What is Agentless Monitoring?
Monitoring IT infrastructures often involves the installation of agent software on target servers to collect performance metrics about the servers and application servers executing on them. Performance metrics collected by the agents are then uploaded to a management server for storage, analysis, alerting and reporting to administrators. The need to install and maintain agent software on all monitored servers is a key drawback with this approach.
Agentless monitoring, on the other hand, collects performance metrics from the infrastructure without installing any agent software on the servers or devices being monitored – which makes the agentless approach easier to manage than agent-based monitoring.
When to Use Agentless Monitoring?
Agentless monitoring is suitable to monitor many IT infrastructure components. Without having to install and manage an agent, various parts of the infrastructure can be monitored and performance metrics obtained in an agentless manner. Whether one wants to use a monitoring agent or go the agentless route is completely based on their organizational policy requirements and custom needs. Given below is a list of IT infrastructure components that are recommended to be monitored in an agentless manner:
- Network devices are monitored in an agentless manner. SNMP polling is used to assess the health of the network devices periodically. At the same time, SNMP traps from the devices and syslog messages are collected to detect asynchronous events.
- Flow-enabled devices export network traffic information to pre-configured nodes, thereby facilitating network traffic analysis and top talker identification.
- Storage devices such as storage arrays (SAN and NAS) can be monitored using SMI-S protocol support that they offer.
- Hypervisors such as VMware vSphere, Oracle VM server, Citrix XenServer, etc. support web services APIs. Metrics collected by polling these APIs provide insights into the health of the hypervisor, the hardware supporting it, and the status of each virtual machine hosted by a hypervisor.
- Hyper-converged infrastructures, such as Nutanix, provide REST APIs through which agentless monitoring can be supported.
- Virtual desktops (such as Citrix XenDesktop and VMware Horizon) can be monitored agentlessly, eliminating the need to install an agent inside every desktop. Resource usage inside the VM operating system , identification of top resource consuming processes, as well as user experience information – such as screen refresh latency and virtual channel usage – can be obtained from every virtual desktop in an agentless manner.
- Public cloud (such as AWS, Azure, etc.) environments do not require agents on the cloud instances. Cloud monitoring data can be obtained from cloud APIs.
Agentless Monitoring with eG Enterprise
eG Enterprise supports both agentless and agent-based monitoring. Administrators can choose which parts of the infrastructure they want to monitor agentlessly, and which components to monitor using agents.
In eG Enterprise, agentless monitoring is implemented using a remote data collector that uses protocols such as SSH for Unix monitoring, Windows management instrumentation (WMI) for Windows servers and applications, HTTP/S for web-based applications, JMX for Java applications, SNMP for network devices, etc.
Agentless Monitoring with eG Enterprise
Agentless monitoring is recommended for network devices and storage devices, as well as for virtualization and virtual desktop monitoring where strong APIs are supported by the underlying platforms. For application servers running on legacy operating systems like Windows and Linux, agent-based monitoring is usually preferred. View supported platforms and technologies »
Infrastructure-Wide Performance Correlation and Automatic Root Cause Diagnosis
- eG Enterprise auto-correlates performance metrics from across the IT infrastructure – user experience to application code, desktops to servers, virtualization to storage, and network to cloud – regardless of whether they are collected in an agentless manner or by using an agent.
- eG Enterprise uses artificial intelligence to differentiate the cause of a performance problem from its effects, and automatically pinpoints the root cause for faster troubleshooting and service restoration.
- Administrators can view the entire infrastructure topology from a service-level view and understand dependencies between various infrastructure tiers.
Comparing Agentless and Agent-based Monitoring
|Ease of deployment
||Easier to deploy. Software installation is required only on the remote data collector.
||Agent deployment is required on each server.
||The remote data collector must be allowed to communicate with the target system on different ports. The data collector may also need to be installed with domain administration privileges in order to access the remote systems.
||Much more secure than agentless monitoring. The agent to application/OS communications are handled internal to the server. Hence, no additional firewall rules need to be configured.
||Introduces additional network traffic as the raw performance data is transported to a remote data collector.
||Bandwidth efficient: Data is collected locally and only the processed results are transported to the console.
|Breadth and depth of monitoring
||Can be limited because not all applications and systems have built-in monitoring capabilities.
||Provides deeper, broader monitoring.
- Agentless monitoring is easy to deploy. Get up and running in a few minutes.
- Low maintenance as just a few data collectors need daily surveillance
- Complete flexibility: Pick the infrastructure tiers to be monitored with agents, and the ones to be monitored in an agentless manner
- Seamless integration of metrics from agent and agentless monitors on a single pane of glass