For the modern enterprise, monitoring multi-platform database performance should be considered a key capability.
Concern over the performance of business-critical databases extends far beyond the confines of the database team.
While the responsibility to address and resolve performance issues usually falls to an organization’s database administrators (DBAs), many other individuals may want or need to know the current state of an enterprise’s databases.
Incorporating the tracking of database performance as a component of comprehensive monitoring efforts puts all the information necessary to understand the state of enterprise computing resources in a centralized location.
This can only be seen as a good thing by the operations team since it reduces the difficulty of maintaining a complex and diverse environment.
IDERA’s Uptime Infrastructure Monitor provides the functionality required to identify potential database performance issues as part of an all-inclusive monitoring scheme.
Uptime enables monitoring of the complete physical, virtual, and cloud environment from a unified and fully customizable interface.
Teams can create custom dashboards that focus on the infrastructure elements that interest them, including groups of database servers or specific database instances. With color-coded alerts and status summaries, Uptime makes it easy to identify problems so they can be addressed promptly.
How Uptime Monitors Database Performance
Uptime uses service monitors that check the availability and performance of infrastructure elements and services. The tool issues an alert if issues are discovered.
Uptime’s functionality is extended by advanced monitors that are available as plug-ins. The advanced monitors employ custom scripts and programs that don’t need to interact with an agent on monitored systems.
Uptime offers advanced monitors for major relational database management systems (RDBMS). It is compatible with Microsoft SQL Server, Oracle Database, MySQL and MariaDB, IBM Db2, PostgreSQL, and SAP (Sybase) ASE (Adaptive Server Enterprise).
Following is an overview of the specific monitors available for each database platform:
Microsoft SQL Server
Uptime offers the following monitors for SQL Servers:
- The basic check monitor verifies that an instance is listening on the server’s ports and if it can process queries. It runs queries against a SQL Server instance and reports or alerts on response and response time.
- The advanced metrics monitor reports on the availability and performance of individual databases. It captures metrics such as user connections, transactions per second, total server memory, and wait times.
- A query monitor runs queries against a SQL Server instance and checks response and response time. Warnings are set if monitored query responses do not produce the expected results.
- The tablespace check monitor keeps track of the size of data files and reports and alerts if user-defined thresholds are exceeded.
For Oracle databases, Uptime provides the following monitors:
- The basic check monitor verifies that an instance is available, listening on the server’s ports, and can be logged into. A SQL script is run against databases or instances and response time is evaluated for use in reporting and alerting.
- The advanced metrics monitor provides reporting on metrics related to availability as well as furnishing data that can be used for performance tuning. Some of the tuning-based information collected includes metrics related to the buffer cache, disk sort ratio, and redo logs.
- The query monitor runs queries against an Oracle instance and checks response and response time. Responses must conform to pre-defined outcomes or warnings and alerts are set.
- The tablespace check monitor reports on the relative size of individual tablespaces within instances and alerts on user-configured thresholds.
- An extendable tablespace check monitor gathers metrics on tablespace capacity such as available space, used space, and percent free space to be used for generating alerts and graphs in Uptime.
MySQL and MariaDB
The following monitors report on MySQL and MariaDB database instances:
- The basic check monitor verifies that an instance is available and accessible by users. A SQL script run against databases reports in response time and valid responses and alerts if issues are detected.
- The advanced metrics monitor checks if an instance is listening on the server’s ports and collects multiple metrics to evaluate its efficiency. Metrics such as slow queries, threads running, and threads connected are monitored and alerts get generated when thresholds are exceeded.
- A replication monitor collects metrics like “seconds behind master”, last SQL error number, and slave SQL running.
- The status monitor plug-in uses status commands to collect information on connections, open tables, and average queries per second.
For IBM Db2, the following monitors are available:
- The basic monitor plug-in runs queries against an instance, checking for valid response and response time and alerts on irregularities.
- The tablespace monitor plug-in collects and reports on tablespace performance across all tablespaces that exist for an instance.
- The buffer pool monitor plug-in gathers metrics regarding buffer performance such as average read time, average write time, and total hit ratio for reporting and alerting.
Monitors are also available for PostgreSQL and SAP that report on system availability, valid response, and response time to serve as an early warning system for database problems.
Download this IDERA solution brief for a more extensive look at monitoring database performance with Uptime. The document includes screenshots that illustrate the customized dashboards that can be created with the tool.