Open
Description
A customer has requested that per-bucket replication be enhanced so as to have target clusters. Suppose we have three clusters, A, B and C, and two buckets b1 and b2. If cluster A is ingest and has both A.b1 & A.b2, then this feature would enable users to configure replication as follows:
A.b1 --> B.b1
A.b2 --> C.b2
Since efficient fullsync replication uses AAE hash trees, per-bucket AAE is a prerequisite of this feature.