tests: "txn_cleanup_locks_batch_size" is flaky

CI link: https://github.com/tikv/client-rust/actions/runs/21196928386/job/60974865100?pr=524
PR: #524

stdout ───

    running 1 test
    init finish with 261 regions
    test txn_cleanup_locks_batch_size ... FAILED

    failures:

    failures:
        txn_cleanup_locks_batch_size

    test result: FAILED. 0 passed; 1 failed; 0 ignored; 0 measured; 5 filtered out; finished in 1.26s
    
  stderr ───
    [2026-01-21T04:15:47Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "127.0.0.1:2379"
    [2026-01-21T04:15:47Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "127.0.0.1:2379"
    [2026-01-21T04:15:47Z INFO  tikv_client::pd::cluster] All PD endpoints are consistent: ["127.0.0.1:2379"]
    [2026-01-21T04:15:47Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "http://127.0.0.1:2379/"
    [2026-01-21T04:15:47Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "http://127.0.0.1:2379/"
    [2026-01-21T04:15:47Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "http://127.0.0.1:2379/"
    [2026-01-21T04:15:47Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "http://127.0.0.1:2379/"
    [2026-01-21T04:15:47Z INFO  tikv_client::pd::client] connect to tikv endpoint: "127.0.0.1:20162"
    [2026-01-21T04:15:47Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "127.0.0.1:20162"
    [2026-01-21T04:15:47Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "127.0.0.1:2379"
    [2026-01-21T04:15:47Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "127.0.0.1:2379"
    [2026-01-21T04:15:47Z INFO  tikv_client::pd::cluster] All PD endpoints are consistent: ["127.0.0.1:2379"]
    [2026-01-21T04:15:47Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "http://127.0.0.1:2379/"
    [2026-01-21T04:15:47Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "http://127.0.0.1:2379/"
    [2026-01-21T04:15:47Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "http://127.0.0.1:2379/"
    [2026-01-21T04:15:47Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "http://127.0.0.1:2379/"
    [2026-01-21T04:15:47Z INFO  tikv_client::pd::client] connect to tikv endpoint: "127.0.0.1:20161"
    [2026-01-21T04:15:47Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "127.0.0.1:20161"
    [2026-01-21T04:15:47Z INFO  tikv_client::pd::client] connect to tikv endpoint: "127.0.0.1:20160"
    [2026-01-21T04:15:47Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "127.0.0.1:20160"
    [2026-01-21T04:15:47Z INFO  tikv_client::pd::client] connect to tikv endpoint: "127.0.0.1:20162"
    [2026-01-21T04:15:47Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "127.0.0.1:20162"
    [2026-01-21T04:15:48Z INFO  tikv_client::pd::timestamp] TSO stream terminated
    [2026-01-21T04:15:48Z INFO  tikv_client::pd::timestamp] TSO stream terminated
    [2026-01-21T04:15:48Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "127.0.0.1:2379"
    [2026-01-21T04:15:48Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "127.0.0.1:2379"
    [2026-01-21T04:15:48Z INFO  tikv_client::pd::cluster] All PD endpoints are consistent: ["127.0.0.1:2379"]
    [2026-01-21T04:15:48Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "http://127.0.0.1:2379/"
    [2026-01-21T04:15:48Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "http://127.0.0.1:2379/"
    [2026-01-21T04:15:48Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "http://127.0.0.1:2379/"
    [2026-01-21T04:15:48Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "http://127.0.0.1:2379/"
    [2026-01-21T04:15:48Z INFO  tikv_client::pd::client] connect to tikv endpoint: "127.0.0.1:20162"
    [2026-01-21T04:15:48Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "127.0.0.1:20162"
    [2026-01-21T04:15:48Z INFO  tikv_client::pd::client] connect to tikv endpoint: "127.0.0.1:20161"
    [2026-01-21T04:15:48Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "127.0.0.1:20161"
    [2026-01-21T04:15:48Z INFO  tikv_client::pd::client] connect to tikv endpoint: "127.0.0.1:20160"
    [2026-01-21T04:15:48Z INFO  tikv_client::common::security] connect to rpc server at endpoint: "127.0.0.1:20160"
    [2026-01-21T04:15:48Z INFO  tikv_client::request::plan] CleanupLocks::execute, inner region error:peer is not leader for region 678, leader may Some(id: 680 store_id: 5)
    thread 'txn_cleanup_locks_batch_size' panicked at tests/failpoint_tests.rs:130:5:
    assertion `left == right` failed
      left: 575
     right: 512
    note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

Start with txn_cleanup_locks_batch_size in tests/failpoint_tests.rs and reproduce the failure using the CI output as a reference. Investigate why the cleanup result can be 575 instead of the expected 512, including the reported not-leader error. Done means the test reliably passes in CI while still checking the intended batch-size behavior.