compose

docker-compose deployment

This directory contains a standalone compose file that adds a scraper cron sidecar on top of the base docker-compose.yml in the repo root.

Usage

From the repo root:

docker compose -f docker-compose.yml -f deploy/compose/docker-compose.scraper.yml up -d

This starts:

• db: Postgres 16 (from the base file)
• api: the GraphQL server on :4000 (from the base file)
• scraper: runs c3x-pricing-api scrape --vendor all once a day at 03:00 UTC via a tiny cron loop. The container sleeps otherwise.

To trigger a scrape manually:

docker compose exec scraper /app/c3x-pricing-api scrape --vendor aws

Why a sidecar and not docker compose run?

• Single compose invocation brings the whole system up.
• Persistent container makes scrape timing visible in docker compose logs scraper.
• No host-level cron required.