Garmin Sync

Python job for syncing Garmin Connect data into PostgreSQL for Grafana dashboards.

Environment variables

The script will automatically read a file named .env from the project root.

Use placeholders like these for now:

GARMIN_EMAIL=
GARMIN_PASSWORD=
GARMIN_ACCOUNT_KEY=personal  # optional, defaults to personal
DATABASE_URL=postgresql://
TIMEZONE=Asia/Jakarta
SYNC_DAYS=7
BACKFILL_START_DATE=
BACKFILL_END_DATE=
FERNET_KEY=

For a one-off September 2025 backfill, set:

BACKFILL_START_DATE=2025-09-01
BACKFILL_END_DATE=2025-09-30

If those two vars are set, the script uses that exact inclusive date range instead of the normal rolling SYNC_DAYS window.

Run

python -m venv .venv
. .venv/bin/activate
pip install -e ".[dev]"
garmin-sync

Database setup

If you want to apply the checked-in schema file directly:

psql "$DATABASE_URL" -f sql/schema.sql

Docker image

Pick an immutable image tag, build it, and push it to the registry your cluster can pull from.

IMAGE=your-registry/garmin-sync:2026-03-20

docker build -t "$IMAGE" .
docker push "$IMAGE"

Kubernetes

This manifest expects a Kubernetes cluster that supports CronJob.spec.timeZone. Create the secret and apply the manifest in the same namespace where you want the CronJob to run. Before applying, update k8s/garmin-sync-cronjob.yaml so its image: value matches the exact tag you pushed.

Create the secret:

kubectl -n your-namespace create secret generic garmin-sync-secrets \
  --from-literal=GARMIN_EMAIL=... \
  --from-literal=GARMIN_PASSWORD=... \
  --from-literal=DATABASE_URL=... \
  --from-literal=FERNET_KEY=...

Apply the CronJob:

kubectl -n your-namespace apply -f k8s/garmin-sync-cronjob.yaml