1、[FAQ] request timed out, possibly due to connection lost
1、确认 PD 之间的网络
2、确认磁盘/CPU 的负载情况
2、 [FAQ] apply request took too long
1、确认部署环境，有没有混布 IO 比较重的程序
2、确认磁盘的负载情况
3、 [FAQ] range failed to find revision pair
升级 PD 版本
4、 [FAQ] rejected stream from remote peer because it was removed
确认 cluster-id 是否和 Leader 有差异，可以判断是否是没有清理干净数据导致的
5、 [FAQ] Prometheus address is not set in PD and etcd
修复：tiup ctl pd --pd=“http://\{pd-ip}:{pd-port}” config set metric-storage http://\{prometheus-ip}:{prometheus-port}
验证：curl http://pdhost:pdport/pd/api/v1/config