Python: Add Multi-Database Support for Cluster Mode Valkey 9.0 #4659

affonsov · 2025-08-27T05:18:01Z

Add Multi-Database Support for Cluster Mode (Valkey 9.0+)

Overview

This PR adds comprehensive support for multi-database functionality in cluster mode, leveraging the new cluster-databases configuration introduced in Valkey 9.0+. Previously, cluster mode was limited to database 0 only, but Valkey 9.0+ allows configuring multiple databases in cluster mode when cluster-databases is set to a value greater than 1.

Key Features

🔧 Core Implementation

Rust Core: Added automatic database selection during cluster client creation
Python Client: Added database_id parameter to cluster client configurations

🎯 Client Configuration

Before (Cluster mode limited to database 0):

client = await GlideClusterClient.create_client(
    GlideClusterClientConfiguration(
        addresses=[NodeAddress("localhost", 6379)]
        # database_id not supported in cluster mode
    )
)

After (Cluster mode supports multiple databases):

client = await GlideClusterClient.create_client(
    GlideClusterClientConfiguration(
        addresses=[NodeAddress("localhost", 6379)],
        database_id=5  # Now supported in Valkey 9.0+ cluster mode
    )
)

Technical Changes

Rust Core (`glide-core`)

Database Selection Logic: Added automatic database selection during cluster client creation
Error Handling: Proper error propagation when database selection fails
Logging: Added detailed logging for database selection operations

Python Client

Configuration: Extended both async and sync cluster configurations with database_id parameter
Validation: Added client-side validation for database_id (must be non-negative integer)
Documentation: Added comprehensive warnings about reconnection behavior

Test Coverage

Integration Tests: Comprehensive test suite for multi-database functionality

Server Compatibility

Valkey 9.0+ Requirements

Configuration: Requires cluster-databases > 1 in server configuration
Default Behavior: Maintains backward compatibility (defaults to database 0)
Error Handling: Graceful handling when multi-database cluster mode is not supported

Test Environment Setup

Cluster Manager: Updated to automatically configure cluster-databases 16 for Valkey 9.0+
Conditional Testing: Tests skip gracefully on older servers or misconfigured clusters

Issue link

This Pull Request is linked to issue (URL): [#4500]

Checklist

Before submitting the PR make sure the following are checked:

This Pull Request is related to one issue.
Commit message has a detailed description of what changed and why.
Tests are added or updated.
CHANGELOG.md and documentation files are updated.
Destination branch is correct - main or release
Create merge commit if merging release branch into main, squash otherwise.

leszek-bq

Looks good to me.

shohamazon

Few points:

I dont understand the need for a new file of test, we should have a single simple test like standalone.
Cant we check for server version in the test instead of printing that failing is expected?

glide-core/src/client/mod.rs

glide-core/tests/test_cluster_client.rs

python/glide-async/python/glide/async_commands/cluster_commands.py

python/glide-async/python/glide/async_commands/standalone_commands.py

python/glide-sync/glide_sync/sync_commands/cluster_commands.py

Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com>

python/glide-shared/glide_shared/config.py

Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com>

yipin-chen

LGTM

shohamazon

Few comment, great work!

glide-core/redis-rs/redis/src/cluster_client.rs

glide-core/tests/test_cluster_client.rs

python/glide-shared/glide_shared/config.py

liorsve · 2025-09-07T11:23:32Z

Please note that we removed the select command for now from the sync client - #4684

refactored extract_client_id to be in utilities Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com>

added custom command test using select Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com>

* - Implement database selection for cluster clients when database_id != 0 - Send SELECT command to all cluster nodes using MultiNode routing - Add comprehensive error handling with clear error messages - Include test for database selection error handling in cluster mode - Support backward compatibility with servers that don't support multi-db cluster mode - Enable cluster-databases=16 for Valkey 9.0+ in cluster manager This enables cluster clients to work with non-default databases on Valkey 9.0+ servers configured with cluster-databases support, while gracefully handling older servers that don't support this feature. Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * update valkey9 multi db tests Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fix lint Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fixing valkey 9 cluster tests Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * add to route to all nodes in the core Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * rust lint fix Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * implement database selection in cluster mode via connection parameters - Add database_id field to ClusterParams and BuilderParams structs - Add database() method to ClusterClientBuilder for setting database ID - Pass database_id through connection info instead of post-connection SELECT - Remove SELECT command from cluster routing - Remove post-connection SELECT command logic from create_cluster_client - Add comprehensive tests for database isolation and reconnection behavior - Verify database ID persistence across node reconnections - Test multi-database cluster support with proper isolation verification This change moves database selection from a post-connection SELECT command to a connection parameter, which is more reliable for cluster mode and handles reconnections automatically. The approach works with servers that support multi-database cluster configurations (like Valkey 9.0+). Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * add valkey9 constraint in the core tests Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fix core test not skipping test when version was lower than valkey 9 Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * Fix version check Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * refactored test test_set_database_id_after_reconnection to be similar from standalone removed is_valkey_9_or_higher Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * removed tests and cleanup Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * renamed builder.database to builder.database_id Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * python: Add multi-database support for cluster and standalone modes - Add database_id parameter to BaseClientConfiguration, GlideClientConfiguration, and GlideClusterClientConfiguration - Implement SELECT command for both cluster and standalone modes with comprehensive documentation - Add validation for database_id parameter (0-15 range, integer type) - Refactor configuration protobuf request creation with helper methods for better maintainability - Add extensive test coverage for database_id configuration and validation - Include production warnings and recommended approaches in SELECT command documentation - Support routing configuration for cluster mode SELECT operations - Ensure database_id persists across reconnections when set in client configuration Documentation: - Added comprehensive docstrings with warnings about SELECT command limitations - Included examples showing recommended database_id configuration approach - Documented reconnection behavior and cluster-specific routing requirements Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fixing database_id restriction Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fixed test test_standalone_client_with_very_high_database_ids Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fix python tests with higher databases id Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fix lint error Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * refactoring out the route option from the select command Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * Remove SELECT command from cluster mode and enable cluster tests - Remove select() method from ClusterCommands in both async and sync clients - Enable cluster mode testing for database_id tests (parametrize True, False) - Delete database_id integration test file The SELECT command is not recommended for cluster mode due to reconnection behavior where nodes revert to configured database_id, not the selected one. Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * removed database_id validation| renamed tests to be cluster/standalone agnostic clean up comments Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * added changelog Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * lint fix Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fix test database_id should be skipped for versions lower than 9.0.0 Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fix test Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * added select rounting refactored extract_client_id to be in utilities Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fix database_id documentation added custom command test using select Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> --------- Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com>

