9.2. 恢复不健康的 etcd 集群

流程

1. 输入以下命令登录到 etcd pod：

   $ oc rsh -n openshift-etcd -c etcd <etcd_pod_name>

   Copy to Clipboard Toggle word wrap

2. 输入以下命令输出 etcd 集群的健康状况：

   sh-4.4# etcdctl endpoint status -w table

   Copy to Clipboard Toggle word wrap

   输出示例

   +------------------------------+-----------------+---------+---------+-----------+------------+-----------+------------+--------------------+--------+
   |          ENDPOINT            |       ID        | VERSION | DB SIZE | IS LEADER | IS LEARNER | RAFT TERM | RAFT INDEX | RAFT APPLIED INDEX | ERRORS |
   +------------------------------+-----------------+---------+---------+-----------+------------+-----------+------------+--------------------+--------+
   | https://192.168.1xxx.20:2379 | 8fxxxxxxxxxx    |  3.5.12 |  123 MB |     false |      false |        10 |     180156 |             180156 |        |
   | https://192.168.1xxx.21:2379 | a5xxxxxxxxxx    |  3.5.12 |  122 MB |     false |      false |        10 |     180156 |             180156 |        |
   | https://192.168.1xxx.22:2379 | 7cxxxxxxxxxx    |  3.5.12 |  124 MB |      true |      false |        10 |     180156 |             180156 |        |
   +-----------------------------+------------------+---------+---------+-----------+------------+-----------+------------+--------------------+--------+

   Copy to Clipboard Toggle word wrap

流程

1. 要确认 etcd pod 失败，请输入以下命令：

   $ oc get pods -l app=etcd -n openshift-etcd

   Copy to Clipboard Toggle word wrap

   输出示例

   NAME     READY   STATUS             RESTARTS     AGE
   etcd-0   2/2     Running            0            64m
   etcd-1   2/2     Running            0            45m
   etcd-2   1/2     CrashLoopBackOff   1 (5s ago)   64m

   Copy to Clipboard Toggle word wrap

   失败的 etcd pod 可能具有 CrashLoopBackOff 或 Error 状态。

2. 输入以下命令删除失败的 pod 及其 PVC：

   $ oc delete pods etcd-2 -n openshift-etcd

   Copy to Clipboard Toggle word wrap