juntekim.com/README.md
Jun-te Kim 45f2c5057c
Some checks are pending
Deploy Home Assistant / deploy (push) Waiting to run
Build juntekim.com / Push-to-juntekim-to-docker-hub (push) Waiting to run
Build juntekim.com / run-on-k8s (push) Blocked by required conditions
Deploy n8n / deploy (push) Waiting to run
Build & Deploy stripe-to-invoice (with DB secrets + migrations) / build (push) Waiting to run
Build & Deploy stripe-to-invoice (with DB secrets + migrations) / Deploy Postgres (PV + PVC + Deployment) (push) Blocked by required conditions
Build & Deploy stripe-to-invoice (with DB secrets + migrations) / Apply runtime secrets (push) Blocked by required conditions
Build & Deploy stripe-to-invoice (with DB secrets + migrations) / Run DB migrations (Atlas) (push) Blocked by required conditions
Build & Deploy stripe-to-invoice (with DB secrets + migrations) / deploy (push) Blocked by required conditions
Terraform Apply / Terraform Apply (push) Waiting to run
Terraform Apply / Terraform Apply - SES (push) Blocked by required conditions
updated README
2026-03-11 23:56:49 +00:00

1.7 KiB

AWS S3 + Terraform Setup (TODO - do first)

  • Create new S3 bucket for Terraform state (e.g. juntekim-terraform-state)
  • Enable versioning on the bucket
  • Set up Terraform with S3 backend pointing to new bucket
  • Use Terraform to define new infra (start with what exists, then expand)
  • Store all future infra changes via version-controlled Terraform in Forgejo

Forgejo Backup (TODO)

  • Set up restic CronJob to back up forgejo-pvc (/data) to S3
    • Mount forgejo-pvc read-only in CronJob
    • Use restic to snapshot to S3 bucket (need: S3_BUCKET, AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, RESTIC_PASSWORD)
    • Schedule: daily
  • Forgejo postgres backup via databasus (separate)
  • Test restore from restic snapshot up

Lets deploy databasus first since i'm using it to sort my back up strategy first

Migrate hyprland

figure out how to do a back up for a small pvc and pv using traefik as the example how would i back everything in ceph storage to aws like i used to do in mist cron job when i it was just the local host un mount the storage class once i got rid of everything

Services still using mist local storage (need to migrate to Ceph)

  • Uptime Kuma (uptime-kuma-pvc, 500Mi)
  • n8n (n8n-pvc, 5Gi)
  • Home Assistant (homeassistant-pvc, 10Gi)
  • DBeaver (dbeaver-pvc, 5Gi)
  • Postgres Prod (postgres-prod-pvc, 20Gi)
  • Postgres Dev (postgres-dev-pvc, 20Gi)
  • Monica (monica-storage-pvc 1Gi + monica-db-pvc 2Gi)
  • Tandoor (tandoor-media-pvc 5Gi + tandoor-postgres-pvc 2Gi)
  • Donetick (donetick-pvc, 1Gi)
  • Papra (papra-pvc, 10Gi)
  • Databasus (databasus-pvc, 500Mi)
  • wger (wger-media-pvc 5Gi + wger-postgres-pvc 2Gi + wger-static-pvc 2Gi)
  • Certs (certs-pvc, 1Mi)
  • Pihole (pihole-pv, 5Gi - Released/unused)