Verifying disk assignment in Maintenance mode in a two-node configuration

Before fully booting the system to ONTAP, you can optionally boot the system to Maintenance mode and verify the disk assignment on the nodes. The disks should be assigned to create a fully symmetric configuration with both sites owning their own disk shelves and serving data, where each node and each pool have an equal number of mirrored disks assigned to them.

Before you begin

The system must be in Maintenance mode.

About this task

New MetroCluster systems have disk assignment completed prior to shipment.

The following table shows example pool assignments for a MetroCluster configuration. Disks are assigned to pools on a per-shelf basis.

Disk shelf (example name)... At site... Belongs to... And is assigned to that node's...
Disk shelf 1 (shelf_A_1_1) Site A Node A 1 Pool 0
Disk shelf 2 (shelf_A_1_3)
Disk shelf 3 (shelf_B_1_1) Node B 1 Pool 1
Disk shelf 4 (shelf_B_1_3)
Disk shelf 9 (shelf_B_1_2) Site B Node B 1 Pool 0
Disk shelf 10 (shelf_B_1_4)
Disk shelf 11 (shelf_A_1_2) Node A 1 Pool 1
Disk shelf 12 (shelf_A_1_4)

If your configuration includes DS460C disk shelves, you should manually assign the disks using the following guidelines for each 12-disk drawer:

Assign these disks in the drawer... To this node and pool...
1 - 6 Local node's pool 0
7 - 12 DR partner's pool 1

This disk assignment pattern minimizes the effect on an aggregate if a drawer goes offline.

Procedure

  1. If your system was received from the factory, confirm the shelf assignments: disk show –v
  2. If necessary, you can explicitly assign disks on the attached disk shelves to the appropriate pool by using the disk assign command.
    Disk shelves at the same site as the node are assigned to pool 0 and disk shelves located at the partner site are assigned to pool 1. You should assign an equal number of shelves to each pool.
    1. If you have not done so, boot each system into Maintenance mode.
    2. On the node on site A, systematically assign the local disk shelves to pool 0 and the remote disk shelves to pool 1: disk assign -shelf disk_shelf_name -p pool
      If storage controller node_A_1 has four shelves, you issue the following commands:
      
      *> disk assign -shelf shelf_A_1_1 -p 0
      *> disk assign -shelf shelf_A_1_3 -p 0
      
      *> disk assign -shelf shelf_A_1_2 -p 1
      *> disk assign -shelf shelf_A_1_4 -p 1
    3. On the node at the remote site (site B), systematically assign its local disk shelves to pool 0 and its remote disk shelves to pool 1: disk assign -shelf disk_shelf_name -p pool
      If storage controller node_B_1 has four shelves, you issue the following commands:
      
      *> disk assign -shelf shelf_B_1_2   -p 0 
      *> disk assign -shelf shelf_B_1_4  -p 0 
      
      *> disk assign -shelf shelf_B_1_1 -p 1
       *> disk assign -shelf shelf_B_1_3 -p 1
    4. Show the disk shelf IDs and bays for each disk: disk show –v