All Products
Search
Document Center

Hologres:Query and analyze historical slow query logs

Last Updated:Oct 15, 2024

Hologres V0.10 and later allow you to query and analyze slow query logs to diagnose and analyze slow queries or failed queries, and take optimization measures for such queries. This topic describes how to query and analyze historical slow query logs in a visualized manner in Hologres.

Limits

  • This feature is supported only by Hologres V0.10 and later. You can view the version of your Hologres instance on the instance details page in the Hologres console. If the version of your Hologres instance is earlier than V0.10, manually upgrade your Hologres instance in the Hologres console or join a Hologres DingTalk group to apply for an instance upgrade. For more information about how to manually upgrade a Hologres instance, see Instance upgrades. For more information about how to join a Hologres DingTalk group, see Obtain online support for Hologres.

  • Up to 2,000 slow query logs can be returned for each query request to ensure system stability and prevent traffic overloads.

  • HoloWeb allows you to query slow query logs within seven days.

Query slow query logs

  1. Log on to the Hologres console. In the left-side navigation pane, click Go to HoloWeb. In the top navigation bar of the HoloWeb console, click Diagnostics and Optimization.

  2. In the left-side navigation pane, click Historical Slow Queries.

  3. On the Historical Slow Query page, configure the parameters that are described in the following table.

    ma1

    Parameter

    Required

    Description

    Instance Name

    Yes

    The name of the instance whose slow queries you want to query. By default, the connected instance is selected.

    Database

    No

    The name of the database whose slow queries you want to query. You can query slow query logs of the specified database only if you have the required permissions on the database. Otherwise, you can query slow query logs only within your Alibaba Cloud account.

    Table Name

    No

    The name of the table for which you want to query slow query logs. You can query slow query logs of the specified table only if you have the required permissions on the table. Otherwise, you can query slow query logs only within your Alibaba Cloud account.

    User

    No

    The Alibaba Cloud account within which you want to query slow query logs.

    Queried Rows

    Yes

    The number of slow query logs to return. A maximum of 2,000 slow query logs can be returned each time.

    Duration

    No

    The execution duration of the query. By default, Hologres collects the logs of queries that consume more than 1 second.

    Metrics

    No

    The dimension from which the query trend is displayed. Valid values: Slow Query and Failed Query.

    Query

    No

    The statement executed in the query. You can use the percent sign (%) in the statement to represent all table names.

    Type

    No

    The type of the SQL statement executed in the query. Statement types such as DDL and DML are supported.

    Query ID

    No

    The ID of the query.

    Virtual Warehouse

    Yes

    The computing resources of the query.

    Time Range

    Yes

    The time range to query. Default value: 10 Minutes. The maximum time range in which slow query logs can be queried is seven days.

  4. After the configuration is complete, click Query to view information in the Query Trend Analysis and Queries sections. For more information, see the View the trend analysis and details of slow queries and failed queries section in this topic.

View the trend analysis and details of slow queries and failed queries

  • Query Trend Analysis

    • View the trend of slow queries and failed queries before and after a specific point in time.

      Click a specific point in time in the trend chart to view the trend of slow queries and failed queries within 10 seconds before and after the point in time.

      单点峰值

    • View the trend of slow queries and failed queries within a time range.

      Select a time range in the trend chart to view the trend of slow queries and failed queries within the selected time range.

      查看区域峰值

  • Queries

    • View the slow query logs that meet the filter conditions.

      The following table describes the fields that are selected by default.

      ma2

      Parameter

      Description

      Query ID

      The ID of the query.

      Database

      The name of the database for which data was queried.

      User Name

      The name of the user who executes the SQL statement.

      Type

      The type of the SQL statement executed in the query.

      Duration

      The total execution duration of the query, including the optimization duration, the startup duration of the query, and the duration for returning the query result.

      Start Time

      The start time of the query.

      Status

      The status of the query.

      Query

      The SQL statement executed in the query. You can click the SQL statement to view its details.

      Engine Type

      The type of the query engine that is used to process the current SQL query.

      Query Plan

      The execution plan of the query.

      Actions

      • Details: Click it to go to the details page of the query. You can view the details of the query on this page. The details can help you optimize the SQL statement.

      • SQL编辑器: Click it to open a new SQL editor for the query.

      • image: Click it to go to the Query Insight page.

    • View other fields.

      If the default fields do not meet your requirements, you can select parameters based on your business requirements by clicking Customize Columns in the upper-right corner of the Queries section. In the dialog box that appears, select more parameters to display.

      自定义列表

    • View the Gantt chart.

      HoloWeb supports Gantt charts. To view the Gantt chart of queries, you can click View Gantt Chart in the upper-right corner of the Queries section. In the dialog box that appears, you can view the execution status of queries that are concurrently executed within a specified time range. This helps you identify performance bottlenecks when queries are concurrently executed.甘特图