MOM 2005 Test Approach

  • Article

Each documented test scenario includes three key areas of the MOM 2005 performance and scalability testing: the hardware, the configuration, and the test results. In each scenario, it is important to note the following:

  • The loads varied in relation to specific scenarios. For example, the number of computers and the frequency of data collection usually varied.

  • In most cases, the load test was higher than the limits supported by MOM 2005 to ensure that performance data was acceptable in supported scenarios. For example, the tests employed 3,000 agents per Management Server and 5,000 agents per management group. MOM 2005 supported limits are 2,000 agents per Management Server and 4,000 agents per management group.

  • Metrics applicable to a particular scenario were used to measure performance and scalability and to compare performance and scalability for that same scenario with different loads.

  • The large agent configuration tests used a RAID 0 array for the database; however, it is generally recommended that you use RAID 1+0 (RAID 10) for optimal performance.

The hardware used to test each scenario is presented in a table and contains the primary specification of each computer used in the test. In cases where the scenario consists of more than one computer, each hardware group is identified in its own table.

The configuration section identifies the specifics of the testing procedure. In sections where more than one configuration exists for the scenario, the details are identified separately.

The results section presents the test data based on the identified scenarios, and provides performance data for each MOM 2005 component. The following scenarios were tested:

  • Large Agent Configuration

  • Small Agent Configuration

  • Single Agent Monitoring

  • Agentless Monitoring

  • MOM Connector Framework

  • Reporting Server

  • Low-Bandwidth Configuration Results