resharding
Changing how data is split across shards, such as adding shards, while the system keeps running.
In more detail
Resharding is a live migration: copy the data, keep the copy current, then cut over while fencing writes to the old owner. It is slow and risky, so it should be planned from growth estimates rather than started when a disk is already full.
Related terms
- shard — One piece of a database that has been split across machines, holding a subset of the data.
- live migration — Moving data or traffic to a new home while the system keeps serving requests.
- cutover — The moment traffic or ownership switches from an old system or shard to a new one.
- fencing — Cutting a failed or doubtful member of a replicated system off from its data so it cannot keep writing while you recover.
Taught in
- Back Of The Envelope Estimation: Storage And Bandwidth — Scalable System Design & Distributed Architecture
- Partitioning, Sharding And Hot Keys — Scalable System Design & Distributed Architecture