add index failed with error “Error 1105 (HY000): [pd] failed to get cluster id” when inject pdleader network partition

[Lily2025]

Bug Report

Please answer these questions before submitting your issue. Thanks!

1. Minimal reproduce step (Required)

1、run workload
2、add index for one table
3、inject pdleader network partition

case config

• name: config_ha_pdleader_to_all_network_partition
  default: |
  tag: "ha-test-add-index"
  workloads:
  - type: sysbench
  storage: s3://benchmark/sysbench_64_7000w
  db: sysbench_64_7000w
  tableNum: 64
  tableSize: 70000000
  threads: 64
  subType: oltp_read_write
  ignoreErrors: "all"
  cases:
  - name: "ha_pdleader_to_all_network_partition"
  warmUpTime: 1m
  faultType: network_partition
  selector: "pdleader_to_all"
  period: "@every 5m"
  faultDuration: 2m
  faultTotalRunTime: 30m
  ddl:
  addIndex:
  - db: "sysbench_64_7000w"
  table: "sbtest1"
  timeout: 60m
  times: 5
  index_col_name: "c"
  job_type: "ingest"
  admin_check: false
  count_check: false

2. What did you expect to see? (Required)

add index success

3. What did you see instead (Required)

add index failed with error “Error 1105 (HY000): [pd] failed to get cluster id” when inject pdleader network partition

"add index failed at 2023-10-29 16:27:07 (Error 1105 (HY000): [pd] failed to get cluster id)
operatorLogs:
[2023-10-29 16:25:16] ###### start adding index
alter table sbtest1 add index index_test_1698567916120 (c)
[2023-10-29 16:25:16] ###### wait for ddl job finish"

4. What is your TiDB version? (Required)

git hash：3682bd82cac05a75241ed9fc9f29a1d320d776a7