/ /

Sync a Replica Set to a Sharded Cluster

This documentation refers to an outdated version of mongosync. View the current documentation for up-to-date guidance on mongosync and instructions on how to upgrade to the latest version.

This page outlines the procedure to synchronize data from a replica set to a sharded cluster using mongosync.

Note

mongosync doesn't support syncing from a sharded cluster to a replica set.

Before You Begin

Define your source replica set and destination sharded cluster.
Define an administrative user that has permissions in both clusters.
Download and install mongosync.

To learn more, see the mongosync setup instructions.

Steps

Connect the replica set and sharded cluster

mongosync must create an initial connection between the source replica set and the destination sharded cluster before you can start to sync data.

To connect cluster0 to cluster1 with mongosync, enter the following command:

mongosync \
      --cluster0 "<cluster0_connection_string>" \
      --cluster1 "<cluster1_connection_string>"

Follow the connection instructions for your cluster architecture to format your connection strings and connect to the mongos instance in your cluster.

Start synchronization

Call the start endpoint to initiate data synchronization.

To sync from a replica set to a sharded cluster, set the sharding option for the start command to shard collections on the destination cluster. For more information, see Sharding Parameters.

Use the sharding.shardingEntries parameter to specify the collections to shard. Collections that you do not list in this array replicate as unsharded. For more information, see Shard Replica Sets and Choose a Shard Key.

The following example starts a sync from a replica set to a sharded cluster:

Request

curl localhost:27182/api/v1/start -XPOST \
--data '
   {
      "source": "cluster0",
      "destination": "cluster1",
      "sharding": {
         "createSupportingIndexes": true,
         "shardingEntries": [
            {
                "database": "accounts",
                "collection": "us_east",
                "shardCollection": {
                   "key": [
                      { "location": 1 },
                      { "region": 1 }
                   ]
                }
            }
         ]
      }
   } '

Response

{"success":true}

Next Steps

You can finalize a migration and transfer your application workload from the source to the destination cluster using the mongosync cutover process.

Limitations

You cannot use the reverse endpoint between a replica set and a sharded cluster.

mongosync allows users to rename collections that the sharding.shardingEntries option includes during sync with some limitations. For details, see Renaming During Sync.
If you use the sharding.createSupportingIndexes option, the indexes are automatically created on the destination cluster during the sync. You cannot create these indexes afterwards on the source cluster.
If you want to create an index to support shard keys manually, you must create the index before mongosync starts or after the migration is complete and mongosync has stopped.

If the source replica set has arbiters, the source replica set must have more than 2 non-arbiter nodes and you must sync from a non-arbiter node.

For more details, see Limitations.

Learn More

Back

Cluster Topologies

Two Sharded Clusters