Info:Mpx:Path Bus 1 Tgt 1 Lun 6 to CETVxxxxxxxx is dead

某日服务器dmesg反复出现以下告警

Info:Mpx:Path Bus 1 Tgt 1 Lun 6 to CETVxxxxxxxx is alive

................

Info:Mpx:Path Bus 1 Tgt 1 Lun 6 to CETVxxxxxxxx is dead

.........

Error:Mpx:Path Bus 1 to VNX CETVxxxxxxxx port SP B3is dead

Error:Mpx:Path Bus 1 to VNX CETVxxxxxxxx port SP B3is dead

Info:Mpx:Path Bus 1 Tgt 1 Lun 6 to CETVxxxxxxxx is dead
登入VNX5400 EMC Unisphere 后台查看SPB日志:

CETVxxxxxxx > System > Monitoring and Alerts > SP Event Logs:

 Info:Mpx:Path Bus 1 Tgt 1 Lun 6 to CETVxxxxxxxx is dead

Description:Fibre Channel loop up on physical port 3.

Description:Fibre Channel loop down on physical port 3.

Description:Fibre Channel loop up on physical port 3.

Description:Fibre Channel loop down on physical port 3.

Description:Fibre Channel loop up on physical port 3.

Description:Fibre Channel loop down on physical port 3.

与服务器端告警相吻合。

可见控制器B光纤或与之联络的光交可能有问题。

进入CETVxxxxxxxx > System > Hardware > Storage Hardware, 通过指示灯闪烁功能定位SP B3 端口

Info:Mpx:Path Bus 1 Tgt 1 Lun 6 to CETVxxxxxxxx is dead

 

由于链路中断次数多且短暂,但影响生产,决定停机更换光纤测试效果。

将SP-B 端口3的光纤更换后,启服务,经过观察,未复现故障。

Info:Mpx:Path Bus 1 Tgt 1 Lun 6 to CETVxxxxxxxx is dead

因此推断此次故障原因为SP-B 3口光纤信号不稳定,更换光纤后解决了问题。