In your case, your 900G hdd is partitioned to two P1 (Larger Data Aggr) And P2 (Smaller Node Root Aggr)
this is usually done automatically during the initialization and looks good.
Data Aggr is created on node-1 using 19 of the P1 Disk Partitions to have the maximum space.
And it have one spare.
3 of the SSD is used as raid-4 (one parity and 2 for data) 4th one is spare. (there is not better way to do this..)
Here is the part I don't agree...
Node-1 root aggr is created with 16 of the P2 disk Partition, And no spares (not a recommended configuration)
Node-2 root aggr is created with 4 of the P2 Disk Partition, And no spares (not a recommended configuration)
Couple thing to notice is.. In this configuration majority of your resources on node-2 is sitting idle.
No Spare disk for root aggr.
Based on documentation, minimum root volume size is 350GB, So creating them equally on both node is the recommended practice.
Which mean assign 10 of P2 disk partitions to node-1 and Node-2 and use 9 of them on each node to create root aggr (1 spare on each node)
Creating one large data aggr or making use of your 2nd Nodes CPU resources by splitting the Aggr to two
should be based on your current and future requirements.
Its not easy and Its not possible to make those changes without re-confiuring the cluster.
I'll recommend you get it right before start using it.
Robin.
