> For the complete documentation index, see [llms.txt](https://documentation.alluxio.io/k8s-operator-en/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://documentation.alluxio.io/k8s-operator-en/release-notes/operator-3-7-0.md).

# Alluxio Kubernetes Operator 3.7.0 Release Notes

## Summary

### New Features

* **Bare-metal deployment** gives you an Ansible collection for deploying and managing Alluxio clusters outside Kubernetes.
* **Built-in observability** deploys Prometheus, Grafana, and Alertmanager by default, with curated dashboards and ready-to-use monitoring for Operator and CSI components.
* **CSI FUSE self-healing** introduces FuseAttachment, a node-scoped resource that improves FUSE mounting, recovery, and cleanup.

### Major Enhancements

* **CSI FUSE reliability** improves automatic replacement of failed FUSE pods, mount cleanup, concurrent-mount handling, and shared-mount recovery.
* **Operational resilience** uses safer default resource requests and limits to reduce pod eviction risk, with clearer logs and error messages for faster diagnosis.

### Bug Fixes

* **Shared CSI FUSE mounts** now recover correctly when one of several workloads sharing a mount is removed after a FUSE pod restart.
* **Diagnostics service discovery** now selects the correct pods.

## Breaking Changes

* The Gemini (cacheOnly) integration, including its custom resource interface and Helm charts, has been removed.
* CSI FUSE now uses the node-scoped FuseAttachment resource. You must start with a clean state for affected node-local CSI FUSE resources when upgrading.

## Upgrade Notes

* Before upgrading, remove or replace any Gemini (cacheOnly) resources and deployment automation.
* Follow the CSI FUSE migration procedure, then validate mounts and recovery behavior on every node.