topic Re: 36 SSD disks, how to create aggregates in ONTAP Discussions

36 SSD disks, how to create aggregates

netappmagic — Sat, 13 Aug 2016 23:31:59 GMT

for total of 36 SSD disks ( one and half shelve) on 2 AFF HA nodes, how would you create aggregates, and how many for spares?

currently, half of them (18 ssd's) is owned by node-1, and the other half is owned by node-2

thanks!

Re: 36 SSD disks, how to create aggregates

prakash_csc — Sun, 14 Aug 2016 02:45:22 GMT

Hey ,

You can keep 1 disk as spare on each node and use the remaining disks to create your aggregate.

This would be around 34 disks for data aggregate, considering that root aggregate is already present for each node.

You can achieve this by forcing disk ownership from second node to first node leaving one spare on each node. After this, you can create aggregate using 34 disks.

Re: 36 SSD disks, how to create aggregates

heightsnj — Sun, 14 Aug 2016 03:19:47 GMT

what about root-data partitioning aggr ? how do I create that?

Re: 36 SSD disks, how to create aggregates

mikkisse — Sun, 14 Aug 2016 08:11:57 GMT

RDP creates, when you initialize your system. You need ONTAP v8.3.x+

https://library.netapp.com/ecmdocs/ECMP1636022/html/GUID-CFB9643B-36DB-4F31-95D4-29EDE648807D.html

RDP works for entry-level and All Flash FAS (AFF) platform models

Re: 36 SSD disks, how to create aggregates

Ashun — Thu, 13 Feb 2025 04:54:29 GMT

SSD RAID groups in SSD local tiers (aggregates)

You should follow these guidelines when sizing your RAID groups composed of SSDs:

All RAID groups in a local tier (aggregate) should have a similar number of drives.
The RAID groups do not have to be exactly the same size, but you should avoid having any RAID group that is less than one half the size of other RAID groups in the same local tier when possible.
For RAID-DP, the recommended range of RAID group size is between 20 and 28.

Using 36 drives Whether one aggr or two aggr is created, there are two raid groups. Each raid group has two drives as parity disks, and the number of spare drives is also taken into account.
The two methods are basically the same in terms of disk utilization, and given the load per node, I think it is a good choice to create one aggr per node