Docs Menu
Docs Home
/
MongoDB Manual
/ / /

Restart a Self-Managed Sharded Cluster

On this page

  • Disable the Balancer
  • Stop Sharded Cluster
  • Start Sharded Cluster
  • Re-Enable the Balancer
  • Validate Cluster Accessibility

The tutorial is specific to MongoDB 6.0. For earlier versions of MongoDB, refer to the corresponding version of the MongoDB Manual.

This procedure demonstrates the shutdown and startup sequence for restarting a sharded cluster. Stopping or starting the components of a sharded cluster in a different order may cause communication errors between members. For example, shard servers may appear to hang if there are no config servers available.

Important

This procedure should only be performed during a planned maintenance period. During this period, applications should stop all reads and writes to the cluster in order to prevent potential data loss or reading stale data.

Disable the balancer to stop chunk migration and do not perform any metadata write operations until the process finishes. If a migration is in progress, the balancer will complete the in-progress migration before stopping.

To disable the balancer, connect to one of the cluster's mongos instances and issue the following command: [1]

sh.stopBalancer()

To check the balancer state, issue the sh.getBalancerState() command.

For more information, see Disable the Balancer.

[1] Starting in MongoDB 6.0.3, automatic chunk splitting is not performed. This is because of balancing policy improvements. Auto-splitting commands still exist, but do not perform an operation. For details, see Balancing Policy Changes.In MongoDB versions earlier than 6.0.3, sh.stopBalancer() also disables auto-splitting for the sharded cluster.
1

Run db.shutdownServer() from the admin database on each mongos router:

use admin
db.shutdownServer()
2

Run db.shutdownServer() from the admin database on each shard replica set member to shutdown its mongod process. Shutdown all secondary members before shutting down the primary in each replica set.

3

Run db.shutdownServer() from the admin database on each of the config servers to shutdown its mongod process. Shutdown all secondary members before shutting down the primary.

1

When starting each mongod, specify the mongod settings using either a configuration file or the command line. For more information on startup parameters, see the mongod reference page.

Configuration File

If using a configuration file, start the mongod with the --config option set to the configuration file path.

mongod --config <path-to-config-file>

Command Line

If using the command line options, start the mongod with the --configsvr, --replSet, --bind_ip, and other options as appropriate to your deployment. For example:

mongod --configsvr --replSet <replica set name> --dbpath <path> --bind_ip localhost,<hostname(s)|ip address(es)>

After starting all config servers, connect to the primary mongod and run rs.status() to confirm the health and availability of each CSRS member.

2

When starting each mongod, specify the mongod settings using either a configuration file or the command line.

Configuration File

If using a configuration file, start the mongod with the --config option set to the configuration file path.

mongod --config <path-to-config-file>

Command Line

If using the command line option, start the mongod with the --replSet, --shardsvr, and --bind_ip options, and other options as appropriate to your deployment. For example:

mongod --shardsvr --replSet <replSetname> --dbpath <path> --bind_ip localhost,<hostname(s)|ip address(es)>

After starting all members of each shard, connect to each primary mongod and run rs.status() to confirm the health and availability of each member.

3

Start mongos routers using either a configuration file or a command line parameter to specify the config servers.

Configuration File

If using a configuration file, start the mongos specifying the --config option and the path to the configuration file.

mongos --config <path-to-config>

For more information on the configuration file, see configuration options.

Command Line

If using command line parameters, start the mongos and specify the --configdb, --bind_ip, and other options as appropriate to your deployment. For example:

Warning

Before binding to a non-localhost (e.g. publicly accessible) IP address, ensure you have secured your cluster from unauthorized access. For a complete list of security recommendations, see Security Checklist for Self-Managed Deployments. At minimum, consider enabling authentication and hardening network infrastructure.

mongos --configdb <configReplSetName>/cfg1.example.net:27019,cfg2.example.net:27019 --bind_ip localhost,<hostname(s)|ip address(es)>

Include any other options as appropriate for your deployment.

Re-enable the balancer to resume chunk migrations.

Connect to one of the cluster's mongos instances and run the sh.startBalancer() command: [2]

sh.startBalancer()

To check the balancer state, issue the sh.getBalancerState() command.

For more information, see Enable the Balancer.

[2] Starting in MongoDB 6.0.3, automatic chunk splitting is not performed. This is because of balancing policy improvements. Auto-splitting commands still exist, but do not perform an operation. For details, see Balancing Policy Changes.In MongoDB versions earlier than 6.0.3, sh.startBalancer() also enables auto-splitting for the sharded cluster.

Connect a mongo shell to one of the cluster's mongos processes. Use sh.status() to check the overall cluster status.

To confirm that all shards are accessible and communicating, insert test data into a temporary sharded collection. Confirm that data is being split and migrated between each shard in your cluster. You can connect a mongo shell to each shard primary and use db.collection.find() to validate that the data was sharded as expected.

Important

To prevent potential data loss or reading stale data, do not start application reads and writes to the cluster until after confirming the cluster is healthy and accessible.

Back

Replace a Self-Managed Config Server