In Simple Log Service (SLS) dashboards, SQL queries that involve large data volumes or long time spans can delay chart rendering by tens of seconds or even cause them to time out. The dashboard acceleration feature uses materialized views to automatically pre-compute data in the background. This process converts complex real-time queries into efficient result reads to enable high-speed dashboard rendering.
Enable accelerated analysis
Log on to the SLS console.
Go to the target project. In the navigation pane on the left, under Dashboard, choose Dashboards.
Click the target dashboard. On the dashboard page, click the Materialized View Acceleration button. The system automatically analyzes the execution performance of the charts in the dashboard.
Smart detection: By default, the system detects charts that support materialized view acceleration and have an SQL execution latency of more than 5 seconds.
Custom threshold: Adjust the analysis threshold based on your performance requirements and click Re-analyze.
Precise selection: The analysis results show the suggested materialized view names and their corresponding charts. Select the check box for each chart that you want to accelerate.
Click Next. Configure the parameters for the materialized view and save the configuration:
Parameter
Description
Default value
Suggestion
Acceleration start time
Specifies the start time for data backfill.
One week ago
If charts frequently involve long-term comparisons, such as year-over-year comparisons, we recommend that you set an earlier start point.
Update interval
Specifies how often to pre-compute data.
15 minutes
For very large data volumes, increase the update interval to balance compute resources.
SLS automatically creates the materialized view in the background. This process asynchronously pre-computes the original query and persists the results. Because acceleration involves processing historical data, the backfill takes some time to complete after the view is created. After the task is created and the pre-computation covers the current query range, the dashboard charts will load in seconds.