Skip to main content
Upgrade controllers

Verify the node4 installation

Contributors netapp-pcarriga netapp-aoife netapp-martyh

You must verify that the physical ports from node2 map correctly to the physical ports on node4. This will enable node4 to communicate with other nodes in the cluster and with the network after the upgrade.

About this task

Refer to References to link to the Hardware Universe to capture information about the ports on the new nodes. You will use the information later in this section.

Physical port layout might vary, depending on the model of the nodes. When the new node boots up, ONTAP will try to determine which ports should host cluster LIFs in order to automatically come into quorum.

If the physical ports on node2 do not map directly to the physical ports on node4, the subsequent section Restore network configuration on node4 must be used to repair network connectivity.

After you install and boot node4, you must verify that it is installed correctly. You must wait for node4 to join quorum and then resume the relocation operation.

At this point in the procedure, the operation will have paused as node4 joins quorum.

  1. Verify that node4 has joined quorum:

    cluster show -node node4 -fields health

    The output of the health field should be true.

  2. Verify that node4 is part of the same cluster as node3 and that it is healthy:

    cluster show

  3. Depending on the ONTAP version running on the HA pair being upgraded, take one of the following actions:

    If your ONTAP version is…​ Then…​

    9.8 to 9.11.1

    Verify that the cluster LIFs are listening on port 7700:

    ::> network connections listening show -vserver Cluster

    9.12.1 or later

    Skip this step and go to Step 5.

    Port 7700 listening on cluster ports is the expected outcome as shown in the following example for a two-node cluster:

    Cluster::> network connections listening show -vserver Cluster
    Vserver Name     Interface Name:Local Port     Protocol/Service
    ---------------- ----------------------------  -------------------
    Node: NodeA
    Cluster          NodeA_clus1:7700               TCP/ctlopcp
    Cluster          NodeA_clus2:7700               TCP/ctlopcp
    Node: NodeB
    Cluster          NodeB_clus1:7700               TCP/ctlopcp
    Cluster          NodeB_clus2:7700               TCP/ctlopcp
    4 entries were displayed.
  4. For each cluster LIF that is not listening on port 7700, set the administrative status of the LIF to down and then up:

    ::> net int modify -vserver Cluster -lif cluster-lif -status-admin down; net int modify -vserver Cluster -lif cluster-lif -status-admin up

    Repeat Step 3 to verify that the cluster LIF is now listening on port 7700.

  5. Switch to advanced privilege mode:

    set advanced

  6. Check the status of the controller replacement operation and verify that it is in a paused state and in the same state it was in before node2 was halted to perform the physical tasks of installing new controllers and moving cables:

    system controller replace show

    system controller replace show-details

  7. If you are working on a MetroCluster system, verify that the replaced controller is configured correctly for the MetroCluster configuration; the MetroCluster configuration should be in a healthy state. Refer to Verify the health of the MetroCluster configuration.

    Reconfigure the intercluster LIFs on MetroCluster node node4, and check cluster peering to restore communication between the MetroCluster nodes before proceeding to Step 6.

    Check the MetroCluster node status:

    metrocluster node show

  8. Resume the controller replacement operation:

    system controller replace resume

  9. Controller replacement will pause for intervention with the following message:

    Cluster::*> system controller replace show
    Node             Status                       Error-Action
    ---------------- ------------------------     ------------------------------------
    Node2(now node4) Paused-for-intervention      Follow the instructions given in
                                                  Step Details
    Step Details:
    To complete the Network Reachability task, the ONTAP network configuration must be
    manually adjusted to match the new physical network configuration of the hardware.
    This includes:
    1. Re-create the interface group, if needed, before restoring VLANs. For detailed
    commands and instructions, refer to the "Re-creating VLANs, ifgrps, and broadcast
    domains" section of the upgrade controller hardware guide for the ONTAP version
    running on the new controllers.
    2. Run the command "cluster controller-replacement network displaced-vlans show"
    to check if any VLAN is displaced.
    3. If any VLAN is displaced, run the command "cluster controller-replacement
    network displaced-vlans restore" to restore the VLAN on the desired port.
    2 entries were displayed.
    Note In this procedure, section Re-creating VLANs, ifgrps, and broadcast domains has been renamed Restoring network configuration on node4.
  10. With the controller replacement in a paused state, proceed to the next section of this document to restore network configuration on the node.

