You can use Database Autonomy Service (DAS) for the daily operations and maintenance (O&M) of ApsaraDB RDS for MySQL to handle database failures, optimize performance, increase efficiency, and reduce O&M costs. DAS is a cloud service based on machine learning and expert experience. It provides self-perception, self-healing, self-optimization, self-O&M, and self-protection features to simplify database O&M tasks and ensure that your database services are stable, secure, and efficient.
Features
The RDS MySQL autonomy service includes the following features:
-
You can diagnose your instance and view visualized results.
-
You can view anomalies, optimization events, auto scaling events, and other events that occurred within a specified time range.
-
You can view, export, and terminate sessions on a target instance. You can also perform 10-second SQL analysis, SQL throttling, and optimization on the target instance.
-
You can view real-time performance trends and data for database instance metrics.
-
You can view and analyze the most recent deadlock in the database.
-
You can view capacity assessment suggestions, performance capacity status, storage capacity usage, and the available days of storage. This feature also uses machine learning and capacity algorithms to predict future storage usage.
-
You can view the storage overview, available days of storage, storage usage of specific tables, storage fragmentation, and diagnoses of storage anomalies for your database instance. This helps you promptly detect storage anomalies and ensure database stability.
-
You can diagnose the parameters of your database instance and view static parameter diagnostic results and the parameter modification history from the last seven days.
-
Performance Insight (New Version)
This feature uses the Performance Schema of MySQL to collect SQL information. It assesses the load on the target database, identifies the root causes of performance issues, and provides governance solutions to ensure database stability.
-
Performance Insight (Old Version)
This is a powerful tool for load monitoring, analysis, and performance tuning. It helps you quickly assess the database load in a simple and intuitive way, identify the root causes of performance issues, and ensure database stability.
-
This feature provides a wide range of performance monitoring metrics and supports custom charts. Its powerful diagnostic capabilities promptly detect and automatically diagnose events on your database instance, providing root cause analysis and suggestions. You can also manually run diagnostics for a specific time range to understand the operational status of the instance.
-
You can view slow SQL statement trends, execution status, and optimization suggestions.
-
This feature uses offline data analysis to analyze and automatically tag all slow SQL statements generated on all instances from the previous day. This process runs at 01:00 every day. It helps you automatically classify slow SQL statements and prioritize them for governance. It also provides governance suggestions and a data exporting feature.
-
Built on full request logging and security auditing, this feature combines search, SQL Explorer, security audit, traffic playback, and stress testing. It helps you retrieve detailed information about SQL statements, troubleshoot performance issues, identify high-risk sources, and validate instance types.
-
You can automatically or manually diagnose instances to view information such as instance health status, alerts, and slow SQL statements.
-
This feature performs scheduled daily inspections and scoring for all database instances. You can also select specific database instances and inspection periods to manually trigger scoring. This helps you understand the operational status of your database instances.
-
You can select the database instances and metrics that you want to monitor. You can also compare metrics across different database instances and use filter interaction for multiple metrics. This helps you monitor the operational status of your database instances.
-
After you enable this feature, you are notified of triggered events based on your settings.
-
This feature automatically diagnoses and optimizes SQL statements and creates indexes. If slow SQL statements occur, this feature promptly diagnoses and optimizes them to help your database system run in an optimal state.
-
You can set trigger conditions for automatic SQL throttling. When the conditions are met, SQL throttling is automatically triggered to control the number of database requests and concurrent SQL statements, ensuring service availability.
-
This feature monitors the storage of your database instance and automatically expands the storage capacity when it is insufficient. This ensures that your business runs smoothly.
-
Automatic scaling for performance optimization
This feature helps your database instance quickly scale to adapt to traffic peaks, ensuring business stability. It also monitors the average CPU utilization in real time and supports automatic scale-in when the average CPU utilization decreases over a specified period.
-
Automatic fragment reclamation
This is a method for automatic space optimization. After you enable this feature, the primary instance automatically runs the `OPTIMIZE TABLE` or `ALTER TABLE` command to reclaim tablespace fragments and organize physical space fragments.
-
This feature predicts the values of performance metrics for the next 24 hours based on the historical data from the last 10 days.
-
This feature scales out the instance type of your database instance at a scheduled time based on a preset policy. The instance automatically reverts to its original instance type after the scale-out duration ends. This feature helps you manage predictable, periodic changes in database load, meeting business requirements while controlling costs.
Authorization
To use DAS features, you must grant the required permissions to Resource Access Management (RAM) users.
-
You can grant permissions to RAM users by attaching a system policy. For more information, see Manage RAM user permissions.
The system policies for DAS are as follows:
-
AliyunHDMFullAccess: Grants the RAM user full permissions to manage DAS.
-
AliyunHDMReadOnlyAccess: Grants the RAM user read-only permissions on DAS.
-
AliyunHDMReadOnlyWithSQLLogArchiveAccess: Grants the RAM user read-only permissions on DAS and permissions to use the Search feature (including data exporting) of SQL Explorer and Audit.
-
-
You can grant permissions to RAM users by creating a custom policy. For more information, see Create a custom policy.
For example, to grant a RAM user read-only permissions:
{ "Action": [ "hdm:Get*", "hdm:Describe*", "hdm:Query*", "hdm:Support*" ], "Effect": "Allow", "Resource": "acs:rds:*:*:*" }