TiDB Operator 1.6.2 Release Notes

Release date: July 14, 2025

TiDB Operator version: 1.6.2

New features

Support compressing backup logs and configuring scheduled tasks to optimize storage usage (#6033, @RidRisR)
Support the new abort restore command to clean up restore related meta data (#6288, @RidRisR)

Support evicting TiKV leaders before deleting a store when scaling in TiKV nodes (#6239, @liubog2008)
When upgrading PD, TiDB Operator uses the new pd/api/v2/ready API to check member readiness (#6243, @csuzhangxc)
Before gracefully restarting TiKV, TiDB Operator flushes backup log files to disk to prevent data loss (#6057, @YuJuncen)
Support automatic retries of failed restore tasks based on policies (#6092, @RidRisR)
Make the --pitrRestoredTs parameter optional in the restore custom resource (CR) (#6135, @RidRisR)
Support namespace-level log backup tracking (#6160, @WangLe1321)
Support specifying forced path-style URLs for FIPS backups (#6250, @3pointer)
Support using a custom S3 endpoint in the backup manager (#6268, @3pointer)

Fix the issue that the TidbCluster CR is incorrectly marked as Ready when some Tiproxy Pods are unhealthy (#6151, @ideascf)
Fix the issue that EBS volume snapshot backup tasks cannot exit normally due to invalid TC format(#6087, @BornChanger)
Fix the issue that incorrect data is restored during PITR due to the gc.ratio-threshold parameter being reset after a TiKV restart (#6267, @YuJuncen)