Hi @Rocky, everyone, Is anyone using the Automated Cluster Snapshot feature in StarRocks and how stable is it?

The Automated Cluster Snapshot is a relatively new feature in StarRocks, introduced in version 3.4.2. It is designed specifically for shared-data (cloud-native) clusters to simplify disaster recovery by automatically backing up both data and metadata (FE state) to object storage. Stability and Status * Beta Status: According to the official documentation, this feature is currently marked as Beta. This means it is functional but may still have bugs or evolving behavior, and it is generally recommended to test it thoroughly in a non-production environment first. * Known Issues: Community feedback and GitHub reports indicate some early-stage stability issues, particularly related to specific storage backends: * Cloud Compatibility: There have been reported issues with uploading snapshots to Azure Blob Storage (e.g.,

UnsupportedFileSystemException

or credential issues). * S3-Compliant Storage: Issues have been noted with MinIO where path-style access configurations might not be correctly applied by the FE during the upload process. * Endpoint Modification: Some users reported that the AWS SDK modified the specified endpoint, causing connection failures to private S3 services. Key Characteristics * Scope: Only available for shared-data clusters. * Retention: It automatically maintains only one (the latest) snapshot. Historical snapshots are deleted once a new one is successfully created. * Mechanism: A snapshot is triggered every time the FE completes a metadata checkpoint (controlled by

automated_cluster_snapshot_interval_seconds

, default 600 seconds). * What's included: All cluster objects (databases, tables, users, privileges, etc.), but it excludes external configurations like local UDF JARs or external catalog config files. Recommendation If you plan to use it, you should: 1. Verify your storage: Ensure your storage volume is correctly configured and that the FE has the necessary permissions/credentials to write to it. 2. Monitor Jobs: Use

SELECT * FROM information_schema.cluster_snapshot_jobs;

to track the success or failure of automated tasks. 3. Check Maturity: Reference the StarRocks Feature Maturity guide for advice on using Beta features. For critical production workloads, you may want to complement this with traditional manual backups until the feature reaches General Availability (GA). References • administration/cluster_snapshot.md • administration/management/FE_configuration.md