LVM: Recovered Path

[Matt Wittasek]

We have an N4000 attached to a VA7410 storage array. This morning, a controller failed. We replaced the controller (Controller 1) and updated the WWN.
The host can now see all the paths (both primary and alternate) and communicate with the array via that controller, however we now get the following errors:
LVM: Performed a switdh for Lun ID = 0 and LVM: Recovered Path
The devices of course vary as there are quite a few luns. LVM seems to be switching paths and recovering constantly - switching from one controller to the other and back again. This error appears to be on the host end, not the array.
Has anyone seen this before, and short of rebooting the host, has anyone found a way to correct the problem?

[Steven E. Protter]

When you replace disk or components hot for example hot swap disks or controllers, this mesage will go into syslog.

The only way I know to clear it is a reboot of the box. It should include an lbolt message.

You can ignore it until you can schedule a maintenance window and boot the box.

SEP

[Helen French]

I would be concerned about these messages especially after a controller replacement. My first guess would be that both controllers are trying to get the 'ownership' of some particular LUNs or disks. I have seen these type of situations, when both controllers in the disk array trying to hold the ownership. Check your syslog carefully and see which LUN is giving the error (may be LUN 0). Check the disk array status from Grid manager and see which controllers owns which LUN. If a specific LUN is not assigned to any controller, then that's your issue. Try to assign it to the proper one or you may have to recreate it (that happens too!).

I would defenitely get all these done with the disk array specialists.