Chapter 1 MySQL Enterprise Monitor Introduction and Architecture

Table of Contents

1.1 MySQL Enterprise Service Manager
1.2 MySQL Enterprise Monitor User Interface
1.3 MySQL Enterprise Advisor Collections
1.4 Monitored MySQL Server
1.5 MySQL Enterprise Monitor Agent
1.5.1 MySQL Enterprise Monitor Agent Basic Architecture
1.5.2 MySQL Enterprise Monitor Agent Proxy Service Architecture
1.5.3 MySQL Enterprise Monitor Agent Aggregator Service Architecture

MySQL Enterprise Monitor is a companion product for MySQL Server that monitors your MySQL instances, notifies you of potential issues and problems, and advises you how to fix the issues. MySQL Enterprise Monitor can monitor all kinds of configurations, from a single MySQL instance that is important to your business, all the way up to a huge farm of database server machines powering a busy web site.

This section describes the basic components that make up the MySQL Enterprise Monitor product. These components can be installed in various configurations depending on your own database and network topology, to give you the best combination of reliable and responsive monitoring data, with minimal overhead on the database server machines. A typical MySQL Enterprise Monitor installation consists of:

The structure of a typical MySQL Enterprise Monitor installation is shown in the following figure.

Figure 1.1 MySQL Enterprise Monitor Architecture

MySQL Enterprise Monitor Architecture

MySQL Enterprise Monitor can monitor one or more MySQL instances, on a single computer or spread across multiple host machines. The agent, MySQL Enterprise Monitor Agent, collects variables, status, and health information from a MySQL instance and sends this information to the MySQL Enterprise Service Manager. You typically have multiple MySQL Enterprise Monitor Agent processes, one monitoring each MySQL instance.

The monitoring data is sent to the MySQL Enterprise Service Manager, which collates information from all the agents. The MySQL Enterprise Service Manager checks whether the values of status values are within reasonable limits. When thresholds are reached, the server can trigger an event (including an alarm and notification) to highlight a potential issue, such as low memory, high CPU usage, or more complex conditions such insufficient buffer sizes and status information. Each test and its associated threshold value form a rule.

Each rule, together its associated alarms and notifications, is known as a MySQL Enterprise Monitor advisor. Advisors are a vital part of the MySQL Enterprise Service Manager, warning about potential problems and providing troubleshooting advice.

You interact with the MySQL Enterprise Service Manager through a web interface, powered by an included web server. This interface, the MySQL Enterprise Monitor User Interface, displays the information collected by the agents. You can view current status for individual MySQL servers or groups of servers. You control and configure all aspects of the service using the MySQL Enterprise Monitor User Interface.

The MySQL Enterprise Monitor User Interface can display statistical and query information as graphs. For example, you can produce graphs over time of aspects such as server load, query numbers, or index usage information. The graphs help to pinpoint the timeframes and trends related to potential issues.

The MySQL Enterprise Monitor Agent can also collect detailed information about database queries, including the row counts and performance times for executing each query. You can correlate the detailed query data with the graph information to identify which queries were executing when you experienced an issue. A system called Query Analyzer presents the query data in various ways.