clustermode vserver for infinite volume (2)

 Create aggregates no larger than 100TB on the FAS6280 and 68TB on the FAS3270 to achieve the
maximum usable capacity for each individual constituent. Creating aggregates larger than this will
result in wasted space. Create aggregates that will be used for the namespace and data
constituents as follows (this requires the aggregates to be created using the Cluster Administration
Aggregate Create tool instead of having the Create Vserver with Infinite Volume wizard create
them):
- FAS6280: 40 3TB disks configured in 2 RAID-DP groups of 20 disks (18 data + 2 parity)
- FAS3270: 20 3TB disks configured in 1 RAID-DP group (18 data + 2 parity)

 Make sure the aggregates are created with the RAID type set to RAID-DP.

 Design the aggregates taking the namespace aggregate and the namespace data protection (DP)
mirror aggregate into consideration, and use the disks accordingly.

 Create the namespace DP mirror aggregate only after the creation of the Infinite Volume.

 Create an aggregate for the namespace DP mirror that is equal in size to the active namespace
aggregate.

 A minimum of four aggregates is required in the cluster to create an Infinite Volume; one aggregate
is required for the Vserver root volume, one for the namespace constituent, and two for the data
constituents.

 Keep the Vserver root aggregate to be the smallest possible three-disk RAID-DP aggregate and
allow the System Manager Create Vserver wizard to create this aggregate as part of the Configure
Aggregate option. You can rename this aggregate as appropriate for your Vserver.

 Before creating a Vserver for Infinite Volume, create all aggregates except for the aggregate needed
for the namespace DP mirror, or else that aggregate will be consumed during the creation of the
Infinite Volume. Verify that the aggregates used for namespace and data constituents are not larger
than 100TB on the FAS6280 and 68TB on the FAS3270.

 If creating an Infinite Volume on a FAS3270, validate that the total Infinite Volume size the wizard is
going to create does not exceed ((62.9TB * (<# aggregates in the Infinite Volume> - 1)) + 10TB). If it
does you need to manually decrease the value to no more than (62.9TB * <# data constituents>) –
10TB for a successful Infinite Volume creation.

 For better throughput, distribute the aggregates evenly across all the available nodes in the cluster