… Valkey 9.0 (#4694) Python: Add Multi-Database Support for Cluster Mode Valkey 9.0 (#4659) * - Implement database selection for cluster clients when database_id != 0 - Send SELECT command to all cluster nodes using MultiNode routing - Add comprehensive error handling with clear error messages - Include test for database selection error handling in cluster mode - Support backward compatibility with servers that don't support multi-db cluster mode - Enable cluster-databases=16 for Valkey 9.0+ in cluster manager This enables cluster clients to work with non-default databases on Valkey 9.0+ servers configured with cluster-databases support, while gracefully handling older servers that don't support this feature. * update valkey9 multi db tests * fix lint * fixing valkey 9 cluster tests * add to route to all nodes in the core * rust lint fix * implement database selection in cluster mode via connection parameters - Add database_id field to ClusterParams and BuilderParams structs - Add database() method to ClusterClientBuilder for setting database ID - Pass database_id through connection info instead of post-connection SELECT - Remove SELECT command from cluster routing - Remove post-connection SELECT command logic from create_cluster_client - Add comprehensive tests for database isolation and reconnection behavior - Verify database ID persistence across node reconnections - Test multi-database cluster support with proper isolation verification This change moves database selection from a post-connection SELECT command to a connection parameter, which is more reliable for cluster mode and handles reconnections automatically. The approach works with servers that support multi-database cluster configurations (like Valkey 9.0+). * add valkey9 constraint in the core tests * fix core test not skipping test when version was lower than valkey 9 * Fix version check * refactored test test_set_database_id_after_reconnection to be similar from standalone removed is_valkey_9_or_higher * removed tests and cleanup * renamed builder.database to builder.database_id * python: Add multi-database support for cluster and standalone modes - Add database_id parameter to BaseClientConfiguration, GlideClientConfiguration, and GlideClusterClientConfiguration - Implement SELECT command for both cluster and standalone modes with comprehensive documentation - Add validation for database_id parameter (0-15 range, integer type) - Refactor configuration protobuf request creation with helper methods for better maintainability - Add extensive test coverage for database_id configuration and validation - Include production warnings and recommended approaches in SELECT command documentation - Support routing configuration for cluster mode SELECT operations - Ensure database_id persists across reconnections when set in client configuration Documentation: - Added comprehensive docstrings with warnings about SELECT command limitations - Included examples showing recommended database_id configuration approach - Documented reconnection behavior and cluster-specific routing requirements * fixing database_id restriction * fixed test test_standalone_client_with_very_high_database_ids * fix python tests with higher databases id * fix lint error * refactoring out the route option from the select command * Remove SELECT command from cluster mode and enable cluster tests - Remove select() method from ClusterCommands in both async and sync clients - Enable cluster mode testing for database_id tests (parametrize True, False) - Delete database_id integration test file The SELECT command is not recommended for cluster mode due to reconnection behavior where nodes revert to configured database_id, not the selected one. * removed database_id validation| renamed tests to be cluster/standalone agnostic clean up comments * added changelog * lint fix * fix test database_id should be skipped for versions lower than 9.0.0 * fix test * added select rounting refactored extract_client_id to be in utilities * fix database_id documentation added custom command test using select --------- Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com>

