Data Storage

This page owns the project storage decision.

Current Choice

The project uses CSV files stored by ab-ticket-bot-jobs.

Files:

  • ticeket-mapping.csv
  • author-conversation.csv
  • ab_ticket_state-meta.csv
  • ab_ticket_state-processed-message-ids.csv
  • ab_ticket_state-processed-fingerprints.csv
  • ab_ticket_state-last-ab-by-author.csv

Why CSV Is Enough Today

  • one long-running worker is the only writer
  • the state model is small and append-or-rewrite oriented
  • operational debugging is easier when operators can inspect files directly

Upgrade Trigger

Move to PostgreSQL when one or more of these become true:

  • multiple writers need safe concurrent updates
  • richer querying becomes necessary
  • data volume makes flat-file rewrite patterns brittle
  • auditability and migrations need stronger guarantees