Step 4: Completing setup and testing your mirrored cluster

Apply power to your disk shelves, then to your filers.

At the prompt, enter the following command:

vol create vol-name -m ndisks[@disk-size]

vol-name is the name for the new volume (without the /vol/ prefix).

ndisks is the number of disks to use; ndisks must be at least 4 and in even increments. Even increments are required because half of the disks specified by ndisks go to one plex from one disk pool and half go to the other plex from another disk pool.

disk-size is a number that specifies the disk size to use, in gigabytes. You must have half of the ndisks available disks of the size you specify in one disk pool and half of the ndisks available disks of the size you specify in another disk pool if you use this option.

Enable the syncmirror_local and clustering licenses by entering the following commands:

license syncmirror_local=code

license cluster=code

If your cluster is...

Then...

Separated by less than 30 meters

1. Reboot the local node, then the partner node.
2. Complete the configuration, as described in Configuring the cluster.

Separated by more than 30 meters and less than 500 meters

1. Reboot the local node, then the partner node.
2. Enable the cluster_remote license by entering the following command:

   license cluster_remote=code

3. Complete the configuration, as described in Configuring the cluster.

Verify the network addresses by entering the following command from one filer:

ifconfig -a

Result

The local node and partner node IP addresses are listed.

Verify that you can create and retrieve files on both filers for each licensed protocol.

From the same filer, enter the following command:

cf enable

Result

Takeover of the partner node is enabled.

Test communication between the local node and partner node.

For example, you can use the fcadmin device_map command to ensure that one filer can access the other filer's disks.

Example:

filer-1> fcadmin device_map

Loop Map for channel 8: 

Translated Map: Port Count 32

                  7  16  17  18  19  32  33  34  35  48  
49  50  51  24  26  27 57  61  42  43  44  45  28  29  30  
52  53  54  37  38  21  22 

Shelf mapping: 

                Shelf 2:  22  21 XXX  19  18  17  16 

.

.

.

From the local node console, enter the following command and verify that the partner node halted:

cf takeover

Result

The local node takes over the partner node. You see a message similar to the following:

takeover completed

Verify that the correct volumes are mirrored by entering the following command:

vol status -r

Check the parameters and options on your cluster with the Cluster Configuration Checker script.

The script is available through download from the NOW Web site at http://now.netapp.com/NOW/tools. Look for the Cluster Configuration Checker (cf-config-check.cgi) file.

Correct any errors you find.

From the same filer, enter the following command:

cf giveback

Result

The partner node reboots and functions normally. A successful giveback ends with the following message on the local node:

giveback completed

Repeat Steps 1 through 8 on the other filer.