We have a FAS8040 (Clustered DataOntap 9.1) that currently has the following configuration:
aggr1_01 - 20 disks (18+2 raid dp) - node1
aggr1_02 - 44 disks (two 22 disk raid groups (20+2 raid dp) - node2
I am adding a new shelf of 24 disks with the intention of adding all 24 disks to aggr1_01 to address performance and capacity issues.
Can you please look over my steps below and let me know if I am missing anything or simply confirm what I am doing is correct?
- Increase the maxraidsize for aggr1_01 from 20 to 22
- storage aggregate modify aggregate aggr1_01 -maxraidsize 22
- Disable disk auto assign
- storage disk option modify -autoassign off
- rack, stack, set shelf id, cable new shelf
- confirm cabling with config advisor
- Assign new unassigned disks to node1
- storage disk assign -disk all -node node1
- add 2 of the new disks to the existing raid group in aggr1_01
- storage aggregate add-disks -aggregate aggr1_01 -raidgroup rg0 -diskcount 2
- add 22 of the new disks to aggr1_01 as a new raid group
- storage aggregate add-disks -aggregate aggr1_01 -raidgroup new -diskcount 22
- OR issue this command instead of the above two commands:
- storage aggregate add-disks -aggregate aggr1_01 -raidgroup all -diskcount 24
- (either way should end up with two 22 (20+2 raid dp) raid groups in aggr1_01, right?)
At this point I should have successfully added 24 new disks to aggr1_01 and it should consist of two 22 disk raid groups, just like the other aggregate aggr1_02.
In order to take advantage of the additional spindle count, do I need to reallocate all the volumes on aggr1_01??
What am I missing or not considering?
Thanks for your help!