Responding to cluster imbalance performance events

Unified Manager generates cluster imbalance warning events when one node in a cluster is operating at a much higher load than other nodes, and therefore potentially affecting workload latencies. These system-defined events provide the opportunity to correct potential performance issues before many workloads are affected by latency.

Before you begin

You must have the Operator, Application Administrator, or Storage Administrator role.

About this task

Unified Manager generates warning events for cluster imbalance threshold policy breaches by comparing the performance capacity used value for all nodes in the cluster to see if there is a load difference of 30% between any nodes.

These steps help you identify the following resources so that you can move high-performing workloads to a lower utilized node:

Steps

  1. Display the Event details page to view information about the event.
  2. Review the Description, which describes the threshold breach that caused the event.
    For example, the message "The performance capacity used counter indicates a load difference of 62% between the nodes on cluster Dallas-1-8 and has triggered a WARNING event based on the system threshold of 30%" indicates that performance capacity on one of the nodes is overused and affecting node performance.
  3. Review the text in the Suggested Actions to move a high-performing volume from the node with the high performance capacity used value to a node with the lowest performance capacity used value.
  4. Identify the nodes with the highest and lowest performance capacity used value:
    1. In the Event Information section, click the name of the source cluster.
    2. In the Cluster / Performance Summary page, click Nodes in the Managed Objects area.
    3. In the Nodes inventory page, sort the nodes by the Performance Capacity Used column.
    4. Identify the nodes with the highest and lowest performance capacity used value and write down those names.
  5. Identify the volume using the most IOPS on the node that has the highest performance capacity used value:
    1. Click the node with the highest performance capacity used value.
    2. In the Node / Performance Explorer page, select Aggregates on this Node from the View and Compare menu.
    3. Click the aggregate with the highest performance capacity used value.
    4. In the Aggregate / Performance Explorer page, select Volumes on this Aggregate from the View and Compare menu.
    5. Sort the volumes by the IOPS column, and write down the name of the volume using the most IOPS, and the name of the aggregate where the volume resides.
  6. Identify the aggregate with the lowest utilization on the node that has the lowest performance capacity used value:
    1. Click Storage > Aggregates to display the Aggregates inventory page.
    2. Select the Performance: All Aggregates view.
    3. Click the Filter button and add a filter where "Node" equals the name of the node with the lowest performance capacity used value that you wrote down in step 4.
    4. Write down the name of the aggregate that has the lowest performance capacity used value.
  7. Move the volume from the overloaded node to the aggregate you identified as having low utilization on the new node.
    You can perform the move operation by using ONTAP System Manager, OnCommand Workflow Automation, ONTAP commands, or a combination of these tools.

After you finish

After a few days, check to see whether you are receiving the same cluster imbalance event from this cluster.