MySQL Router is part of InnoDB Cluster and is lightweight middleware that provides transparent routing between your application and back-end MySQL Servers. It is used for a wide variety of use cases, such as providing high availability and scalability by routing database traffic to appropriate back-end MySQL servers. The pluggable architecture also enables developers to extend MySQL Router for custom use cases.
For additional details about how Router is part of InnoDB Cluster, see MySQL AdminAPI.
For client applications to handle failover, they need to be aware of the InnoDB cluster topology and know which MySQL instance is the PRIMARY. While it is possible for applications to implement that logic, MySQL Router can provide and handle this functionality for you.
MySQL uses Group Replication to replicate databases across multiple servers while performing automatic failover in the event of a server failure. When used with a MySQL InnoDB Cluster, MySQL Router acts as a proxy to hide the multiple MySQL instances on your network and map the data requests to one of the cluster instances. As long as there are enough online replicas and communication between the components is intact, applications will be able to contact one of them. MySQL Router also makes this possible by having applications connect to MySQL Router instead of directly to MySQL.
The recommended deployment model for MySQL Router is with InnoDB Cluster, with Router sitting on the same host as the application.
The steps for deploying MySQL Router with an InnoDB Cluster after configuring the cluster are:
Install MySQL Router.
Bootstrap InnoDB Cluster, and test.
Bootstrapping automatically configures MySQL Router for an existing InnoDB Cluster by using
--bootstrapand other command-line options. During bootstrap, Router connects to the cluster, fetches its metadata, and configures itself for use. Bootstrapping is optional.
For additional information, see Chapter 3, Deploying MySQL Router.
Set up MySQL Router for automatic startup.
Configure your system to automatically start MySQL Router when the host is rebooted, a process similar to how the MySQL server is configured to start automatically. For additional details, see Section 5.1, “Starting MySQL Router”.
For example, after creating a MySQL InnoDB Cluster, you might configure MySQL Router using:
$> mysqlrouter --bootstrap localhost:3310 --directory /opt/myrouter --user snoopy
This example bootstraps MySQL Router to an existing InnoDB Cluster where:
localhost:3310is a member of an InnoDB cluster, and either the PRIMARY or bootstrap will redirect to a PRIMARY in the cluster.
Because the optional
--directorybootstrap option was used, this example creates a self-contained installation with all generated directories and files at
/opt/myrouter/. These files include
log/, and a fully functional MySQL Router configuration file named
Only the host's system user named
snoopywill have access to
--bootstrap and related
options for ways to modify the bootstrap configuration process.
For example, passing in
Unix domain socket connections because only TCP/IP connections are
enabled by default.