Summary: Keep this in mind when designing storage on a Netapp Filer.
Date: Around 2015
Refactor: 7 March 2025: Checked links and formatting.
When configuring a NetApp and designing the aggregate size there is something really important to keep in mind. So bear with me while I write it down. In a NetApp filer are disks. Disks themselves are point of failures, and that's why we use RAID. In this story I'll keep to raid_dp, which stand for raid double parity. This means that in each raid set that is configured two disks can fail before you're screwed. However, since NetApps work with WAFL (write anywhere file layout) you as a administrator don't work with disks and RAID, you work with aggregates and volumes. Also consider, that if you add disks to an aggregate they are automatically added to the RAID until it is full and a new RAID set will be created. Now a raid_dp set can contain up to 28 disks, but the default is 16. That means that if you have 20 disks, and don't change the default two RAID sets will be created, one containing 16 disks, and one containing 4 disks. That means two things: First, you loose 4 disks to parity, two to spare and you only keep 14 disks left for actual storage. Second, large RAID sets operate faster than small raid sets. That is because more disks can write at the same time in a large raid set (sales will tell you they're both evenly fast but ask any netapp engineer and he'll tell you the same.). So this meams you get an aggregate that's slow, since it will be as fast as the slowest RAID set. Than a final note, keep in mind the maximum size for an aggregate on a NetApp ONTAP 7.3 filer is 16 TB. So keep all this in mind when you're designing your aggregates.
note: You cannot remove disks from an aggregate! This is very important, so grow your aggregates with caution, it's easy to grow, impossible to shrink. If you've grown your aggregate too much then you'll need to destroy it to regain those spares.