* Go: Add MUSL support (#4476) * Go: Switch to MUSL binary Signed-off-by: James Xin <james.xin@improving.com> Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * rustup add musl targets Signed-off-by: James Xin <james.xin@improving.com> Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * lint Signed-off-by: James Xin <james.xin@improving.com> Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * rustup add musl targets in Makefile Signed-off-by: James Xin <james.xin@improving.com> Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Add pathing Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Trigger CodeQL on main Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Fix typo Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Fix typo Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Modify CD Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Modify CD Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Modify CD Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Modify CD Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update docs Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Format Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Fix Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Fix Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Address feedback Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update doc Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> --------- Signed-off-by: James Xin <james.xin@improving.com> Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> Co-authored-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> Co-authored-by: Alexey Temnikov <alexey.temnikov@improving.com> * Java: Migration guide for Jedis compatibility layer (#4672) (#4681) * Java: Migration guide for Jedis compatibility layer Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * [Backport to 2.1] Python: Add Multi-Database Support for Cluster Mode Valkey 9.0 (#4694) Python: Add Multi-Database Support for Cluster Mode Valkey 9.0 (#4659) * - Implement database selection for cluster clients when database_id != 0 - Send SELECT command to all cluster nodes using MultiNode routing - Add comprehensive error handling with clear error messages - Include test for database selection error handling in cluster mode - Support backward compatibility with servers that don't support multi-db cluster mode - Enable cluster-databases=16 for Valkey 9.0+ in cluster manager This enables cluster clients to work with non-default databases on Valkey 9.0+ servers configured with cluster-databases support, while gracefully handling older servers that don't support this feature. * update valkey9 multi db tests * fix lint * fixing valkey 9 cluster tests * add to route to all nodes in the core * rust lint fix * implement database selection in cluster mode via connection parameters - Add database_id field to ClusterParams and BuilderParams structs - Add database() method to ClusterClientBuilder for setting database ID - Pass database_id through connection info instead of post-connection SELECT - Remove SELECT command from cluster routing - Remove post-connection SELECT command logic from create_cluster_client - Add comprehensive tests for database isolation and reconnection behavior - Verify database ID persistence across node reconnections - Test multi-database cluster support with proper isolation verification This change moves database selection from a post-connection SELECT command to a connection parameter, which is more reliable for cluster mode and handles reconnections automatically. The approach works with servers that support multi-database cluster configurations (like Valkey 9.0+). * add valkey9 constraint in the core tests * fix core test not skipping test when version was lower than valkey 9 * Fix version check * refactored test test_set_database_id_after_reconnection to be similar from standalone removed is_valkey_9_or_higher * removed tests and cleanup * renamed builder.database to builder.database_id * python: Add multi-database support for cluster and standalone modes - Add database_id parameter to BaseClientConfiguration, GlideClientConfiguration, and GlideClusterClientConfiguration - Implement SELECT command for both cluster and standalone modes with comprehensive documentation - Add validation for database_id parameter (0-15 range, integer type) - Refactor configuration protobuf request creation with helper methods for better maintainability - Add extensive test coverage for database_id configuration and validation - Include production warnings and recommended approaches in SELECT command documentation - Support routing configuration for cluster mode SELECT operations - Ensure database_id persists across reconnections when set in client configuration Documentation: - Added comprehensive docstrings with warnings about SELECT command limitations - Included examples showing recommended database_id configuration approach - Documented reconnection behavior and cluster-specific routing requirements * fixing database_id restriction * fixed test test_standalone_client_with_very_high_database_ids * fix python tests with higher databases id * fix lint error * refactoring out the route option from the select command * Remove SELECT command from cluster mode and enable cluster tests - Remove select() method from ClusterCommands in both async and sync clients - Enable cluster mode testing for database_id tests (parametrize True, False) - Delete database_id integration test file The SELECT command is not recommended for cluster mode due to reconnection behavior where nodes revert to configured database_id, not the selected one. * removed database_id validation| renamed tests to be cluster/standalone agnostic clean up comments * added changelog * lint fix * fix test database_id should be skipped for versions lower than 9.0.0 * fix test * added select rounting refactored extract_client_id to be in utilities * fix database_id documentation added custom command test using select --------- Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * [Backport 2.1] Go: Add Multi-Database Support for Cluster Mode Valkey 9.0 (#4696) Go: Add Multi-Database Support for Cluster Mode Valkey 9.0 (#4660) * - Implement database selection for cluster clients when database_id != 0 - Send SELECT command to all cluster nodes using MultiNode routing - Add comprehensive error handling with clear error messages - Include test for database selection error handling in cluster mode - Support backward compatibility with servers that don't support multi-db cluster mode - Enable cluster-databases=16 for Valkey 9.0+ in cluster manager This enables cluster clients to work with non-default databases on Valkey 9.0+ servers configured with cluster-databases support, while gracefully handling older servers that don't support this feature. * update valkey9 multi db tests * fix lint * fixing valkey 9 cluster tests * add to route to all nodes in the core * rust lint fix * implement database selection in cluster mode via connection parameters - Add database_id field to ClusterParams and BuilderParams structs - Add database() method to ClusterClientBuilder for setting database ID - Pass database_id through connection info instead of post-connection SELECT - Remove SELECT command from cluster routing - Remove post-connection SELECT command logic from create_cluster_client - Add comprehensive tests for database isolation and reconnection behavior - Verify database ID persistence across node reconnections - Test multi-database cluster support with proper isolation verification This change moves database selection from a post-connection SELECT command to a connection parameter, which is more reliable for cluster mode and handles reconnections automatically. The approach works with servers that support multi-database cluster configurations (like Valkey 9.0+). * add valkey9 constraint in the core tests * fix core test not skipping test when version was lower than valkey 9 * Fix version check * refactored test test_set_database_id_after_reconnection to be similar from standalone removed is_valkey_9_or_higher * removed tests and cleanup * renamed builder.database to builder.database_id * go: Add multi-database support for cluster mode clients - Add DatabaseId field to baseClientConfiguration for both standalone and cluster clients - Implement WithDatabaseId() method for ClusterClientConfiguration - Add database ID validation with proper error handling for negative values - Refactor standalone client to use shared database ID logic from base configuration - Add Select() and SelectWithOptions() methods for cluster clients with routing support - Include comprehensive test coverage for database isolation, reconnection persistence, and error handling - Add backward compatibility support - clients default to database 0 when no database_id specified - Add server version compatibility checks (cluster multi-database requires Valkey 9.0+) - Update documentation with warnings about SELECT command limitations and recommended configuration approach This enables cluster mode clients to connect to specific databases at initialization time, with the database selection persisting across reconnections, unlike the SELECT command which resets on reconnection. * fixed go tests - batch tests - database id tests * removed selectWithOptions * go: Remove Select command from cluster clients - Remove Select method from ClusterClient and related interfaces - Delete comprehensive database_id integration tests - Remove Select command from batch operations - Remove Select examples and tests - Add example for cluster client with database_id configuration The Select command is being removed in favor of using the database_id configuration parameter, which provides persistent database selection across reconnections. * remove config validation simplified tests * added changelog * removing batch test, and readding standalone select and removing batch select --------- Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * [Backport 2.1] Java: Multi-Database Support for Cluster Mode Valkey 9.0 (#4697) Java: Multi-Database Support for Cluster Mode Valkey 9.0 (#4658) * - Implement database selection for cluster clients when database_id != 0 - Send SELECT command to all cluster nodes using MultiNode routing - Add comprehensive error handling with clear error messages - Include test for database selection error handling in cluster mode - Support backward compatibility with servers that don't support multi-db cluster mode - Enable cluster-databases=16 for Valkey 9.0+ in cluster manager This enables cluster clients to work with non-default databases on Valkey 9.0+ servers configured with cluster-databases support, while gracefully handling older servers that don't support this feature. * update valkey9 multi db tests * fix lint * fixing valkey 9 cluster tests * add to route to all nodes in the core * rust lint fix * implement database selection in cluster mode via connection parameters - Add database_id field to ClusterParams and BuilderParams structs - Add database() method to ClusterClientBuilder for setting database ID - Pass database_id through connection info instead of post-connection SELECT - Remove SELECT command from cluster routing - Remove post-connection SELECT command logic from create_cluster_client - Add comprehensive tests for database isolation and reconnection behavior - Verify database ID persistence across node reconnections - Test multi-database cluster support with proper isolation verification This change moves database selection from a post-connection SELECT command to a connection parameter, which is more reliable for cluster mode and handles reconnections automatically. The approach works with servers that support multi-database cluster configurations (like Valkey 9.0+). * add valkey9 constraint in the core tests * fix core test not skipping test when version was lower than valkey 9 * Fix version check * refactored test test_set_database_id_after_reconnection to be similar from standalone removed is_valkey_9_or_higher * removed tests and cleanup * renamed builder.database to builder.database_id * Java: implement multi-database support for cluster mode - Move databaseId from GlideClientConfiguration to BaseClientConfiguration - Add databaseId validation and protobuf integration - Update SELECT command with AllNodes routing and reconnection warnings - Add connection management for database selection after auth - Add comprehensive tests with Valkey 9+ version guards Completes Java client implementation for Valkey 9 cluster multi-database support. * fixing database id restriction * fixed java test error select_command_invalid_database_standalone * fixed lint errors * Refactored select command to route to all nodes per default Removed the option to provide a route * java lint fix * java: Remove SELECT command from cluster client - Remove select() method from GlideClusterClient and ConnectionManagementClusterCommands interface - Delete MultiDatabaseTests.java which tested SELECT command functionality - Remove SELECT command tests from SharedCommandTests.java - Add focused test for cluster database_id configuration in ClusterClientTests - Remove SELECT import from GlideClusterClient The SELECT command is being removed from the cluster client API as database selection should be handled through client configuration (databaseId parameter) rather than runtime commands, which provides better consistency and avoids reconnection issues. * remove database_id validatation and simplified tests * added changelog * fix documentation * fix databaseId documentation --------- Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * [Backport 2.1] Node: Add Multi-Database Support for Cluster Mode (Valkey 9.0) (#4698) Node: Add Multi-Database Support for Cluster Mode (Valkey 9.0) (#4657) * - Implement database selection for cluster clients when database_id != 0 - Send SELECT command to all cluster nodes using MultiNode routing - Add comprehensive error handling with clear error messages - Include test for database selection error handling in cluster mode - Support backward compatibility with servers that don't support multi-db cluster mode - Enable cluster-databases=16 for Valkey 9.0+ in cluster manager This enables cluster clients to work with non-default databases on Valkey 9.0+ servers configured with cluster-databases support, while gracefully handling older servers that don't support this feature. * update valkey9 multi db tests * fix lint * fixing valkey 9 cluster tests * add to route to all nodes in the core * rust lint fix * implement database selection in cluster mode via connection parameters - Add database_id field to ClusterParams and BuilderParams structs - Add database() method to ClusterClientBuilder for setting database ID - Pass database_id through connection info instead of post-connection SELECT - Remove SELECT command from cluster routing - Remove post-connection SELECT command logic from create_cluster_client - Add comprehensive tests for database isolation and reconnection behavior - Verify database ID persistence across node reconnections - Test multi-database cluster support with proper isolation verification This change moves database selection from a post-connection SELECT command to a connection parameter, which is more reliable for cluster mode and handles reconnections automatically. The approach works with servers that support multi-database cluster configurations (like Valkey 9.0+). * add valkey9 constraint in the core tests * fix core test not skipping test when version was lower than valkey 9 * Fix version check * refactored test test_set_database_id_after_reconnection to be similar from standalone removed is_valkey_9_or_higher * removed tests and cleanup * renamed builder.database to builder.database_id * node changes * move tests to sharedtests.ts * reafactoring to select to all nodes being in the core * node: Remove SELECT command support from cluster client - Remove select() method from GlideClusterClient - Remove createSelect import from cluster client - Remove extensive test coverage for SELECT functionality - Remove database ID validation tests from client internals - Add minimal database ID test for cluster client configuration - Clean up ConfigurationError import that's no longer needed This change removes the SELECT command implementation that was added for Valkey 9.0+ cluster support, likely due to reliability concerns with database switching in cluster mode or to simplify the API. * added changelog * fix test, and removed comment --------- Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * [Backport 2.1] Python sync: removed select command (#4705) Python sync: removed select command (#4684) removed sync select command Signed-off-by: Lior Sventitzky <liorsve@amazon.com> * [Backport 2.1] CI/CD: Added self hosted macOS ARM runners (#4706) * CI: Added self hosted macOS runners (#4683) added self hosted runners for macos Signed-off-by: Lior Sventitzky <liorsve@amazon.com> * CD: removed self hosted mac arm runners from required CD platforms #4700 (#4702) removed self hosted mac from cd, reverted npm-cd Signed-off-by: Lior Sventitzky <liorsve@amazon.com> --------- Signed-off-by: Lior Sventitzky <liorsve@amazon.com> * [backport 2.1] Python: fixed pypi-cd workflow, fixed release tests (#4708) Python: fixed pypi-cd workflow, fixed release tests (#4703) fixed pypi upload, fixed release tests Signed-off-by: Lior Sventitzky <liorsve@amazon.com> * Fix FFI tests for non-alpine distros (#4711) Signed-off-by: Jeremy Parr-Pearson <jeremy.parr-pearson@improving.com> * MUSL Java Build Fixes (#4712) * Use zigbuild for easier cross-compilation Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Fix Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Use zigbuild for easier cross-compilation Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Use zigbuild for easier cross-compilation Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Fix musl classifier / detection Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Fix musl classifier / detection Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Fix musl classifier / detection Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update docs Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Exclude arm musl test Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update arm runner Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update go exclusion Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Format Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> --------- Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Revert go upload step change (#4716) Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * CD: Fix Node CD delete artifact issue (#4717) Signed-off-by: James Xin <james.xin@improving.com> * CD: Fix Go musl CD issue (#4719) Signed-off-by: James Xin <james.xin@improving.com> * CD: Fix Node CD delete artifact issue 2nd try (#4718) * CD: Fix Node CD delete artifact issue 2nd try Signed-off-by: James Xin <james.xin@improving.com> * Linter Signed-off-by: James Xin <james.xin@improving.com> * remove temporary testing Signed-off-by: James Xin <james.xin@improving.com> --------- Signed-off-by: James Xin <james.xin@improving.com> * Python: Update ExpiryType enum references in hash field expiration docs (#4715) python: Update ExpiryType enum references in hash field expiration docs Replace deprecated ExpiryType enum values with new names in docstrings: - EX → SEC - PX → MILLSEC - EXAT → UNIX_SEC - PXAT → UNIX_MILLSEC - KEEPTTL → KEEP_TTL Updates documentation for hash field expiration commands in both async and sync clients. Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * [Backport 2.1] Python sync/CD: added description to pypi package of the sync client (#4720) Python sync/CD: added description to pypi package of the sync client (#4714) added description to sync client Signed-off-by: Lior Sventitzky <liorsve@amazon.com> * Java: Compatibility layer module (#4692) * Java: Compatibility layer module Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: fix integration tests Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Fix codeql for java code Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Fix codeql java build Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Fix codeql java build Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Fix codeql java build Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Fix codeql java build Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Fix codeql java build Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Fix codeql java build Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Fix codeql java build Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Update java cd Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: lint issues fixed Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Add signing configuration to jedis-compatibility build.gradle Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: update dependency Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Update valkey-glide-jedis-compatibility file path Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Fix build.gradle publishing code Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> --------- Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Fix: Copy GPG secring to jedis-compatibility in Java CD workflow (#4723) fix(java-cd): copy secring.gpg to jedis-compatibility for signing Signed-off-by: jbrinkman <joe.brinkman@improving.com> * [Backport 2.1] Python Sync: revert license format (#4732) Python Sync: revert license format (#4731) revert license format Signed-off-by: Lior Sventitzky <liorsve@amazon.com> * Java: Fix Jedis compatibility layer valkey-glide dependency (#4737) * Java: Fix Jedis compatibility layer valkey-glide dependency Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * [Backport 2.1] Python: improve UDS socket error handling (#4733) (#4755) Python: improve UDS socket error handling (#4733) fixed uds error handling Signed-off-by: Lior Sventitzky <liorsve@amazon.com> Co-authored-by: Lior Sventitzky <liorsve@amazon.com> --------- Signed-off-by: James Xin <james.xin@improving.com> Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> Signed-off-by: Lior Sventitzky <liorsve@amazon.com> Signed-off-by: Jeremy Parr-Pearson <jeremy.parr-pearson@improving.com> Signed-off-by: jbrinkman <joe.brinkman@improving.com> Co-authored-by: James Xin <james.xin@improving.com> Co-authored-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> Co-authored-by: Alexey Temnikov <alexey.temnikov@improving.com> Co-authored-by: prateek-kumar-improving <prateek.kumar@improving.com> Co-authored-by: affonsov <67347924+affonsov@users.noreply.github.com> Co-authored-by: Lior Sventitzky <liorsve@amazon.com> Co-authored-by: Jeremy Parr-Pearson <jeremy.parr-pearson@improving.com>

