Subscribe
Accepted Solution

Bricked controller - need to netboot - lab environment

I managed to brick one of the controllers in my FAS2220 HA pair running 8.1.2 (7 Mode). 

 

It originally had 4 disks assigned and was sat there in an active/active config with only the root vol across 3 disks with one as a spare.  I had to give up the DS2246 shelf attached to it for U requirements and basically did a p*ss poor job of decomissioning it.  Turns out 2 of the root vol disks were located there, the shelf is long gone now.  I have two disks spare (now unowned) on the first controller to assign to the bricked controller but booting into OnTAP just isn't going to happen without netboot I fear.

 

Is there anywhere I can get the boot image from?  I don't have a subscription contract any more as this unit was decommed a while back, gutted.

 

I'm getting:

 

PANIC: raid: there are no data or parity disks in SK process rc on release

 

Re: Bricked controller - need to netboot - lab environment

Hi, unfortunately I'm afraid without a Support contract not sure where you'll be able to obtain the netboot.tgz file from.

 

However, you should however be able to boot into ONTAP using the boot media:

  • abort the AUTOBOOT process
  • drop to the LOADER-A> prompt
  • perform a boot_primary from the boot media.

Also, assuming it was a 3 disk RAID-DP and "just" a degraded aggregate, you could change to a RAID-4, leaving one spare for the system.

 

Hope this helps.

 

Cheers,

Grant.

 

Re: Bricked controller - need to netboot - lab environment

You do not need netboot to reinitalize disks. Simply interrupt boot with ^C when prompted for special boot menu and select 4. This will keep your existing NetApp binaries on boot device.

 

Do not forget that you will need to re-enter all your licenses. If you did not record them, you need access to support site to fetch.

Re: Bricked controller - need to netboot - lab environment

To be clear the Boot Menu Option 4 will reinitialise all the disks owned by that system. Not sure that is what is being asked - but might be wrong!

 

Since there is still a single disk available from the root aggregate, it should simply reconstruct once the spares are made available, with all data intact.

 

 

Re: Bricked controller - need to netboot - lab environment

Well, the error is "no data or parity disk". Of course it is possible that disk labels were upgraded to newer version in which case one indeed needs newer (Data) ONTAP. But that's just speculation without seeing full boot log. But if the problem is simply that root aggregate had been lost, then no netboot is needed.

 

Re: Bricked controller - need to netboot - lab environment

Thanks for the replies.

 

I ended up unassigning a couple of disks from the first controller and then ran an option 4 on the bricked controller.  It must have grabbed them and rebuilt the root vol as it is now running.


Thanks for the help, really appreciate it.