ManageEngine OpManager supports agent-based and agentless monitoring enabling, users to choose the monitoring method that best suits their requirements. Both approaches offer various advantages. Hence, users should understand the difference between each and choose accordingly.
Agent-based monitoring enhances server monitoring by installing lightweight agents on end servers. While the user interface remains the same, in the backend, the agent collects real-time performance data and sends it to the OpManager server, thereby reducing operational load on the server and improving monitoring efficiency.
Agentless monitoring allows the monitoring server to directly poll the devices for availability, performance and overall health of the devices. Since monitoring is handled directly from the OpManager server, users only need to configure the monitoring server, making deployment simpler.
In this document, you will see a detailed comparison between agent-based and agentless monitoring.
| Parameters | Agent Monitoring | WMI Monitoring |
|---|---|---|
| Credential | No credentials required for monitoring. | Requires valid credentials for continuous monitoring. |
| Port Dependency | Uses the endpoint’s default dynamic ports. No additional ports need to be opened. | Requires WMI and RPC ports to be open between the OpManager server and the endpoint. |
| Communication | Communication is initiated from the endpoint to the OpManager server. | Communication is initiated from the OpManager server to the endpoint. |
| Installation | Requires installing an agent on the endpoint server. | No agent installation is required. |
| Polling Impact | Not affected by endpoint changes such as IP or DNS updates. | Not affected as long as credentials and connectivity remain unchanged. |
| Bulk Discovery | Bulk discovery is supported through external tools such as GPO, SCCM, or other endpoint management solutions. | Supports multiple discovery methods such as IP range-based discovery and CSV import, along with scheduling options. |
| Data collection during downtime | If the OpManager server goes down for maintenance, the agent continues collecting data for up to 6 hours. | Data collection stops while the OpManager server is under maintenance. |
| Additional Monitors | Supports Log File Monitoring. | Supports the following additional monitors:
|
| Core Monitoring Metrics |
| Notifications based on polling intervals:
|
| Interface Monitoring | Does not rely on SNMP agents. | Relies on SNMP agents. |
| Direct execution on end servers | In script monitoring, scripts can be executed directly on the end servers. In Network Path Analysis, the source can be configured as the installed end server. | In script monitoring, scripts are executed on the OpManager server. In Network Path Analysis, the source must be configured as the OpManager server. |
| Threshold Violation Alerts | Threshold violation alerts are generated after the data is received from the agent and processed (maximum delay of 5 minutes from polling). | Threshold violation alerts are generated during the polling interval. |
| Limitations of monitors |
| Interfaces cannot be monitored using WMI credential alone. |
| Parameters | Agent Monitoring | CLI Monitoring |
|---|---|---|
| Credential | Credentials are not required for monitoring. | Credentials are required and must remain synchronised, if not, the user account may be locked. |
| Port Dependency | Uses the endpoint’s default dynamic port. No additional ports need to be opened. | Requires Telnet (23) or SSH (22) ports to be enabled on both the OpManager server and the endpoin |
| Communication | Communication is initiated from the endpoint to the OpManager server. | Communication is initiated from the OpManager server to the endpoint. |
| Installation | Requires installing an agent on the endpoint server. | No agent installation required |
| Polling Impact | Not affected by endpoint changes such as IP or DNS updates. | Not affected as long as credentials and connectivity remain unchanged. |
| Bulk Discovery | Bulk discovery can be achieved via endpoint management tools such as Endpoint Management solutions. | Supports multiple bulk discovery methods such as IP range—based discovery and CSV import, with scheduling options available. |
| Data collection during downtime | If the OpManager server goes down for maintenance, the agent continues collecting data for up to 6 hours. | Data collection stops during the maintenance period. |
| Common Monitors |
| |
| Additional Monitors | Supports Script Monitoring. | Supports the following additional monitors:
|
| Core Monitoring Metrics |
| Notifications based on polling intervals:
|
| Threshold Violation Alerts | Alerts are generated after the data is received from the agent and processed (maximum delay of 5 minutes from polling). | Alerts are generated during the configured polling interval. |
| Limitations of monitors |
| Interfaces cannot be monitored using CLI credential alone. |
Based on the comparison between agent-based and agentless monitoring, the following scenarios will help you determine which monitoring approach would best work for you.
| Agent-Based Monitoring | Agentless Monitoring |
|---|---|
|
|
To learn the installation steps for Windows and Unix agents, click here