* - Implement database selection for cluster clients when database_id != 0 - Send SELECT command to all cluster nodes using MultiNode routing - Add comprehensive error handling with clear error messages - Include test for database selection error handling in cluster mode - Support backward compatibility with servers that don't support multi-db cluster mode - Enable cluster-databases=16 for Valkey 9.0+ in cluster manager This enables cluster clients to work with non-default databases on Valkey 9.0+ servers configured with cluster-databases support, while gracefully handling older servers that don't support this feature. Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * update valkey9 multi db tests Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fix lint Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fixing valkey 9 cluster tests Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * add to route to all nodes in the core Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * rust lint fix Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * implement database selection in cluster mode via connection parameters - Add database_id field to ClusterParams and BuilderParams structs - Add database() method to ClusterClientBuilder for setting database ID - Pass database_id through connection info instead of post-connection SELECT - Remove SELECT command from cluster routing - Remove post-connection SELECT command logic from create_cluster_client - Add comprehensive tests for database isolation and reconnection behavior - Verify database ID persistence across node reconnections - Test multi-database cluster support with proper isolation verification This change moves database selection from a post-connection SELECT command to a connection parameter, which is more reliable for cluster mode and handles reconnections automatically. The approach works with servers that support multi-database cluster configurations (like Valkey 9.0+). Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * add valkey9 constraint in the core tests Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fix core test not skipping test when version was lower than valkey 9 Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * Fix version check Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * refactored test test_set_database_id_after_reconnection to be similar from standalone removed is_valkey_9_or_higher Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * removed tests and cleanup Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * renamed builder.database to builder.database_id Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * python: Add multi-database support for cluster and standalone modes - Add database_id parameter to BaseClientConfiguration, GlideClientConfiguration, and GlideClusterClientConfiguration - Implement SELECT command for both cluster and standalone modes with comprehensive documentation - Add validation for database_id parameter (0-15 range, integer type) - Refactor configuration protobuf request creation with helper methods for better maintainability - Add extensive test coverage for database_id configuration and validation - Include production warnings and recommended approaches in SELECT command documentation - Support routing configuration for cluster mode SELECT operations - Ensure database_id persists across reconnections when set in client configuration Documentation: - Added comprehensive docstrings with warnings about SELECT command limitations - Included examples showing recommended database_id configuration approach - Documented reconnection behavior and cluster-specific routing requirements Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fixing database_id restriction Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fixed test test_standalone_client_with_very_high_database_ids Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fix python tests with higher databases id Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fix lint error Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * refactoring out the route option from the select command Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * Remove SELECT command from cluster mode and enable cluster tests - Remove select() method from ClusterCommands in both async and sync clients - Enable cluster mode testing for database_id tests (parametrize True, False) - Delete database_id integration test file The SELECT command is not recommended for cluster mode due to reconnection behavior where nodes revert to configured database_id, not the selected one. Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * removed database_id validation| renamed tests to be cluster/standalone agnostic clean up comments Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * added changelog Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * lint fix Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fix test database_id should be skipped for versions lower than 9.0.0 Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fix test Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * added select rounting refactored extract_client_id to be in utilities Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * fix database_id documentation added custom command test using select Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> --------- Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com>

