subtask failover shouldn't change state of subtask

[D3Hunter]

Bug Report

Please answer these questions before submitting your issue. Thanks!

1. Minimal reproduce step (Required)

subtask failover shouldn't change state of subtask, we should keep the state, and let scheduler handle it depends on whether the subtask is idempotent(this part will be changed when we fix #46735)

tidb/disttask/framework/storage/task_table.go

Line 641 in aba8342

if err := sqlexec.FormatSQL(sql, "update mysql.tidb_background_subtask set state = %? ,exec_id = (case ", proto.TaskStatePending); err != nil {

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

3. What did you see instead (Required)

4. What is your TiDB version? (Required)

master

[D3Hunter]

refer #47298 too

[D3Hunter]

currently, we take all subtasks as idempotent and run the failovered subtask again directly, so this issue can be closed.