Restore network configuration on node4

After you confirm that node4 is in quorum and can communicate with node3, verify that node2’s VLANs, interface groups and broadcast domains are seen on node4. Also, verify that all node4 network ports are configured in their correct broadcast domains.

About this task

For more information on creating and re-creating VLANs, interface groups, and broadcast domains, refer to References to link to Network Management.

Note If you are changing the port speed of the e0a and e1a cluster ports on AFF A800 or AFF C800 systems, you might observe malformed packets being received after the speed conversion. See NetApp Bugs Online Bug ID 1570339 and the knowledge base article CRC errors on T6 ports after converting from 40GbE to 100GbE for guidance.
  1. List all the physical ports that are on upgraded node2 (referred to as node4):

    network port show -node node4

    All physical network ports, VLAN ports and interface group ports on the node are displayed. From this output you can see any physical ports that have been moved into the Cluster broadcast domain by ONTAP. You can use this output to aid in deciding which ports should be used as interface group member ports, VLAN base ports or standalone physical ports for hosting LIFs.

  2. List the broadcast domains on the cluster:

    network port broadcast-domain show

  3. List the network port reachability of all ports on node4:

    network port reachability show

    The output from the command looks similar to the following example:

    clusterA::*> reachability show -node node2_node4
      (network port reachability show)
    Node         Port       Expected Reachability       Reachability Status
    ---------    --------  ---------------------------  ---------------------
                 a0a        Default:Default             no-reachability
                 a0a-822    Default:822                 no-reachability
                 a0a-823    Default:823                 no-reachability
                 e0M        Default:Mgmt                ok
                 e0a        Cluster:Cluster             misconfigured-reachability
                 e0b        Cluster:Cluster             no-reachability
                 e0c        Cluster:Cluster             no-reachability
                 e0d        Cluster:Cluster             no-reachability
                 e0e        Cluster:Cluster             ok
                 e0e-822    -                           no-reachability
                 e0e-823    -                           no-reachability
                 e0f        Default:Default             no-reachability
                 e0f-822    Default:822                 no-reachability
                 e0f-823    Default:823                 no-reachability
                 e0g        Default:Default             misconfigured-reachability
                 e0h        Default:Default             ok
                 e0h-822    Default:822                 ok
                 e0h-823    Default:823                 ok
    18 entries were displayed.

    In the above example, node2_node4 is just booted after controller replacement. It has several ports that have no reachability and are pending a reachability scan.

  4. Repair the reachability for each of the ports on node4 with a reachability status other than ok. Run the following command, first on any physical ports, then on any VLAN ports, one at a time:

    network port reachability repair -node node_name -port port_name

    The output looks like the following example:

    Cluster ::> reachability repair -node node2_node4 -port e0h
    Warning: Repairing port "node2_node4: e0h" may cause it to move into a different broadcast domain, which can cause LIFs to be re-homed away from the port. Are you sure you want to continue? {y|n}:

    A warning message, as shown above, is expected for ports with a reachability status that might be different from the reachability status of the broadcast domain where it is currently located.

    Review the connectivity of the port and answer y or n as appropriate.

    Verify that all physical ports have their expected reachability:

    network port reachability show

    As the reachability repair is performed, ONTAP attempts to place the ports in the correct broadcast domains. However, if a port’s reachability cannot be determined and does not belong to any of the existing broadcast domains, ONTAP will create new broadcast domains for these ports.