* Go: Add MUSL support (#4476) * Go: Switch to MUSL binary Signed-off-by: James Xin <james.xin@improving.com> Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * rustup add musl targets Signed-off-by: James Xin <james.xin@improving.com> Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * lint Signed-off-by: James Xin <james.xin@improving.com> Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * rustup add musl targets in Makefile Signed-off-by: James Xin <james.xin@improving.com> Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Add pathing Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Trigger CodeQL on main Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update CI Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Fix typo Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Fix typo Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Modify CD Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Modify CD Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Modify CD Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Modify CD Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update docs Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Format Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Fix Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Fix Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Address feedback Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update doc Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> --------- Signed-off-by: James Xin <james.xin@improving.com> Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> Co-authored-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> Co-authored-by: Alexey Temnikov <alexey.temnikov@improving.com> * Java: Migration guide for Jedis compatibility layer (#4672) (#4681) * Java: Migration guide for Jedis compatibility layer Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * [Backport to 2.1] Python: Add Multi-Database Support for Cluster Mode Valkey 9.0 (#4694) Python: Add Multi-Database Support for Cluster Mode Valkey 9.0 (#4659) * - Implement database selection for cluster clients when database_id != 0 - Send SELECT command to all cluster nodes using MultiNode routing - Add comprehensive error handling with clear error messages - Include test for database selection error handling in cluster mode - Support backward compatibility with servers that don't support multi-db cluster mode - Enable cluster-databases=16 for Valkey 9.0+ in cluster manager This enables cluster clients to work with non-default databases on Valkey 9.0+ servers configured with cluster-databases support, while gracefully handling older servers that don't support this feature. * update valkey9 multi db tests * fix lint * fixing valkey 9 cluster tests * add to route to all nodes in the core * rust lint fix * implement database selection in cluster mode via connection parameters - Add database_id field to ClusterParams and BuilderParams structs - Add database() method to ClusterClientBuilder for setting database ID - Pass database_id through connection info instead of post-connection SELECT - Remove SELECT command from cluster routing - Remove post-connection SELECT command logic from create_cluster_client - Add comprehensive tests for database isolation and reconnection behavior - Verify database ID persistence across node reconnections - Test multi-database cluster support with proper isolation verification This change moves database selection from a post-connection SELECT command to a connection parameter, which is more reliable for cluster mode and handles reconnections automatically. The approach works with servers that support multi-database cluster configurations (like Valkey 9.0+). * add valkey9 constraint in the core tests * fix core test not skipping test when version was lower than valkey 9 * Fix version check * refactored test test_set_database_id_after_reconnection to be similar from standalone removed is_valkey_9_or_higher * removed tests and cleanup * renamed builder.database to builder.database_id * python: Add multi-database support for cluster and standalone modes - Add database_id parameter to BaseClientConfiguration, GlideClientConfiguration, and GlideClusterClientConfiguration - Implement SELECT command for both cluster and standalone modes with comprehensive documentation - Add validation for database_id parameter (0-15 range, integer type) - Refactor configuration protobuf request creation with helper methods for better maintainability - Add extensive test coverage for database_id configuration and validation - Include production warnings and recommended approaches in SELECT command documentation - Support routing configuration for cluster mode SELECT operations - Ensure database_id persists across reconnections when set in client configuration Documentation: - Added comprehensive docstrings with warnings about SELECT command limitations - Included examples showing recommended database_id configuration approach - Documented reconnection behavior and cluster-specific routing requirements * fixing database_id restriction * fixed test test_standalone_client_with_very_high_database_ids * fix python tests with higher databases id * fix lint error * refactoring out the route option from the select command * Remove SELECT command from cluster mode and enable cluster tests - Remove select() method from ClusterCommands in both async and sync clients - Enable cluster mode testing for database_id tests (parametrize True, False) - Delete database_id integration test file The SELECT command is not recommended for cluster mode due to reconnection behavior where nodes revert to configured database_id, not the selected one. * removed database_id validation| renamed tests to be cluster/standalone agnostic clean up comments * added changelog * lint fix * fix test database_id should be skipped for versions lower than 9.0.0 * fix test * added select rounting refactored extract_client_id to be in utilities * fix database_id documentation added custom command test using select --------- Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * [Backport 2.1] Go: Add Multi-Database Support for Cluster Mode Valkey 9.0 (#4696) Go: Add Multi-Database Support for Cluster Mode Valkey 9.0 (#4660) * - Implement database selection for cluster clients when database_id != 0 - Send SELECT command to all cluster nodes using MultiNode routing - Add comprehensive error handling with clear error messages - Include test for database selection error handling in cluster mode - Support backward compatibility with servers that don't support multi-db cluster mode - Enable cluster-databases=16 for Valkey 9.0+ in cluster manager This enables cluster clients to work with non-default databases on Valkey 9.0+ servers configured with cluster-databases support, while gracefully handling older servers that don't support this feature. * update valkey9 multi db tests * fix lint * fixing valkey 9 cluster tests * add to route to all nodes in the core * rust lint fix * implement database selection in cluster mode via connection parameters - Add database_id field to ClusterParams and BuilderParams structs - Add database() method to ClusterClientBuilder for setting database ID - Pass database_id through connection info instead of post-connection SELECT - Remove SELECT command from cluster routing - Remove post-connection SELECT command logic from create_cluster_client - Add comprehensive tests for database isolation and reconnection behavior - Verify database ID persistence across node reconnections - Test multi-database cluster support with proper isolation verification This change moves database selection from a post-connection SELECT command to a connection parameter, which is more reliable for cluster mode and handles reconnections automatically. The approach works with servers that support multi-database cluster configurations (like Valkey 9.0+). * add valkey9 constraint in the core tests * fix core test not skipping test when version was lower than valkey 9 * Fix version check * refactored test test_set_database_id_after_reconnection to be similar from standalone removed is_valkey_9_or_higher * removed tests and cleanup * renamed builder.database to builder.database_id * go: Add multi-database support for cluster mode clients - Add DatabaseId field to baseClientConfiguration for both standalone and cluster clients - Implement WithDatabaseId() method for ClusterClientConfiguration - Add database ID validation with proper error handling for negative values - Refactor standalone client to use shared database ID logic from base configuration - Add Select() and SelectWithOptions() methods for cluster clients with routing support - Include comprehensive test coverage for database isolation, reconnection persistence, and error handling - Add backward compatibility support - clients default to database 0 when no database_id specified - Add server version compatibility checks (cluster multi-database requires Valkey 9.0+) - Update documentation with warnings about SELECT command limitations and recommended configuration approach This enables cluster mode clients to connect to specific databases at initialization time, with the database selection persisting across reconnections, unlike the SELECT command which resets on reconnection. * fixed go tests - batch tests - database id tests * removed selectWithOptions * go: Remove Select command from cluster clients - Remove Select method from ClusterClient and related interfaces - Delete comprehensive database_id integration tests - Remove Select command from batch operations - Remove Select examples and tests - Add example for cluster client with database_id configuration The Select command is being removed in favor of using the database_id configuration parameter, which provides persistent database selection across reconnections. * remove config validation simplified tests * added changelog * removing batch test, and readding standalone select and removing batch select --------- Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * [Backport 2.1] Java: Multi-Database Support for Cluster Mode Valkey 9.0 (#4697) Java: Multi-Database Support for Cluster Mode Valkey 9.0 (#4658) * - Implement database selection for cluster clients when database_id != 0 - Send SELECT command to all cluster nodes using MultiNode routing - Add comprehensive error handling with clear error messages - Include test for database selection error handling in cluster mode - Support backward compatibility with servers that don't support multi-db cluster mode - Enable cluster-databases=16 for Valkey 9.0+ in cluster manager This enables cluster clients to work with non-default databases on Valkey 9.0+ servers configured with cluster-databases support, while gracefully handling older servers that don't support this feature. * update valkey9 multi db tests * fix lint * fixing valkey 9 cluster tests * add to route to all nodes in the core * rust lint fix * implement database selection in cluster mode via connection parameters - Add database_id field to ClusterParams and BuilderParams structs - Add database() method to ClusterClientBuilder for setting database ID - Pass database_id through connection info instead of post-connection SELECT - Remove SELECT command from cluster routing - Remove post-connection SELECT command logic from create_cluster_client - Add comprehensive tests for database isolation and reconnection behavior - Verify database ID persistence across node reconnections - Test multi-database cluster support