witness: tikv leader are unbalanced during br full restore data

## Bug Report

### What did you do?
https://tcms.pingcap.net/dashboard/executions/plan/1621300
1、enable witness
2、br full restore
export AWS_ACCESS_KEY_ID=minioadmin;export AWS_SECRET_ACCESS_KEY=minioadmin;/br restore full --send-credentials-to-tikv=true --pd http://tc-pd.e2e-oltp-tikv-network-fault-rpt-witness-tps-1621300-1-379:2379 --storage s3://benchmark/tpcc-1000-4.0 --s3.endpoint http://minio.pingcap.net:9000 --check-requirements=false

### What did you expect to see?
pd leader are balanced during br full restore data

### What did you see instead?
pd leader are unbalanced during br full restore data
https://clinic.pingcap.com.cn/portal/#/orgs/31/clusters/7200346070474861503?from=1676461181&to=1676477650

### What version of PD are you using (`pd-server -V`)?
./pd-server -V
Release Version: v6.7.0-alpha
Edition: Community
Git Commit Hash: e168bd65773163529eb917ac4edbec030a9a3617
Git Branch: heads/refs/tags/v6.7.0-alpha
UTC Build Time: 2023-02-15 11:03:36
2023-02-15T19:43:37.135+0800 INFO k8s/client.go:132 it should be noted that a long-running command will not be interrupted even the use case has ended. For more information, please refer to https://github.com/pingcap/test-infra/discussions/129