ADD LEAF

Add a leaf node to a cluster.

Syntax

ADD LEAF user[:'password']@'host':port [INTO GROUP {1|2}]

Remarks

Warning

Addresses used when running ADD LEAF should be stable addresses for optimal function and to avoid a split cluster situation.

If a leaf has a different IP when it is started from the one passed to ADD LEAF, then the master aggregator will not know how to connect to it. The recommended course of action in such a situation is to scrap the leaf node, configure it anew and add it to the cluster. Be careful to use a stable address.

It is also recommended to reconfigure the master aggregator with a stable address during BOOTSTRAP AGGREGATOR, otherwise, you might hit the same issue with the master.

If you need to change the address of a node, you need to remove it from the cluster and re-add it. Make sure to always have an available master aggregator, otherwise, you could end up losing the database.

  • user specifies the user other nodes in the cluster will use to connect to the leaf. This user should have super privileges as it will be used to run all queries and cluster management commands on the leaf.

  • password (default "") is optional. The password needed to log in to the leaf. This password can’t be changed once the leaf is added without first removing the leaf.

  • port value is 3306.

  • [INTO GROUP {1|2}] specifies the availability group, if applicable, to which you want to add the leaf. See Managing High Availability for more information. If you do not explicitly specify a group, SingleStore will automatically add the leaf into the smallest group less than or equal to redundancy_level.

  • If you run ADD LEAF on a host/port pair that already exists in SHOW LEAVES, SingleStore will return an error.

  • This command can be run on the master aggregator node, or a child aggregator node (see Cluster Management Commands ).

  • ADD LEAF doesn’t automatically run REBALANCE PARTITIONS to move new data onto the leaf. REBALANCE needs to be run manually to move data. If you plan to add multiple leaves its most efficient to add them all first and run a single REBALANCE to move data onto the new leaves.

  • If the leaf was previously a part of the cluster any partitions still on the leaf when its added will be reattached as part of ADD LEAF.

  • This command causes implicit commits. Refer to COMMIT for more information.

  • Refer to the Permission Matrix for the required permission.

Example

ADD LEAF 'root'@'192.168.1.110':3306;
Query OK, 1 row affected (0.00 sec)

Related Topics

Last modified: July 31, 2023

Was this article helpful?