Intelligent Distribution Column Recommendation

Context

Intelligent distribution column recommendation is crucial in distributed database systems. It optimizes data storage and query performance by analyzing past query logs and suggesting the best distribution column for the data table. This results in efficient data sharding and storage, ultimately improving query performance.

If the current console does not support this feature, contact technical support.

Prerequisites

To activate this feature, enable Historical Query Monitoring on the Monitoring Settings > Monitoring Collection page. Once enabled, the system will generate historical query logs after a period of query operations. For details, see Monitoring Collection.

Creating an Inference Task

1. Log in to the GaussDB(DWS) console.
2. On the Clusters > Dedicated Clusters page, locate the cluster to be monitored.
3. In the Operation column of the target cluster, click Monitoring Panel.
4. In the navigation tree, choose Tools > Intelligent Distribution Column Recommendation.
5. To receive column distribution recommendations, click Create Inference Task and choose the cluster, database, and time segment you want to analyze.
6. Click OK.

Viewing the Inference Result

1. Log in to the GaussDB(DWS) console.
2. On the Clusters > Dedicated Clusters page, locate the cluster to be monitored.
3. In the Operation column of the target cluster, click Monitoring Panel.
4. In the navigation tree, choose Tools > Intelligent Distribution Column Recommendation.
5. View the recommendation job list. The following table describes the related parameters.

   Table 1 Parameters in the intelligent distribution column recommendation job list

   Parameter	Description
   Task ID	ID generated after a column distribution recommendation job is created.
   Database Name	Name of the database.
   Log Start Time	Start time for historical query logs used in data analysis. The default value is one week prior to the current time.
   Log End Time	End time for historical query logs used in data analysis. The default value is the current time.
   Status	Real-time status of a recommendation job based on the current column distribution recommendation job. Created: The creation is successful. Entities extracted: The log data is being processed. Plans processed: The recommendation result is being generated. Finish: The recommendation job has been completed. Failure: The recommendation job fails to be executed.
   Operation	Click Show Inference Result to view the column distribution recommendation. Click Delete to delete the record.