执行rocketmq_v1alpha1_broker_cr.yaml，size 2改成1的时候，没有删除旧的 Broker 组对应的 StatefulSet 和 Pod，缩容现在还不支持吗？

[snow-netizen]

[zhihui921016]

您好，您的邮件我已经收到。

[oss-sentinel-ai]

Assessment by github-manager-bot

Type: Question
Priority: Medium

Summary

User is asking about scale-down support in rocketmq-operator. When changing size from 2 to 1 in rocketmq_v1alpha1_broker_cr.yaml, the old Broker StatefulSet and Pod are not deleted.

Analysis

This is a valid question about operator functionality. The issue title indicates the user wants to know if scale-down is supported.

Suggestions

1. Check if the operator version supports scale-down operations
2. Review operator documentation for scale-down procedures
3. Consider filing a feature request if scale-down is not currently supported

Verdict

Valid question — Requires clarification on operator scale-down capabilities.

---

Automated assessment by github-manager-bot

[RockteMQ-AI]

Issue Evaluation

Category: unclassified | Status: Needs Triage

This issue could not be automatically classified. A maintainer will review it manually.

---

Automated evaluation by RockteMQ-AI

[RockteMQ-AI]

Issue Evaluation

Category: question

This is a question about scale-down behavior in the RocketMQ Operator.

Currently, the RocketMQ Operator does not fully support automatic cleanup when scaling down Broker replicas (e.g., from size 2 to 1). The remaining StatefulSet pod retains its ordinal index, and the removed pod's data (CommitLog, ConsumeQueue) is not automatically migrated or cleaned up.

Current behavior:

• Scaling down removes the highest-ordinal pod
• The old pod's PVC may remain (depending on persistentVolumeClaimRetentionPolicy)
• Queue assignments on the removed broker need to be rebalanced by consumers

Recommendation:

• Before scaling down, ensure all message queues on the target broker have been migrated
• Use mqadmin updateSubGroup and mqadmin deleteTopic to clean up if needed
• This is a known limitation — consider filing a feature request for automated scale-down with data migration

---

Automated evaluation by RockteMQ-AI

[RockteMQ-AI]

Hi @snow-netizen! 👋

Thanks for your contribution. This issue has been noted and flagged for maintainer triage.

---

🤖 Automated issue analysis by github-manager

[RockteMQ-AI]

Issue Evaluation

Category: question

Scale-down (reducing broker group count) is not currently supported by the RocketMQ Operator. Changing size from 2 to 1 in the CR does not automatically clean up the removed broker group's StatefulSet and Pods.

This is a known limitation. The operator currently supports scale-up but not scale-down, as scale-down requires careful handling of:

• Message data migration from removed brokers
• Consumer rebalance coordination
• Topic reassignment

Recommendation: For now, manually delete the removed broker group's StatefulSet and Pods after updating the CR. Full scale-down support would require a significant enhancement to the operator.

---

Automated evaluation by github-manager

[RockteMQ-AI]

Issue Evaluation

Category: question

Scale-down (reducing broker group size) is not fully supported in the current operator. When you reduce the size field, the operator removes the StatefulSet replicas but does not clean up the associated resources (StatefulSet, PVC, etc.) or handle the graceful migration of topics/queues from the removed brokers.

This is a known limitation. Manual cleanup is currently required after reducing the broker group size.

---

Automated evaluation by RockteMQ-AI