In the ever-evolving landscape of database management and replication, Tungsten Clustering and Tungsten Replicator continue to be at the forefront, providing robust solutions for high-performance and scalable database environments.
The latest releases, version 6.1.24 for Tungsten Clustering and version 7.1.1 for Tungsten Replicator, bring a host of bug fixes, behavior changes, and new features that further enhance their capabilities.
Tungsten Clustering 6.1.24: Bug Fixes and Enhanced Failover Handling
The critical bug fix in Tungsten Clustering 6.1.24 addresses an issue with the Tungsten Manager. In previous versions, the manager sometimes waited longer than the specified timeout for replicator purge during failover. This bug has been rectified, ensuring a smoother failover process and minimizing downtime.
Tungsten Clustering 7.1.1: Enhanced Features, Functionality, and Bug Fixes
Behavior Changes
The tpm diag command, a diagnostic tool, now collects ps -aux output in addition to ps -efl when possible. This enhancement provides a more comprehensive view of system processes, aiding administrators in diagnosing and troubleshooting issues effectively.
Tungsten Connector sees a change in the default value of mgr-policy-fence-slave setting. Now, a replica data source is put offline 60 seconds after its replicator goes offline, preventing routing traffic to stale replicas and improving overall cluster performance.
These changes may affect scripts and integration tools. Users should review and update their configurations accordingly.
Improved CLI Functionality
The tapi command now supports the --no-connectors=1 argument in APIv2 via the new CLI argument --no-connectors, enhancing control over connectors.
The tpm mysql command now accepts the --force argument to function on Witness nodes that have a running database.
Bug Fixes and Manager Enhancements
A bug when upgrading from 6.x to 7.x that prevented switches/failovers from working has been fixed. Additionally, the change of datasource-group-id in the INI will now be correctly reflected after a tpm update.
Tungsten Manager addresses a thread synchronization issue that could cause memory leaks, enhancing overall system stability.
Installations will now succeed on hosts running Ruby 3.x, expanding compatibility and ensuring a smooth deployment experience.
The tpm report command now provides more accurate reporting by correctly displaying security information as blank when none exists.
The tpm update command no longer fails to display foreign-owned dot-files and directories, streamlining the update process.
The tpm mysql command now accepts the -e argument as a shortcut for the --execute MySQL CLI client argument, enhancing flexibility in MySQL command execution.
v6 API configuration properties within the INI file are now recognized by v7, ensuring a seamless transition for users upgrading from version 6.x to 7.x.
Tungsten Replicator 7.1.1: Command-line Tools Overhaul and Bug Fixes
Behavior Changes
Since Tungsten Replicator is a part of Tungsten Clustering, it has similar updates to the tpm diag command that now also collects ps -aux output, providing more comprehensive diagnostic information.
Bug Fixes and Installation Improvements
Installation on hosts running Ruby 3.x is now supported, ensuring compatibility with the latest Ruby versions.
The tpm report, tpm update and tpm mysql commands see the same improvements as mentioned in the Tungsten Clustering section.
Upgrade Today for a More Powerful Database Management Experience
These releases highlight the commitment of the Tungsten team to delivering a reliable, high-performance database clustering and replication solution. Users are encouraged to upgrade to the latest versions to take advantage of these improvements and ensure the optimal performance of their database environments. As always, thorough testing in a non-production environment is recommended before applying updates to a live system to avoid any unforeseen issues.
Comments
Add new comment