I have an environment that consist of FAS2040 (two nodes) where one of the nodes failed. At that time the previous admin moved aggr0 (that contains the root volume) from that node to the second node because the controller broke down and he wasn't able to replace it for a month.
Filer1 is the controller to which the aggregate was moved and from which we need to remove it and bring it back to Filer2.
The Data ONTAP is 8.0.3 7mode.
Below is the action plan that I have. I'm not certain for the last part that is making the aggregate to be root and how to choose again the old root volume.
Summary:
1. Move a nonroot aggregate from filer <Filer1> to filer <Filer2>.
2. Mark the old root volume on the aggregate as root volume on filer <Filer2>.
Action plan
1. On the source storage system <Filer 1> to locate the disks that contain the aggregate:
aggr status <aggr_name> -r
2. Take the aggregate offline from FILER1.
For the traditional volumes:
FILER1>aggr offline <aggr_name>
Switch to diag mode (if needed).
FILER1> priv set diag
Offline the aggregate:
FILER1*> aggr offline <aggr_name> -a
WARNING: The aggregate contains one or more volumes. In order to offline the aggregate, all of these volumes must be unmounted. Do you want to unmount all of the volumes hosted by the aggregate? {y|n}: y
FILER1*> aggr status --------- ------- test offline
FILER1*>priv set admin
3. Remove the ownership of the disks
>priv set advanced
*>disk remove_ownership <disk name>
(if needed the command can be executed with force FILER1*>disk remove_ownership -f <disk name>)
>priv set admin
4. Log in the other node <Filer 2> in maintenance mode and assign the disks.
Reboot the filer. When prompted for special boot menu press ctrl+c
Select option 5) Maintenance mode boot
Verify that the disks are visible
>disk show -a
>disk assign <disk name>
Check that no further disks belonging to the original aggregate are left on the original node <Filer 1>.
FILER1>aggr status –r
5. Bring the aggregate online (after it was moved)
FILER2>aggr status -r
Online the relocated aggregate from FILER2:
FILER2> aggr online <aggr_name>
6. After the aggregate is online you need to mark it as root aggregate.
>aggr options <aggr_name> <option>
You should the option root
7. Choose the aggr0 as the root aggregate. After that reboot the system again.
8. Go to the special boot menu for which you need to press ctrl+c
The menu should look like this:
(1) Normal boot. (2) Boot without /etc/rc. (3) Change password. (4) Initialize all disks. (4a) Same as option 4, but create a flexible root volume. (5) Maintenance mode boot.
Without choosing any option use the following command
vol_pick_root <vol_name> <aggr_name>
9.Reboot the filer
>halt
In the loader prompt perform a normal boot
loader>boot_ontap
Thank you in advance for the help!
The main problem that I have is with steps 7 and 8. I have tried with the vol options command to mark the old root volume as root again and it failed with the message below.
But when I try to use it in maintenance mode it pops with the following message:
vol: Please use the "aggr" command for maintenance operations on traditional volumes and aggregates.
Now, maintenance mode does not support "vol" command as suggested. Why did you run "vol options"? Following the steps, you have already set root flag for the aggregate in step 6 by " aggr options <aggr> root". Check if root flag is set for the aggr by "aggr status" In maintenance mod, volume level operation for FlexVol is not possible. To setting root flag for volume, just do step 8 after setting root flag for aggregate.
@YIshikawa The reason why I'm worried at step 8 is because I'm not sure if the vol_pick_root <vol_name> <aggr_name> command will overwrite the root volume or will only point it as a root volume without overwriting the information that is actully still there. Are you aware which action it will take?
It does not overwrite any volume. Just manipulate volume options.
BTW, after vol_pick_root, you will go back to special boot menu. Then, go "1) Normal boot". Or, if you would like to reboot, press 8 here in stead of running "halt".
