This topic describes how to optimize dataset performance by configuring the cache duration, using the built-in cache of a slow database, and enabling full table scan.

Prerequisites

A dataset is created. For more information, see Create a dataset.

Configure the cache duration for a dataset

After caching is enabled, you can set the cache duration to 30 minutes or 1 hour.
  • After the cache duration elapses, the cached data becomes invalid, and queries on the report page generate new cache. Subsequent SQL statements that are executed within the cache duration will obtain data from the new cache.
  • Cache is cleared by dataset. When you manually trigger cache clearing or after the cache duration elapses, the cache of all charts generated based on the dataset is cleared.
  1. Log on to the Quick BI console.
  2. Click the Workspace tab. In the left-side navigation pane, click Datasets.
  3. On the Datasets page, find the target dataset, click the More icon in the Actions column, and select Cache Configuration.
    You can also right-click the target dataset and select Cache Configuration.
  4. In the Cache Configuration dialog box, select Enable Caching, set Cache Duration, and click OK.
    Enable Caching is selected by default.Cache Configuration

Use the built-in cache of a slow database

For slow databases such as MaxCompute (formally ODPS) and Hive, a built-in cache is provided to optimize dataset performance.

Enable full table scan

We recommend that you enable full table scan to query data from datasets that are created based on a MaxCompute partitioned table.

  1. Log on to the Quick BI console.
  2. Click the Workspace tab. In the left-side navigation pane, click Datasets.
  3. On the Datasets page, find the target dataset, click the More icon in the Actions column, and select Edit Properties.
    You can also right-click the target dataset and select Edit Properties.
  4. In the Edit Properties dialog box, select Scan Full Table to enable full table scan.
  5. Click Save.
  6. Optional:If full table scan is slow, you can use field-based filtering or SQL statement-based filtering to speed up the scan.
    • Field-based filtering
      1. On the Datasets page, find the target dataset and click the edit icon in the Actions column.
      2. On the dataset edit page, click Set Filter in the top menu bar.
      3. In the Set Filter Fields dialog box, add a filter field.
    • SQL statement-based filtering