What 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. The need to install and maintain agent software on all monitored servers is a key drawback with this approach. Agentless monitoring 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.
Agent vs Agentless Monitoring
There are several differences between agent-based and agentless monitoring.
- When it comes to agent-based monitoring, agent deployment is required on each server which makes deployment more difficult. Agentless monitoring is much easier to deploy and software installation is required only on the remote data collector.
- Agent monitoring is more secure than agentless and no additional firewall rules need to be configured.
- When it comes to network overhead, agentless monitoring introduces additional network traffic as the raw performance data is transported to a remote data collector. With agent monitoring, data is collected locally and only the processed results are transported to the console.
- Finally, the depth of monitoring can be limited with agentless monitoring because not all applications and systems have built-in monitoring capabilities. Agent-based monitoring provides deeper, broader monitoring.
As you can see, there are pros and cons to each and a combination approach to monitoring might be the best option.
When to use Agentless Monitoring?
Measuring the availability, response time and behaviour of each and every business transaction is key to understanding the user journey. When a user performs a transaction on a digital business service, the application owner needs to know:
- Monitoring of network devices (routers, switches, firewalls, etc.) must be predominantly agentless – using SNMP polling, SNMP traps, flow data, etc.
- Monitoring of storage platforms: It is not possible to install agents on storage devices. Command line interfaces, SNMP, and SMI-S are some of the ways in which storage platforms can be monitored in an agentless manner.
- Monitoring of advanced networking platforms: Devices like Citrix NetScaler, F5 BigIP, etc., perform a wide range of functions. These devices act as network accelerators, VPN concentrators, load balancers, protocol proxies, etc. These devices run as appliances and administrators will not be able to install agents on them. Hence, an agentless monitoring approach is preferred.
- Monitoring of virtualization platforms (VMware vSphere, Citrix Hypervisor, Nutanix Acropolis, etc.) is also agentless. Most virtualization vendors do not recommend installing agents on their server platforms. REST/Web service APIs provide a great degree of detail about the performance of the hypervisors and VMs running on them.
- Monitoring of cloud platforms (AWS, Azure, Google Cloud, Alibaba Cloud, etc.) must be agentless. Customers do not have direct access to the cloud infrastructure. So, agent-based monitoring is not an option. APIs such as CloudWatch are used to monitor the cloud platforms.
- Monitoring of SaaS applications (Salesforce, Microsoft 365, etc.) also must be agentless. APIs from the respective SaaS platform vendors must be used.
- Monitoring of real user experience for web applications: Real User Monitoring for web applications is based on Javascript injection techniques. This is done in an agentless manner: the Javascript executes on client browsers, performance metrics are sent to a RUM collector, which then aggregates metrics and reports performance data to the management server.
How does Agentless Monitoring work?
With eG Enterprise, agentless monitoring is implemented in eG Enterprise using remote agents. A remote agent is capable of monitoring a number of systems and applications remotely, i.e., without requiring an agent to be locally installed on the system that is to be monitored. For monitoring Microsoft Windows systems and applications, a remote agent uses Netbios/perfmon to communicate with the operating system/applications. For monitoring Unix systems, secure shell (SSH) is used. In addition, for specific applications, the remote agent uses application-specific protocols to communicate with the application (e.g., SQLNet for Oracle databases, HTTP for WebLogic and WebSphere application servers, JDBC for Sybase, etc.). Agentless monitoring of virtual environments and cloud technologies uses the respective APIs. SNMP is widely used for agentless monitoring of network devices.
