Mamba Core Road Map

[GOmare]

Feature Work

• Add flag to turn ON/OFF automated flattening
• Revert all JSON functions and also indicate the MySQL version and MariaDB we support after this revert
• Auto-deploying of the mamba_etl_scheduler.sql script
• Put back ticks around ALL column labels
• Combine Patient, Person, Person Attribute and Person Name tables into one Patient Table
• Externalise the User report JSON files so that we just read them in as configs
• Add MambaETL fetch LineList reports via API (currently on aggregate values supported)

Improvements

• Test MambaETL performance on a large PROD dataset (Masaka regional referral database) - over 30 million observations or PROD data
• Look into and analyse Debezium as a better alternative solution to support/implement the incremental update feature
• Improve the implementation of the insert and update scripts on the mamba Z table. It is currently a performance chokepoint

Bugs

• Multi-select questions should include their answers as a list of comma separated values under one column (this is not working though there is an SP related (as if) to this.

House Keeping

• Make sure the scheduler Table always has max: 10 records - analysis_db._mamba_etl_schedule