ID

7fcd79fa-1b2f-4d43-91f8-e94ea8527de8

BE-7859 Exp 242 Investigate and backfill missing Favorites experiment data

- source :: https://hungryroot.atlassian.net/browse/BE-7859

- tags :: Hungryroot API Project

- epic :: EP-650 Exp 242: Favorites + User Cookbooks

- sprint :: HR BE Sprint 111

- status :: Status - To Do

- priority :: Priority - Medium

- points ::

Summary

Fanny provided CSV files showing customer data missing from Exp 242 Favorites reporting. Suspected cause: data exists in prod MySQL but did not make it to the data lake/warehouse.

Investigation steps

1. Check if ~CustomerTag~ rows exist in MySQL for affected customers:

#+begin_src sql

SELECT * FROM customer_tag WHERE customer_id = <id>;

#+end_src

2. *Row exists → ETL problem.* Pipeline feeding ~shared_model.dim_customer_tags~ missed it (replication lag, filter bug, failed job). Escalate to data eng.

3. *Row missing → app never wrote it.* Trace why ~update_customer_tags~ (~app/services/api/tag.py~) was not called for that customer. May require a backfill.

References

- Slack thread :: https://hungryroot.slack.com/archives/C0B7CUKNZ6F/p1780531230664279

- Source of truth :: ~app/models/customer_tag.py~ → DB table ~customer_tag~

- ETL chain :: Django app writes → ~customer_tag~ (MySQL) → ETL/dbt → ~shared_model.dim_customer_tags~