ONTAP Discussions

System Overheaded, Filer Head B Failed

SeanPerryman
2,856 Views

I have an FAS2050 in my production environment, and contoller B has had issues with heat before. Hadn't swapped out the controller because of the old "if it ain't broke, don't fix it" and it was running; just was alerting about temperature (though the temp was fine). The other day during an electrical storm the AC unit for the server room went out, and as a result most everthing thermal'd and shut down. Upon restarting the NetApp Filer, I am getting the message FAS2050a has taken over this node on the controller B na_admin portal, and a similar message on controller A.

 

Since I had an extra user controller I sourced, I swapped the CF card and installed the controller. Now, during the boot, it tells me that IN0, IN1, and IN2 ports on IOM B have been disabled due to excessive Phy changes. It is also telling me that there are no disks attached to the system (though controller A is functioning just fine and the system is in use).

 

It would see that I need to assign disks to this controller, but will doing so interrupt the operation of controller A? Unfortunately I didn't set the system up (or mess with the configuration before), but my understanding was that controller B was just a fail over for when controller A had trouble. Is that a correct assumpton, or do some of the disks need to be assigned to the B controller to load balance the disk operations?

 

 

2 REPLIES 2

SeanPerryman
2,838 Views

On the healthy node running the disk show command shows all the disks. On the replacement now, running disk show only shows the disks from the healthy node. Running show disk -v on the replacement node  shows all the disks.Checking the system ID of the replacement node, it matches what the disk show -v output is for the FAS2050b. Is there a command I can run that will bring those disks back online for the B controller maybe?

SeanPerryman
2,830 Views

Just needed to assign the disks to the new controller, reboot, run cf giveback on the health node. Problem solved!

 

 

https://library.netapp.com/ecm/ecm_download_file/ECMM1280334

Public