  5. If interface group configuration does not match the new controller physical port layout, modify it by using the following steps.

    1. You must first remove physical ports that should be interface group member ports from their broadcast domain membership. You can do this by using the following command:

      network port broadcast-domain remove-ports -broadcast-domain broadcast_domain_name -ports node_name:port_name

    2. Add a member port to an interface group:

      network port ifgrp add-port -node node_name -ifgrp ifgrp -port port_name

    3. The interface group is automatically added to the broadcast domain about a minute after the first member port is added.

    4. Verify that the interface group was added to the appropriate broadcast domain:

      network port reachability show -node node_name -port ifgrp

      If the interface group’s reachability status is not ok, assign it to the appropriate broadcast domain:

      network port broadcast-domain add-ports -broadcast-domain broadcast_domain_name -ports node:port

  6. Assign appropriate physical ports to the Cluster broadcast domain:

    1. Determine which ports have reachability to the Cluster broadcast domain:

      network port reachability show -reachable-broadcast-domains Cluster:Cluster

    2. Repair any port with reachability to the Cluster broadcast domain, if its reachability status is not ok:

      network port reachability repair -node node_name -port port_name

  7. Move the remaining physical ports into their correct broadcast domains by using one of the following commands:

    network port reachability repair -node node_name -port port_name

    network port broadcast-domain remove-port

    network port broadcast-domain add-port

    Verify that there are no unreachable or unexpected ports present. Check the reachability status for all physical ports by using the following command and examining the output to confirm the status is ok:

    network port reachability show -detail

  8. Restore any VLANs that might have become displaced by using the following steps:

    1. List displaced VLANs:

      cluster controller-replacement network displaced-vlans show

      Output like the following should display:

      Cluster::*> displaced-vlans show
      (cluster controller-replacement network displaced-vlans show)
      Node        Base Port     VLANs
      ---------   ---------     ------------------------------------------------------
      Node1       a0a           822, 823
                  e0e           822, 823
    2. Restore VLANs that were displaced from their previous base ports:

      cluster controller-replacement network displaced-vlans restore

      The following is an example of restoring VLANs that have been displaced from interface group a0a back onto the same interface group:

      Cluster::*> displaced-vlans restore -node node2_node4 -port a0a -destination-port a0a

      The following is an example of restoring displaced VLANs on port "e0e" to "e0h":

      Cluster::*> displaced-vlans restore -node node2_node4 -port e0e -destination-port e0h

      When a VLAN restore is successful, the displaced VLANs are created on the specified destination port. The VLAN restore fails if the destination port is a member of an interface group, or if the destination port is down.

      Wait about one minute for newly restored VLANs to be placed into their appropriate broadcast domains.

    3. Create new VLAN ports as needed for VLAN ports that are not in the cluster controller-replacement network displaced-vlans show output but should be configured on other physical ports.

  9. Delete any empty broadcast domains after all port repairs have been completed:

    network port broadcast-domain delete -broadcast-domain broadcast_domain_name

  10. Verify port reachability:

    network port reachability show

    When all ports are correctly configured and added to the correct broadcast domains, the network port reachability show command should report the reachability status as ok for all connected ports, and the status as no-reachability for ports with no physical connectivity. If any ports report a status other than these two, perform the reachability repair and add or remove ports from their broadcast domains as instructed in Step 4.

  11. Verify that all ports have been placed into broadcast domains:

    network port show

  12. Verify that all ports in the broadcast domains have the correct maximum transmission unit (MTU) configured:

    network port broadcast-domain show

  13. Restore LIF home ports, specifying the Vserver(s) and LIF(s) home ports, if any, that need to be restored:

    1. List any LIFs that are displaced:

      displaced-interface show

    2. Restore LIF home ports:

      displaced-interface restore-home-node -node node_name -vserver vserver_name -lif-name LIF_name

  14. Verify that all LIFs have a home port and are administratively up:

    network interface show -fields home-port, status-admin