Skip to main content

Guidelines for upgrading controllers with ARL

Contributors netapp-pcarriga netapp-aoife netapp-aherbin
PDFs

To understand whether you can use aggregate relocation (ARL) to upgrade a pair of controllers running ONTAP 9.5 to ONTAP 9.7 depends on the platform and the configuration of both the original and replacement controllers.

Supported upgrades for ARL

When you upgrade a pair of nodes using this ARL procedure for ONTAP 9.5 to ONTAP 9.7, you must verify that ARL can be performed on the original and replacement controllers.

You should check the size of all defined aggregates and number of disks supported by the original system. You must then compare the aggregate sizes and number of disks supported to the aggregate size and number of disks supported by the new system. Refer to References to link to the Hardware Universe where this information is available. The aggregate size and the number of disks supported by the new system must be equal to or greater than the aggregate size and number of disks supported by the original system.

You should validate in the cluster mixing rules whether new nodes can become part of the cluster with the existing nodes, when the original controller is replaced. For more information about cluster mixing rules, refer to References to link to the Hardware Universe.

Note Before performing an AFF system upgrade, you must upgrade ONTAP to release versions 9.5P1 or later. These release levels are required for a successful upgrade.
Note If you are upgrading a system that supports internal drives (for example, an FAS2700 or AFF A250) but does NOT have internal drives, refer to References and use the procedure in the Aggregate Relocation to Manually Upgrade Controller Hardware content that is correct for your version of ONTAP.

If you are using ONTAP 9.6P11, 9.7P8, or later releases, it is recommended to enable Connectivity, Liveliness, and Availability Monitor (CLAM) takeover to return the cluster into quorum when certain node failures occur. The kernel-service command requires advanced privilege level access. For more information, see: NetApp KB Article SU436: CLAM takeover default configuration changed.

Controller upgrade using ARL is supported on systems configured with SnapLock Enterprise andSnapLock Compliance volumes.

Two-node switchless clusters

If you are upgrading nodes in a two-node switchless cluster, you can leave the nodes in the switchless cluster while performing the upgrade. You do not need to convert them to a switched cluster.

Upgrades not supported for ARL

You cannot perform the following upgrades:

  • To replacement controllers that do not support the disk shelves connected to the original controllers

    Refer to References to link to the Hardware Universe for disk-support information.

  • To entry level controllers with internal drives, for example: an FAS 2500.

    If you want to upgrade entry level controllers with internal drives, refer to References to link to Upgrade by moving volumes or storage and go to the procedure Upgrading a pair of nodes running clustered Data ONTAP by moving volumes.

MetroCluster FC configuration

In a MetroCluster FC configuration, you must replace the disaster recovery/failover site nodes as soon as possible. Mismatch in controller models within a MetroCluster configuration isn't supported because controller model mismatch can cause disaster recovery mirroring to go offline. Use the -skip-metrocluster-check true command to bypass MetroCluster checks when you're replacing nodes at the second site.

Troubleshoot

You might encounter a failure while upgrading the node pair. The node might crash, aggregates might not relocate, or LIFs might not migrate. The cause of the failure and its solution depend on when the failure occurred during the upgrade procedure.

Refer to the table describing the different phases of the procedure in the section Overview of the ARL upgrade. Information about the failures that can occur is listed by the phase of the procedure.

If any problems occur while upgrading the controllers, refer to the Troubleshoot section. The information about failures that can occur is listed by the phase of the procedure in the ARL upgrade sequence.

If you do not find a solution to the problem you encountered, contact technical support.