We have an existing aggregate created with X380 10TB NL-SAS disks, it currently consists of two RAID-Tec Groups of 18 disks each. We now want to add 12 additional disks to the aggregate, 6 to each RAID Group... so we issue this command: "aggr add-disks -aggregate data -raidgroup rg0 -diskcount 6 -simulate true"
We then get this response:
Error: command failed: Addition of disks would fail for aggregate "data" on node "node01". Reason: 6 disks needed from Pool0, but no matching disks
are available in that pool.
Yet we have the spare disks needed when we run the "disk show -spare".. and they are in Pool0... and they are X380 disks... so not really sure why this is happening??
(we have also set the RAID Max Size to 29)... any suggestions?
#Cannot add spare disks to aggregate...
dawn sleet
dawn sleet
Found one thing that looks strange... if we run a "disk show -disk ..." on a disk in the existing aggregate, is shows "Owner/Home: node01 / node02" but if we do the same on one of the spare diske is shows "Owner/Home: node01 / node01".. the aggregate is owned by node01 and the spare disks are all assigned to node01... We tried to specify the disks like "aggr add-disks -aggregate data -raidgroup rg0 -disklist 1.5.0,1.5.1,1.5.2,1.5.3,1.5.4,1.5.5 -simulate true"... but we then get this error: "Error: command failed: The specified disks do not have the same ownership as aggregate "node02". Use the "storage aggregate show-spare-disks -original-owner node02" command to display compatible spare disks."
...and if we try to assign the disks to the other node and try to add them directly, we get this eror: "Error: command failed: Assigned owner for specified disk list (node01:0a.05.0) does not match assigned owner for aggregate (data)."
ahhh ffs... the aggregate was not correctly given back because of a file-lock veto from when we did an ONTAP upgrade...
slender halo
Also, if multiple disk types exist you will see a similar message (like SSD and spinners)
Aggr add-disk … -disktype FSAS
Also may help
dawn sleet
Yep... but if you specify the disks with disklist, I guess the dsktype doesn't make sense 🙂 Anyway I find it odd that you cannot add capacity to a failed over aggregate... I know it's a long stretch, but what if one controller dies and you are low on space at the same time?... then you would just be out of luck? 😉 Also I guess the error messages could be a bit more specific... I eventually found the cause, and yes we should have checked the ONTAP upgrade process completed... which it didn't because of a file-lock veto, which also still boggles me that NetApp doesn't give you the possibility to just ignore those locks as you start the update... but there we are... I could find a lot more small issues that NetApp could fix if they had the time 🙂 Beeing as old as me, I can remember that the first 5 years working with NetApp we actually had a NetApp developer visit us once a year, just so that we could ask questions... that we actually very insightful... 🙂
winged swallow
Hi @dawn sleet can you post the output of these 2 commands:
- ::> node run <node name> disk show (need to only see the disk column, rest you can hide)
- ::> disk show -fields type,model
dawn sleet
Hi <@743120101942886442> can you post the output of these 2 commands:
1. ::> no...
Hi... yep I know this, but the problem was that the aggregate I was trying to add disks to was in a failed over state (on the other node) so not "at home"... this caused all these messages which looked like there wasn't any suitable disks avaliable to add... (but of cause there was)... trust me, I tried a few "tricks" like changing the ownership of the spare disks back and forth between two nodes etc.. but apparently you are unable to add disks to a failed over aggregate... of cause as soon as the aggregate returned home, there was no problems at all and the disks were added.. 🙂
winged swallow
Hi... yep I know this, but the problem was that the aggregate I was trying to ad...
Ok got it. Glad you were able to add the disks once it failed back.
dawn sleet
Ok got it. Glad you were able to add the disks once it failed back.
So was I 🙂 But I still don't understand why NetApp will not allow me to add it... and the error messages could have been more helpful... like "the aggregate is in a failed over state...." 🙂
brazen inlet
I think it would have worked if you had changed the disk ownership of the spare disks over to the node that currently held the aggregate. Spare disk do not get reassigned on takeover so they were still on the old node