Before putting a
Data ONTAP system into a production environment, you must test your setup to ensure that everything works.
About this task
Testing for proper setup of a
Data ONTAP system
with a storage array includes the following:
- Checking your system to ensure that the configuration is as expected
- Verifying that there are two paths to storage
- Testing normal controller failover
- Testing path failover and controller failover
- Use the following commands to confirm that the results of configuration are what you expect:
|Use this command...
| disk show -v
||Check whether all array LUNs are visible.
| sysconfig -v
FC initiator ports on the Data ONTAP system, switch ports, and array LUNs are used.
| storage array show-config
||Display connectivity to back-end storage arrays.
| sysconfig -r
||Check the aggregate configuration and ensure that spare array LUNs are available.
- Ensure that there are two paths to each array LUN so that the
Data ONTAP system can continue to operate when running on a single path:
- Enter the following command:
storage array show-config
- Check whether two paths to the array LUNs are shown.
If you do not see two paths to the array LUNs, check zoning, host group configuration, and cabling.
- Look at the adapters shown to see whether all paths are on a single adapter.
If you see both paths through only one Data ONTAP system
FC initiator port (the system’s 0c port, for example) this is an indication that the back-end zoning is redundantly crossed. This is not a supported configuration.
Note: Do not continue with testing until you see two paths.
- Test normal controller failover.
- On the partner node (vs2 in the example configuration), enter the following commands to perform a cf takeover; a takeover of vs1 should occur:
vs2> cf status
vs2> cf takeover (vs1 goes down; vs2 takes over)
vs1/vs2> df (vs2 looks at vs1)
vs1/vs2> partner (to switch to the partner)
vs2 (takeover)> cf status
vs2 (takeover)> cf giveback (to return to normal operation)
- Repeat the same commands on the local Data ONTAP system (vs1 in the example configuration).
- Test path failover and cluster failover:
- On the local
Data ONTAP system (vs1 in the example configuration), enter the following commands:
vs1> fcadmin offline 0a
vs1> storage show disk -p (you should see only one path) or enter storage array show-config
vs1> fcadmin offline
vs1> fcadmin offline HA pair (takeover should occur)
- On the partner node (vs2 in the example configuration), enter the following commands:
cf takeovercf giveback
- After both the local node (vs1) and the partner node (vs2) are back online, go to the partner node (vs2 in the example) and repeat the procedure.