Dear friends,
We have a Oracle Linux (6.4) IBM blade server with some Hitachi storage LUNS attached, but, after a hardware maintenance, we start to receiving message:
[root@sdvlin03 ~]# dmesg | grep Cable
qla2xxx [0000:03:05.0]-8038:1: Cable is unplugged...
qla2xxx [0000:03:05.1]-8038:2: Cable is unplugged...
[root@sdvlin03 ~]# cat /sys/class/fc_host/host*/port_state
Linkdown
Linkdown
HBA WWN remais the same, so we don't have zonning issues at this point.
Running "nodefind" inside our brocade switch, WWN is not located:
SanBroc01:FID128:admin> nodefind 21:00:00:e0:8b:9a:8d:7a
No device found
So, the LUNS are not avaliable anymore. We check connections and everything looks fine. So, we perform some software actions trying to fix the problem, but none of them works.
We run Kernel update, HBA driver update, linux updates, restart multipath services, several reboots, extract and reinsert server from chassis, etc...
Red Hat has a document about it, but it's accessible only by subscription.
https://access.redhat.com/solutions/32767
Suggestions are very well welcomed.
Thanks in advance