removed the redundant tables and simplified our ETL.
We migrating the caching layer behind the FDM validation endpoint from node-based Redis to serverless Valkey. Valkey is a Redis replacement that Amazon is gradually pushing users toward, so this helps prevent future issues down the road if Redis availability on AWS becomes limited.
API changes
- New journalline_wid and raw_updated_ts columns in journal entry line output