with proper isolation verification This change moves database selection from a post-connection SELECT command to a connection parameter, which is more reliable for cluster mode and handles reconnections automatically. The approach works with servers that support multi-database cluster configurations (like Valkey 9.0+). * add valkey9 constraint in the core tests * fix core test not skipping test when version was lower than valkey 9 * Fix version check * refactored test test_set_database_id_after_reconnection to be similar from standalone removed is_valkey_9_or_higher * removed tests and cleanup * renamed builder.database to builder.database_id * Java: implement multi-database support for cluster mode - Move databaseId from GlideClientConfiguration to BaseClientConfiguration - Add databaseId validation and protobuf integration - Update SELECT command with AllNodes routing and reconnection warnings - Add connection management for database selection after auth - Add comprehensive tests with Valkey 9+ version guards Completes Java client implementation for Valkey 9 cluster multi-database support. * fixing database id restriction * fixed java test error select_command_invalid_database_standalone * fixed lint errors * Refactored select command to route to all nodes per default Removed the option to provide a route * java lint fix * java: Remove SELECT command from cluster client - Remove select() method from GlideClusterClient and ConnectionManagementClusterCommands interface - Delete MultiDatabaseTests.java which tested SELECT command functionality - Remove SELECT command tests from SharedCommandTests.java - Add focused test for cluster database_id configuration in ClusterClientTests - Remove SELECT import from GlideClusterClient The SELECT command is being removed from the cluster client API as database selection should be handled through client configuration (databaseId parameter) rather than runtime commands, which provides better consistency and avoids reconnection issues. * remove database_id validatation and simplified tests * added changelog * fix documentation * fix databaseId documentation --------- Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * [Backport 2.1] Node: Add Multi-Database Support for Cluster Mode (Valkey 9.0) (#4698) Node: Add Multi-Database Support for Cluster Mode (Valkey 9.0) (#4657) * - Implement database selection for cluster clients when database_id != 0 - Send SELECT command to all cluster nodes using MultiNode routing - Add comprehensive error handling with clear error messages - Include test for database selection error handling in cluster mode - Support backward compatibility with servers that don't support multi-db cluster mode - Enable cluster-databases=16 for Valkey 9.0+ in cluster manager This enables cluster clients to work with non-default databases on Valkey 9.0+ servers configured with cluster-databases support, while gracefully handling older servers that don't support this feature. * update valkey9 multi db tests * fix lint * fixing valkey 9 cluster tests * add to route to all nodes in the core * rust lint fix * implement database selection in cluster mode via connection parameters - Add database_id field to ClusterParams and BuilderParams structs - Add database() method to ClusterClientBuilder for setting database ID - Pass database_id through connection info instead of post-connection SELECT - Remove SELECT command from cluster routing - Remove post-connection SELECT command logic from create_cluster_client - Add comprehensive tests for database isolation and reconnection behavior - Verify database ID persistence across node reconnections - Test multi-database cluster support with proper isolation verification This change moves database selection from a post-connection SELECT command to a connection parameter, which is more reliable for cluster mode and handles reconnections automatically. The approach works with servers that support multi-database cluster configurations (like Valkey 9.0+). * add valkey9 constraint in the core tests * fix core test not skipping test when version was lower than valkey 9 * Fix version check * refactored test test_set_database_id_after_reconnection to be similar from standalone removed is_valkey_9_or_higher * removed tests and cleanup * renamed builder.database to builder.database_id * node changes * move tests to sharedtests.ts * reafactoring to select to all nodes being in the core * node: Remove SELECT command support from cluster client - Remove select() method from GlideClusterClient - Remove createSelect import from cluster client - Remove extensive test coverage for SELECT functionality - Remove database ID validation tests from client internals - Add minimal database ID test for cluster client configuration - Clean up ConfigurationError import that's no longer needed This change removes the SELECT command implementation that was added for Valkey 9.0+ cluster support, likely due to reliability concerns with database switching in cluster mode or to simplify the API. * added changelog * fix test, and removed comment --------- Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * [Backport 2.1] Python sync: removed select command (#4705) Python sync: removed select command (#4684) removed sync select command Signed-off-by: Lior Sventitzky <liorsve@amazon.com> * [Backport 2.1] CI/CD: Added self hosted macOS ARM runners (#4706) * CI: Added self hosted macOS runners (#4683) added self hosted runners for macos Signed-off-by: Lior Sventitzky <liorsve@amazon.com> * CD: removed self hosted mac arm runners from required CD platforms #4700 (#4702) removed self hosted mac from cd, reverted npm-cd Signed-off-by: Lior Sventitzky <liorsve@amazon.com> --------- Signed-off-by: Lior Sventitzky <liorsve@amazon.com> * [backport 2.1] Python: fixed pypi-cd workflow, fixed release tests (#4708) Python: fixed pypi-cd workflow, fixed release tests (#4703) fixed pypi upload, fixed release tests Signed-off-by: Lior Sventitzky <liorsve@amazon.com> * Fix FFI tests for non-alpine distros (#4711) Signed-off-by: Jeremy Parr-Pearson <jeremy.parr-pearson@improving.com> * MUSL Java Build Fixes (#4712) * Use zigbuild for easier cross-compilation Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Fix Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Use zigbuild for easier cross-compilation Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Use zigbuild for easier cross-compilation Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Fix musl classifier / detection Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Fix musl classifier / detection Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Fix musl classifier / detection Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update docs Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Exclude arm musl test Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update arm runner Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Update go exclusion Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Format Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> --------- Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * Revert go upload step change (#4716) Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> * CD: Fix Node CD delete artifact issue (#4717) Signed-off-by: James Xin <james.xin@improving.com> * CD: Fix Go musl CD issue (#4719) Signed-off-by: James Xin <james.xin@improving.com> * CD: Fix Node CD delete artifact issue 2nd try (#4718) * CD: Fix Node CD delete artifact issue 2nd try Signed-off-by: James Xin <james.xin@improving.com> * Linter Signed-off-by: James Xin <james.xin@improving.com> * remove temporary testing Signed-off-by: James Xin <james.xin@improving.com> --------- Signed-off-by: James Xin <james.xin@improving.com> * Python: Update ExpiryType enum references in hash field expiration docs (#4715) python: Update ExpiryType enum references in hash field expiration docs Replace deprecated ExpiryType enum values with new names in docstrings: - EX → SEC - PX → MILLSEC - EXAT → UNIX_SEC - PXAT → UNIX_MILLSEC - KEEPTTL → KEEP_TTL Updates documentation for hash field expiration commands in both async and sync clients. Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> * [Backport 2.1] Python sync/CD: added description to pypi package of the sync client (#4720) Python sync/CD: added description to pypi package of the sync client (#4714) added description to sync client Signed-off-by: Lior Sventitzky <liorsve@amazon.com> * Java: Compatibility layer module (#4692) * Java: Compatibility layer module Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: fix integration tests Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Fix codeql for java code Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Fix codeql java build Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Fix codeql java build Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Fix codeql java build Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Fix codeql java build Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Fix codeql java build Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Fix codeql java build Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Fix codeql java build Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Update java cd Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: lint issues fixed Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Add signing configuration to jedis-compatibility build.gradle Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: update dependency Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Update valkey-glide-jedis-compatibility file path Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Java: Fix build.gradle publishing code Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> --------- Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * Fix: Copy GPG secring to jedis-compatibility in Java CD workflow (#4723) fix(java-cd): copy secring.gpg to jedis-compatibility for signing Signed-off-by: jbrinkman <joe.brinkman@improving.com> * [Backport 2.1] Python Sync: revert license format (#4732) Python Sync: revert license format (#4731) revert license format Signed-off-by: Lior Sventitzky <liorsve@amazon.com> * Java: Fix Jedis compatibility layer valkey-glide dependency (#4737) * Java: Fix Jedis compatibility layer valkey-glide dependency Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> * [Backport 2.1] Python: improve UDS socket error handling (#4733) (#4755) Python: improve UDS socket error handling (#4733) fixed uds error handling Signed-off-by: Lior Sventitzky <liorsve@amazon.com> Co-authored-by: Lior Sventitzky <liorsve@amazon.com> --------- Signed-off-by: James Xin <james.xin@improving.com> Signed-off-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> Signed-off-by: Prateek Kumar <prateek.kumar@improving.com> Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com> Signed-off-by: Lior Sventitzky <liorsve@amazon.com> Signed-off-by: Jeremy Parr-Pearson <jeremy.parr-pearson@improving.com> Signed-off-by: jbrinkman <joe.brinkman@improving.com> Co-authored-by: James Xin <james.xin@improving.com> Co-authored-by: Alex Rehnby-Martin <alex.rehnby-martin@improving.com> Co-authored-by: Alexey Temnikov <alexey.temnikov@improving.com> Co-authored-by: prateek-kumar-improving <prateek.kumar@improving.com> Co-authored-by: affonsov <67347924+affonsov@users.noreply.github.com> Co-authored-by: Lior Sventitzky <liorsve@amazon.com> Co-authored-by: Jeremy Parr-Pearson <jeremy.parr-pearson@improving.com>

