Cloudera Docs

Tuning Impala Virtual Warehouses

This topic describes how to tune Impala Virtual Warehouses in Cloudera Data Warehouse (CDW) Private Cloud.

When you tune Impala Virtual Warehouses, you can disable the auto-suspend feature, set the minimum and maximum executor nodes allocated for the warehouse and you can set the scale up and scale down delay which determines auto-scaling behavior.

Required role: DWAdmin

  1. Log in to the CDP web interface and navigate to the Data Warehouse service.
  2. In the Data Warehouse service, navigate to the Overview page.
  3. On the Overview page under Virtual Warehouses, click the edit icon for an Impala Virtual Warehouse in the upper right corner of the tile.
  4. The next page provides properties that you can adjust to tune auto-suspend and auto-scaling for your Virtual Warehouse:
    1. Set the auto-suspend behavior:
      • If you do not want your Virtual Warehouse to auto-suspend, click Disable AutoSuspend. If you enable this feature, your Virtual Warehouse does not suspend itself, even if it is idle and no workloads are being processed.
      • If you do not want to disable auto-suspend, set the AutoSuspend Timeout. This sets the time, in seconds, that it takes for the Virtual Warehouse to automatically suspend itself. A Virtual Warehouse auto-suspends itself when the auto-scaler has scaled back to the last executor group and those executors are idle.
    2. Adjust the minimum or maximum number of executor nodes as needed:

      Setting and adjusting the minimum and maximum number of executor nodes per Virtual Warehouse is very similar to setting the number of nodes for on-premises clusters. Keep in mind the number of concurrent queries, the complexity of queries, and the volume of queries in your workloads to determine the appropriate number of executor nodes to set on each Virtual Warehouse instance.

    3. Set the Scale Up Delay and the Scale Down Delay to fine-tune when the auto-scaler starts scaling up and the number of executor groups to meet workload demand.
      • Scale Up Delay: Sets the length of time in seconds that the system waits before adding more executors when it detects queries waiting in the queue to execute. The time to auto-scale up is affected by how the underlying Kubernetes system is configured.
      • Scale Down Delay: Sets the length of time in seconds that the system waits before it removes executors when it detects idle executor groups. As with the Scale Up Delay setting, the time to auto-scale down is affected by how the underlying Kubernetes system is configured.
  5. (Optional) If you need to tune your Impala Virtual Warehouse to run more queries per executor group, select Use Legacy Multithreading Mode.
  6. Click Apply in the upper right of the page to save your changes.