Kubernetes operator that watches per-user Flickr download Jobs in a namespace, restarts failed Jobs after a configurable delay, and retrieves pod logs and exit codes for failed Jobs before restarting them. Designed to run alongside Immich (self-hosted photo management).



docker run --rm xomoxcc/flickr-immich-k8s-sync-operator:latest
| Variable | Description | Default |
|---|---|---|
LOGURU_LEVEL | Log verbosity (DEBUG, INFO, WARNING, ...) | DEBUG |
NAMESPACE | Namespace to watch | flickr-downloader |
JOB_NAMES | Comma-separated Job names to monitor (required) | -- |
CHECK_INTERVAL | Seconds between check cycles | 60 |
RESTART_DELAY | Seconds to wait after failure before restart | 3600 |
SKIP_DELAY_ON_OOM | Skip restart delay when failure reason is OOMKilled | false |
The operator runs as a single-replica Deployment with a namespace-scoped ServiceAccount. See the README for full RBAC and Deployment manifests.
python:3.14-slim-trixiepythonuser)tini -- with flickr-immich-k8s-sync-operator as default CMDlinux/amd64, linux/arm64Content type
Image
Digest
sha256:62e234956…
Size
122.3 MB
Last updated
7 months ago
docker pull xomoxcc/flickr-immich-k8s-sync-operator