ESXi 主机故障导致 ONTAP Tools rook-ceph pod 损坏
适用于
- NetApp ONTAP 工具适用于 VMware vSphere (OTV) 10.x
- VMware vSphere
问题
- 在计划外 ESXi 主机关闭或崩溃后,OTV 插件无法连接到 NetApp 存储系统。
- vCenter 插件显示
no healthy upstream错误,OTV 插件 UI 和 Swagger 界面均无法访问。 - 观察到多个 OTV 相关的 pod 处于
CrashLoopBackOff或错误状态,包括 rook-ceph 组件(mgr、mon、osd)。- 从 OTV diag shell 运行
kr get pod命令:
- 从 OTV diag shell 运行
root@otv1:/home/diag# kr get podsNAME READY STATUS RESTARTS AGEcsi-rbdplugin-provisioner-57fb48c589-k76f8 5/5 Running 30 (133m ago) 103dcsi-rbdplugin-rgngl 3/3 Running 18 (133m ago) 103drook-ceph-crashcollector-atlmavotvnode1-7865857b4f-b4c92 1/1 Running 6 (133m ago) 103drook-ceph-mgr-a-779c5fbcc-rxkfk 1/2 CrashLoopBackOff 432 (114s ago) 103drook-ceph-mon-a-665989c77b-qfshh 1/2 CrashLoopBackOff 314 (4m7s ago) 103drook-ceph-operator-79cc9bfb75-wsq2w 1/1 Running 8 (133m ago) 103drook-ceph-osd-0-7766d89c6f-h6nx9 0/2 Init:0/3 153 (2m10s ago) 57drook-ceph-osd-prepare-atlmavotvnode1-l5tls 0/1 Completed 0 57drook-ceph-tools-699f655bb7-tjnbd 1/1 Running 5 (133m ago) 103d
- 来自 rook-ceph-mon pod 的示例日志输出:
2026-06-09T16:33:15.428641389Z stderrF [db/db_impl/db_impl_compaction_flush.cc:3496] Compaction error: Corruption: block checksum mismatch: stored=3922177252, computed=190925503, type=4 in /var/lib/ceph/mon/ceph-a/store.db/225181.sst offset 216990 size 60822026-06-09T16:33:15.423996324Z stderrF MonitorDBStore.h:355: ceph_abort_msg("failed to write to db")2026-06-09T16:33:15.424069239Z stderrF *** Caught signal (Aborted) ***