(leader) Upgrading - PD For each PD replica - If it’s the leader of PD cluster, transfer the leadership - Upgrade PD V0 PD V0 (leader) PD V0 PD V1 PD V0 (leader) PD V0 PD V0 PD V0 PD V1 (leader) PD PingCAP.com Upgrading - PD For each PD replica - If it’s the leader of PD cluster, transfer the leadership - Upgrade PingCAP.com Upgrading - TiKV For each TiKV replica - Evict region leaders - Wait

Store 4 PD PD PD PD: “Hey, Node1, create a new replica of Region 2 in Node 4, and transfer your leadership of Region 2 to Node 2” How scale-out works inside TiDB TiKV Node 3 Store 3 TiKV Node 1 Store

impact on the cluster operation. Before disabling a leader, you need to switch the leadership. During the leadership change, performance jitter of about 3 seconds will occur. 10.2.4.2.2 Minority of replicas (BR) • Alleviate the issue that the latency of the PITR log backup progress increases when Region leadership migration occurs #13638 @YuJuncen • Enhance support for connection reuse of log backup and PITR @YuJuncen • Alleviate the issue that the latency of the PITR log backup progress increases when Region leadership migration occurs #13638 @YuJuncen • TiCDC • Fix the issue that restarting the changefeed might

impact on the cluster operation. Before disabling a leader, you need to switch the leadership. During the leadership change, performance jitter of about 3 seconds will occur. 10.2.4.2.2 Minority of replicas (BR) • Alleviate the issue that the latency of the PITR log backup progress increases when Region leadership migration occurs #13638 @YuJuncen • Enhance support for connection reuse of log backup and PITR @YuJuncen • Alleviate the issue that the latency of the PITR log backup progress increases when Region leadership migration occurs #13638 @YuJuncen • TiCDC • Fix the issue that restarting the changefeed might

impact on the cluster operation. Before disabling a leader, you need to switch the leadership. During the leadership change, performance jitter of about 3 seconds will occur. 10.2.4.2.2 Minority of replicas @YuJuncen • Alleviate the issue that the latency of the PITR log backup progress increases when Region leadership migration occurs #13638 @YuJuncen • TiCDC • Fix the issue that restarting the changefeed might @RidRisR • Alleviate the issue that the latency of the PITR log backup progress increases when Region leadership migration occurs #13638 @YuJuncen • Improve the speed of merging SST files during data restore

impact on the cluster operation. Before disabling a leader, you need to switch the leadership. During the leadership change, performance jitter of about 3 seconds will occur. 10.2.4.2.2 Minority of replicas (BR) • Alleviate the issue that the latency of the PITR log backup progress increases when Region leadership migration occurs #13638 @YuJuncen • Enhance support for connection reuse of log backup and PITR @YuJuncen • Alleviate the issue that the latency of the PITR log backup progress increases when Region leadership migration occurs #13638 @YuJuncen • TiCDC • Fix the issue that restarting the changefeed might

impact on the cluster operation. Before disabling a leader, you need to switch the leadership. During the leadership change, performance jitter of about 3 seconds will occur. 10.2.4.2.2 Minority of replicas (BR) • Alleviate the issue that the latency of the PITR log backup progress increases when Region leadership migration occurs #13638 @YuJuncen • Enhance support for connection reuse of log backup and PITR @YuJuncen • Alleviate the issue that the latency of the PITR log backup progress increases when Region leadership migration occurs #13638 @YuJuncen • TiCDC • Fix the issue that restarting the changefeed might

impact on the cluster operation. Before disabling a leader, you need to switch the leadership. During the leadership change, performance jitter of about 3 seconds will occur. 10.2.4.2.2 Minority of replicas (BR) • Alleviate the issue that the latency of the PITR log backup progress increases when Region leadership migration occurs #13638 @YuJuncen • Enhance support for connection reuse of log backup and PITR @YuJuncen • Alleviate the issue that the latency of the PITR log backup progress increases when Region leadership migration occurs #13638 @YuJuncen • TiCDC • Fix the issue that restarting the changefeed might

impact on the cluster operation. Before disabling a leader, you need to switch the leadership. During the leadership change, performance jitter of about 3 seconds will occur. 10.2.4.2.2 Minority of replicas (BR) • Alleviate the issue that the latency of the PITR log backup progress increases when Region leadership migration occurs #13638 @YuJuncen • Enhance support for connection reuse of log backup and PITR @YuJuncen • Alleviate the issue that the latency of the PITR log backup progress increases when Region leadership migration occurs #13638 @YuJuncen • TiCDC • Fix the issue that restarting the changefeed might

impact on the cluster operation. Before disabling a leader, you need to switch the leadership. During the leadership change, performance jitter of about 3 seconds will occur. 10.2.4.2.2 Minority of replicas (BR) • Alleviate the issue that the latency of the PITR log backup progress increases when Region leadership migration occurs #13638 @YuJuncen • Enhance support for connection reuse of log backup and PITR @YuJuncen • Alleviate the issue that the latency of the PITR log backup progress increases when Region leadership migration occurs #13638 @YuJuncen • TiCDC • Fix the issue that restarting the changefeed might