affonsov requested a review from a team as a code owner August 27, 2025 05:18

affonsov had a problem deploying to AWS_ACTIONS August 27, 2025 05:18 — with GitHub Actions Error

affonsov force-pushed the python/affonsov-cluster-multidb-support branch from 1d4cb9a to bbb3dba Compare August 27, 2025 16:55

affonsov had a problem deploying to AWS_ACTIONS August 27, 2025 16:55 — with GitHub Actions Error

leszek-bq reviewed Aug 27, 2025

View reviewed changes

affonsov force-pushed the python/affonsov-cluster-multidb-support branch from bbb3dba to 9f24dfc Compare August 28, 2025 01:33

affonsov had a problem deploying to AWS_ACTIONS August 28, 2025 01:33 — with GitHub Actions Error

affonsov had a problem deploying to AWS_ACTIONS August 28, 2025 01:52 — with GitHub Actions Error

alexr-bq approved these changes Aug 28, 2025

View reviewed changes

affonsov force-pushed the python/affonsov-cluster-multidb-support branch from a6d54ef to 558a099 Compare September 2, 2025 23:23

affonsov had a problem deploying to AWS_ACTIONS September 2, 2025 23:23 — with GitHub Actions Error

shohamazon requested changes Sep 3, 2025

View reviewed changes

This was referenced Sep 3, 2025

Node: Add Multi-Database Support for Cluster Mode (Valkey 9.0) #4657

Merged

Java: Multi-Database Support for Cluster Mode Valkey 9.0 #4658

Merged

Go: Add Multi-Database Support for Cluster Mode Valkey 9.0 #4660

Merged

affonsov had a problem deploying to AWS_ACTIONS September 4, 2025 20:06 — with GitHub Actions Error

fix test database_id should be skipped for versions lower than 9.0.0

a6d4749

Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com>

affonsov had a problem deploying to AWS_ACTIONS September 4, 2025 21:30 — with GitHub Actions Error

jduo reviewed Sep 4, 2025

View reviewed changes

python/glide-shared/glide_shared/config.py Show resolved Hide resolved

fix test

7c9264a

Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com>

affonsov had a problem deploying to AWS_ACTIONS September 4, 2025 22:46 — with GitHub Actions Error

affonsov had a problem deploying to AWS_ACTIONS September 4, 2025 23:45 — with GitHub Actions Error

yipin-chen approved these changes Sep 5, 2025

View reviewed changes

shohamazon approved these changes Sep 7, 2025

View reviewed changes

affonsov added 2 commits September 8, 2025 10:43

added select rounting

0a9cd97

refactored extract_client_id to be in utilities Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com>

fix database_id documentation

d8cc808

added custom command test using select Signed-off-by: affonsov <67347924+affonsov@users.noreply.github.com>

affonsov had a problem deploying to AWS_ACTIONS September 8, 2025 17:44 — with GitHub Actions Error

affonsov requested a deployment to AWS_ACTIONS September 8, 2025 17:44 — with GitHub Actions Waiting

affonsov requested a deployment to AWS_ACTIONS September 8, 2025 18:56 — with GitHub Actions Waiting

affonsov merged commit d1b0ae7 into main Sep 8, 2025
75 of 81 checks passed

affonsov deleted the python/affonsov-cluster-multidb-support branch September 15, 2025 15:20

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Python: Add Multi-Database Support for Cluster Mode Valkey 9.0 #4659

Python: Add Multi-Database Support for Cluster Mode Valkey 9.0 #4659

Uh oh!

affonsov commented Aug 27, 2025 •

edited

Loading

Uh oh!

leszek-bq left a comment

Uh oh!

shohamazon left a comment •

edited

Loading

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

yipin-chen left a comment

Uh oh!

shohamazon left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

liorsve commented Sep 7, 2025

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

8 participants

Python: Add Multi-Database Support for Cluster Mode Valkey 9.0 #4659

Python: Add Multi-Database Support for Cluster Mode Valkey 9.0 #4659

Uh oh!

Conversation

affonsov commented Aug 27, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Add Multi-Database Support for Cluster Mode (Valkey 9.0+)

Overview

Key Features

🔧 Core Implementation

🎯 Client Configuration

Technical Changes

Rust Core (glide-core)

Python Client

Test Coverage

Server Compatibility

Valkey 9.0+ Requirements

Test Environment Setup

Issue link

Checklist

Uh oh!

leszek-bq left a comment

Choose a reason for hiding this comment

Uh oh!

shohamazon left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

yipin-chen left a comment

Choose a reason for hiding this comment

Uh oh!

shohamazon left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

liorsve commented Sep 7, 2025

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

8 participants

affonsov commented Aug 27, 2025 •

edited

Loading

Rust Core (`glide-core`)

shohamazon left a comment •

edited

Loading