Home Arrow Icon Knowledge base Arrow Icon Global Arrow Icon How does MongoDB handle data migration between shards
    MongoDB data migration shards balancer process chunk migration resharding sharding database management MongoDB monitoring jumbo chunks


How does MongoDB handle data migration between shards


MongoDB handles data migration between shards through the balancer process. The balancer is responsible for ensuring that data is evenly distributed across the shards in a sharded cluster. Here are the steps involved in data migration between shards:

1. Balancing

- Automatic Balancing: MongoDB automatically balances data by migrating chunks from one shard to another. This ensures that data is evenly distributed across the shards.

2. Chunk Migration

- Manual Migration: MongoDB can perform manual chunk migration for specific collections. This is useful for distributing data during bulk inserts.

3. Range Migration

- Automatic Range Migration: MongoDB migrates data one range at a time. The balancer does not wait for the current migration's delete phase to complete before starting the next range migration.

4. Resharding

- Reshard-to-Shard Technique: MongoDB uses resharding to spread data across the shards in the cluster quickly. This involves intentionally sharding into a temporary shard key that is different from the desired shard key and then resharding into the desired shard key.

5. Chunk Migration Procedure

- Source Shard: The source shard starts the move when it receives an internal command. During the migration process, operations to the chunk are sent to the source shard.

- Destination Shard: The destination shard builds any indexes required by the source that do not exist on the destination. The destination shard begins requesting documents in the chunk and starts receiving copies of the data.

- Synchronization: After receiving the final document in the chunk, the destination shard starts a synchronization process to ensure that it has the changes to the migrated documents that occurred during the migration.

- Metadata Update: The source shard connects to the config database and updates the cluster metadata with the new location for the chunk.

- Delete Phase: After the source shard completes the update of the metadata, and once there are no open cursors on the chunk, the source shard deletes its copy of the documents[3][5].

6. Jumbo Chunks

- Jumbo Chunks: If a chunk exceeds the maximum size or the number of documents in the chunk exceeds the maximum allowed, MongoDB does not migrate the chunk. Instead, MongoDB attempts to split the chunk. If the split is unsuccessful, MongoDB labels the chunk as jumbo to avoid repeated attempts to migrate the chunk[3].

7. Monitoring and Maintenance

- Monitoring: MongoDB provides tools to monitor the balancer process and ensure that data is evenly distributed across the shards.

- Maintenance: MongoDB provides commands to manage the balancer and ensure that data migration is efficient and scalable.

Conclusion

MongoDB handles data migration between shards through the balancer process, which ensures that data is evenly distributed across the shards in a sharded cluster. The balancer migrates data one range at a time, and MongoDB provides tools to monitor and manage the balancer process.

Citations:
[1] https://stackoverflow.com/questions/48438644/migration-data-between-shards-with-mongodb
[2] https://hackernoon.com/optimize-data-migration-in-mongodb-resharding-techniques-for-speed-and-scalability
[3] https://www.mongodb.com/docs/v3.0/core/sharding-chunk-migration/
[4] https://www.mongodb.com/community/forums/t/shard-migration-leads-to-service-disruption/247228
[5] https://www.mongodb.com/docs/v6.2/core/sharding-balancer-administration/