FAS2820 ONTAP 9.14 maximize data aggregate sizing

[jessango]

Hello.

We have a FAS2820 HA pair with 12 internal drives + 36 external drives (4TB NLSAS drives).

We are trying to figure out the best approach to maximize the available size of the aggregates in a symmetric configuration.

The internal shelf (12 drives) is using ADP.

Root aggregates have 5 ADP root partitions (1 raid group using RAID_DP).

Our main concern is whether we can create 2 data aggregates (1 per node) with the following config (would leave 4 entire drives as spare and 2 partitioned drives also):

- rg0: 16 entire disks (RAID_DP > 14 data drives + 2 parity drives)

- rg1: 5 data partitions (RAID _DP > 2 parity partitions + 3 data partitions)

I wonder if we will break any best practices if we create this aggregate layout. 

Any help is appreciated. 

Best regards. 

[TMACMD]

First, verify all disks are split evenly. (24 per node)

Can check with {disk show -owner node-01 ; disk show -owner node-02} making sure each has 24

Then manually create the aggregates

aggr create -diskcount 22 -maxraidsize 22 -raidtype raid_tec -node node-01 -aggregate sata_01

aggr create -diskcount 22 -maxraidsize 22 -raidtype raid_tec -node node-02 -aggregate sata_02

By default, ONTAP allows up to 20 disks with Raid-DP. Moving to raid-tec allows up to 29 in the raid group.

You should end up with all drives ultimately being partition (root-Data), maximizing your capacity.

In my example, I left 2 spares per node. You could increase the "22" above to be "23" and get a little more capacity.

aggr create -diskcount 23 -maxraidsize 23 -raidtype raid_tec -node node-01 -aggregate sata_01

aggr create -diskcount 23 -maxraidsize 23 -raidtype raid_tec -node node-02 -aggregate sata_02