CA SystemEDGE User Guide › Concepts › Guidelines for Using the SystemEDGE Agent
Guidelines for Using the SystemEDGE Agent
You can gain the most value from the SystemEDGE agent by setting up effective management policies for your systems, networks, and applications. In particular, follow these guidelines:
- Manage the agent in managed mode from CA Server Automation and restrict SNMP-based changes or monitor these changes so that you have one consistent point of agent configuration.
- Limit SNMP access to the agent through access control lists and binding to private interfaces.
- Create groups in your management software based on the following:
- Operating systems
- Database systems
- Clients
- File servers
- Email servers
- Web servers
- Physical location
Note: For more information about creating groups, see the CA Server Automation Administration Guide or CA eHealth Administration Guide.
- Define configuration policies for groups of systems instead of on an individual system basis. Use CA Server Automation to deploy configuration sets to groups of systems.
- Include meaningful information in the system location and contact field of your sysedge.cf file. For example, include information such as Rack 0, Slot 1, Atlanta, and email: it@yourdomain.com.
- Use a well known range of row indexes across your monitoring tables. When defining these indexes and reserving rows, keep in mind the following:
- Use large ranges of index numbers for each type of monitoring to enable growth.
- Use standard index entries for specific types of monitoring entries. For example, always use row 10,000 for monitoring the total amount of CPU available.
- Use CA Server Automation to apply your self-monitoring entries to groups of systems.
Note: For more information about using standard index rows, see Row Creation for the Self Monitor Table in the chapter “Self Monitoring”.
- Consider the following as you create entries that you can use across multiple systems:
- Monitor total CPU utilization states (which enable the configuration to be portable across single- and multi-processor systems).
- Monitor thresholds and configure SystemEDGE to send a limited number of traps (for example, two or three) to prevent flooding of CA Server Automation.
Note: For more information about effectively monitoring thresholds and clearing traps, see Monitor Table Flags in the chapter "Self Monitoring."
- When you are managing a large network (hundreds or thousands of systems) and polling each system for granular data (at intervals of less than 15 minutes), do the following:
- Use history collection to gain highly granular data at the agent level, and let the management system poll the History table. You can use SystemEDGE for short-term history collection and eHealth (or another management system) for the long-term historical view.
- For more information, see the chapter "History Collection."
- Push the monitoring out to the agent, and configure the agent to send traps based on these self-monitoring entries.
- Limit the number of potential traps from a single monitored entity by using the stateful monitors.
For more usage tips, see the following sections in this guide:
- Recommendations for Configuring Security in the chapter “Agent Configuration”
- Recommendations for Process and Service Monitoring in the chapter “Process and Service Monitoring”
- Recommendations for Log File Monitoring in the chapter “Log File Monitoring”
- Recommendations for Using Extensions in the chapter “Custom MIB Objects”
Copyright © 2013 CA.
All rights reserved.
|
|