ONTAP Hardware

few questions regarding aggreagates


Hi everyone,

I am expecting implementation of fas 3160A and 4 disk shelves ds4234, each with 600GB sas HDD, only block access (FC SAN)

System will be configured with ontap 8.0.2

My idea is to assign 2 shelves to filer1 and 2 shelves to filer2

Then i will have 48-3=45 free disks for data on each filer.

First idea is to create 3 aggregates per filer, existing of 15 disks per aggregate (6 aggreagtes total)

But since this is ontap 8.0.2 with support for larger aggregates, i have some ideas

First, on website i checked that 3160 supports maximum 28 disks per aggregate (http://www.netapp.com/us/products/storage-systems/fas3100/fas3100-tech-specs.html). Is this true also with ontap 8.x?

From the link i cannot see ontap version, can i assign more disks per aggregate with 8.x version on this model? And if i can, what do you think of an idea to create 2 large aggregates (one per filer)? I would get more space, but will this overload CPU's and cache on filers?

So i need some advices from your experience : 3x15 disks aggregates per filer, or larger aggregates (ofc if supported - lets say 45 disks aggregate with 2 spares)

Thank you



one more thing :

from "storage system tehnical faq" guide i have found table as attached

So i think that optimal solution would be to create 1 aggregate per filler consisting of 45 (3 disks are for root vol) disks

in raid dp. Is raid group size of 15 good for this?

Is this OK overall?

You've almost answered your own question!

There is no need for a dedicated root aggregate in my opinion. You do need, however, spare disks, ideally 2 per controller.

So you are good to go with:

- 45 disks, RG size 15, 3 hot spares per controller


- 46 disks, RG size 16, 2 hot spares per controller



Actually with 32-bit aggregates (pre-ONTAP 8 & ONTAP 8.x as well), you can have up to 33x 600GB drives in one aggregate. With 64-bit aggregates (ONTAP 8.x), you can have up to 105x 600GB drives in a single aggregate on FAS3160.

This doc explains a lot of detail around RAID group sizes, aggregate limits, spares policies, etc:




thank you

46 disks size, rg size 16, that will be it