All Products
Search

This Product

    ApsaraMQ for Kafka:Resolve topic partition skew

    Document Center

    ApsaraMQ for Kafka:Resolve topic partition skew

    Last Updated:Sep 21, 2026

    Symptoms

    • Partition skew causes high disk usage on some disks, which prevents full utilization of all disk performance and capacity.

    • Partition skew can cause excessive traffic on some nodes, triggering single-node throttling.

    Possible cause

    The partition count was not set to a recommended multiple during topic creation.

    Solution

    1. Log on to the ApsaraMQ for Kafka console. In the navigation pane on the left, click Instances.

    2. In the top menu bar, select a region. On the Instances page, click the name of the target instance.

    3. In the navigation pane on the left, click Topics, and then click the name of the target topic.

    4. On the Topic Details page:

      Non-serverless instances

      1. Click the Configuration Information tab. Next to Partitions, click Increase Partitions.

      2. In the Increase Partitions panel, use the up and down arrows for Partitions to set the partition count to a multiple of the recommended number.

      Note

      If you cannot add partitions due to business requirements and your topic uses cloud storage, use the partition rebalancing feature to adjust partition placement. After rebalancing, new data is written to disks with lower usage. The original data is deleted only after its messages expire, at which point disk usage decreases. For more information, see Partition rebalancing and traffic redirection.

      Serverless instances

      1. Next to Partition Replicas, click Edit.

      2. In the Edit Partition panel, use the arrows for Partition to set the value to a multiple of the recommended number.