Three proposed MySQL Shell and Router enhancements could simplify cluster automation, reduce connection disruptions during router maintenance, and support scaling reads across multiple replica levels. Explore a Python module for MySQL Shell AdminAPI, router connection draining, and cascading read replicas, and contribute feedback to the respective GitHub discussions:
https://github.com/mysql/mysql-community/issues/149
https://github.com/mysql/mysql-community/issues/148
https://github.com/mysql/mysql-community/issues/147
Explore three proposed enhancements to MySQL Shell and MySQL Router designed to simplify database topology management, support less disruptive router maintenance, and extend read scalability in MySQL InnoDB Cluster and InnoDB ClusterSet environments. In this MySQL Summit session, a MySQL developer presents proposals for a standalone Python module for MySQL Shell AdminAPI, connection draining in MySQL Router, and cascading read replica support.
The first proposal focuses on making MySQL Shell AdminAPI accessible directly from Python applications. AdminAPI provides operations for deploying, configuring, and managing MySQL topologies, including InnoDB Cluster, InnoDB ReplicaSet, and InnoDB ClusterSet. Currently, external Python applications must invoke MySQL Shell as a separate process to access these operations. The proposed importable Python module would allow applications to call selected AdminAPI operations within the same Python process, potentially simplifying integration with Kubernetes operators and other database management tools.
The presentation examines technical considerations for this Python module, including shared C++ implementation code, Python runtime compatibility, dependency management, packaging, concurrency, and automated testing. Community feedback is requested to help determine the initial API scope and supported environments.
The second proposal introduces connection draining for MySQL Router. During router upgrades or replacements, shutting down an active router can interrupt existing database connections. The proposed draining capability would stop a router from accepting new connections while allowing established sessions to continue until they close. Administrators could initiate and monitor draining remotely through MySQL Shell and metadata-based state management.
The session also discusses how draining could interact with connection pools, monitoring systems, idle connections, and automatic router shutdown. These design considerations aim to support more controlled router maintenance and replacement workflows.
The third proposal addresses cascading read replicas in MySQL InnoDB Cluster and InnoDB ClusterSet. Rather than requiring every read replica to connect directly to cluster members, cascading replication would allow intermediate replicas to serve as replication sources for additional downstream replicas. This approach could help distribute replication traffic and extend read capacity.
The presentation examines the tradeoffs of multilevel replication, including increased replication lag, intermediate replica failures, automatic connection failover, and replica health monitoring. It also explores how MySQL Shell could manage cascading replica topologies and how MySQL Router policies could select appropriate read targets based on replication health and lag.
The session concludes by inviting community feedback on all three proposals, particularly from developers, database administrators, and maintainers of existing MySQL management tools.
These capabilities are proposed enhancements under discussion and should not be interpreted as generally available MySQL functionality.
https://github.com/mysql/mysql-community/issues/149
https://github.com/mysql/mysql-community/issues/148
https://github.com/mysql/mysql-community/issues/147
Explore three proposed enhancements to MySQL Shell and MySQL Router designed to simplify database topology management, support less disruptive router maintenance, and extend read scalability in MySQL InnoDB Cluster and InnoDB ClusterSet environments. In this MySQL Summit session, a MySQL developer presents proposals for a standalone Python module for MySQL Shell AdminAPI, connection draining in MySQL Router, and cascading read replica support.
The first proposal focuses on making MySQL Shell AdminAPI accessible directly from Python applications. AdminAPI provides operations for deploying, configuring, and managing MySQL topologies, including InnoDB Cluster, InnoDB ReplicaSet, and InnoDB ClusterSet. Currently, external Python applications must invoke MySQL Shell as a separate process to access these operations. The proposed importable Python module would allow applications to call selected AdminAPI operations within the same Python process, potentially simplifying integration with Kubernetes operators and other database management tools.
The presentation examines technical considerations for this Python module, including shared C++ implementation code, Python runtime compatibility, dependency management, packaging, concurrency, and automated testing. Community feedback is requested to help determine the initial API scope and supported environments.
The second proposal introduces connection draining for MySQL Router. During router upgrades or replacements, shutting down an active router can interrupt existing database connections. The proposed draining capability would stop a router from accepting new connections while allowing established sessions to continue until they close. Administrators could initiate and monitor draining remotely through MySQL Shell and metadata-based state management.
The session also discusses how draining could interact with connection pools, monitoring systems, idle connections, and automatic router shutdown. These design considerations aim to support more controlled router maintenance and replacement workflows.
The third proposal addresses cascading read replicas in MySQL InnoDB Cluster and InnoDB ClusterSet. Rather than requiring every read replica to connect directly to cluster members, cascading replication would allow intermediate replicas to serve as replication sources for additional downstream replicas. This approach could help distribute replication traffic and extend read capacity.
The presentation examines the tradeoffs of multilevel replication, including increased replication lag, intermediate replica failures, automatic connection failover, and replica health monitoring. It also explores how MySQL Shell could manage cascading replica topologies and how MySQL Router policies could select appropriate read targets based on replication health and lag.
The session concludes by inviting community feedback on all three proposals, particularly from developers, database administrators, and maintainers of existing MySQL management tools.
These capabilities are proposed enhancements under discussion and should not be interpreted as generally available MySQL functionality.