MirrorMaker 2双向同步场景咨询：每个需复制主题是否在链路两端各需6个系统主题？主题爆炸风险探讨

Great question—this is a super common concern when setting up bidirectional Kafka syncs with MirrorMaker 2 (MM2), so let’s break this down clearly to ease your worries.

First, let’s correct a critical misunderstanding: MM2 does NOT create 6 system topics per replicated business topic. Instead, system topics are tied to replication links (the sync relationship between a source and target cluster), not individual business topics. That’s the key detail that prevents "topic explosion."

What System Topics Does MM2 Actually Create? #

For a single unidirectional replication link (e.g., Cluster A → Cluster B), MM2 generates a fixed set of system topics (typically 5-6, depending on your config) on the target cluster. These include:

• mm2-[source-cluster]-to-[target-cluster]-heartbeats: Tracks liveness of MM2 workers
• mm2-[source-cluster]-to-[target-cluster]-checkpoints: Stores offset checkpoints for replicated topics
• mm2-[source-cluster]-to-[target-cluster]-offsets: Maps source topic offsets to target offsets
• mm2-[source-cluster]-to-[target-cluster]-status: Tracks replication task health and progress
• mm2-[source-cluster]-to-[target-cluster]-configs: Stores dynamic configuration for the replication link
• (Optional) mm2-[source-cluster]-to-[target-cluster]-topics: Used if you’re auto-discovering topics to replicate

Bidirectional Sync: How This Scales #

Bidirectional sync is just two separate unidirectional links working in tandem: Cluster A → Cluster B, and Cluster B → Cluster A. Each link gets its own set of system topics:

• On Cluster B, you’ll have the 5-6 system topics for the A→B link
• On Cluster A, you’ll have the 5-6 system topics for the B→A link

The total number of system topics here is 10-12 across both clusters—no matter how many business topics you’re replicating. Even if you sync 1000+ business topics, you still only have these 10-12 system topics. That’s a night-and-day difference from the per-topic scenario you were worried about.

Quick Tips to Keep Things Clean #

If you want to minimize clutter even further:

• Customize the system topic prefix with the mm2.prefix config to group them neatly
• Disable optional system topics if you don’t need them (e.g., turn off auto-discovery if you’re explicitly listing topics to sync)
• Use Kafka’s built-in topic filtering tools to separate system topics from your business topics in monitoring UIs

内容的提问来源于stack exchange，提问作者teeboy