added flush log backup tasks (pingcap#2718)

Author: YuJuncen

en/restart-a-tidb-cluster.md

@@ -69,3 +69,5 @@ When the number of TiKV region leaders drops to zero, according to the value of
     1. TiDB Operator calls the PD API and adds evict-leader-scheduler for the TiKV store.
     2. When the number of TiKV region leaders drops to zero, TiDB Operator deletes the Pod and recreates it.
     3. When the new Pod becomes ready, remove the evict-leader-scheduler for the TiKV store by calling the PD API.
+
+Before deleting a Pod, if a log backup task is running, TiDB Operator attempts to flush the backup logs to the external storage. To disable this feature, add the annotation `tidb.pingcap.com/tikv-restart-without-flush-log-backup` to the TidbCluster CR and set it to any value.

zh/restart-a-tidb-cluster.md

@@ -58,3 +58,5 @@ kubectl -n ${namespace} annotate pod ${tikv_pod_name} tidb.pingcap.com/evict-lea
     1. 调用 PD API，为对应 TiKV store 添加 evict-leader-scheduler。
     2. 当 TiKV region leader 数掉到 0 时，删除 Pod 并重建 Pod。
     3. 当新的 Pod Ready 后，调用 PD API 删除对应 TiKV store 的 evict-leader-scheduler。
+
+在删除 Pod 之前，如果检测到有日志备份任务正在运行，TiDB Operator 会尝试强制将备份日志写出到外部存储。若要禁用此功能，你可以在 TidbCluster CR 中添加 annotation `tidb.pingcap.com/tikv-restart-without-flush-log-backup` 并设置为任意值。