[TiDB Usage Environment] Production Environment / Testing / PoC
[TiDB Version]
[Reproduction Path] What operations were performed when the issue occurred
[Encountered Issue: Issue Phenomenon and Impact]
[Resource Configuration] Go to TiDB Dashboard - Cluster Info - Hosts and take a screenshot of this page
[Attachments: Screenshots / Logs / Monitoring]
After upgrading dmctl to v8.0.0, there is a bug. How can I roll back to v7.6.0?
Back up the current cluster, stop the current DM cluster, download the old version, replace the binary files, restore the configuration files, start the DM cluster, and verify the cluster status.
Generally, in this situation, both the old v7.6.0 and the new v8.0.0 are installed. So uninstalling the new dmctl will naturally revert to the old dmctl.
Stop DM-worker and DM-master: First, you need to stop the running DM-worker and DM-master processes. You can use the appropriate stop commands or service management tools to stop these services.
Download dmctl version v7.6.0: Visit the official GitHub repository of TiDB Data Migration (DM) or other reliable package management tools to download version v7.6.0 of dmctl.
Replace the dmctl binary file: Replace the previously installed v8.0.0 version of the dmctl binary file with the downloaded v7.6.0 version in the same location.
Restart DM-worker and DM-master: Use the appropriate start commands or service management tools to restart the DM-worker and DM-master services.
Verify the rollback is successful: Use the dmctl --version command to check the currently running dmctl version and ensure it shows v7.6.0.