Back to Floor
Indexer Clustering • Room 2

Bucket Replication

When an indexer receives data, it writes to a hot bucket. That bucket is the "primary" copy. The Cluster Manager instructs other peers to create "non-primary" copies to satisfy the RF.

If a peer goes down, the Cluster Manager initiates "bucket fixup" — redistributing copies across remaining peers to restore the RF.

Knowledge Check

Prove your understanding to clear the room (Rewards XP)
❤️❤️❤️
Question 1 of 1
In a 3-peer cluster with RF=3 and SF=2, one peer goes offline permanently. What happens?
ANothing; data is lost.
BThe Cluster Manager initiates bucket fixup, creating new copies on the remaining 2 peers to restore the replication and search factors.
CThe entire cluster shuts down.
DUsers must manually copy buckets.