How Does eG Enterprise Monitor A Linux Systems Group?

eG Enterprise monitors the Linux systems in an agentless manner. Before starting to monitor the Linux systems, ensure the following pre-requisites are fulfilled.

Pre-Requisites for monitoring Linux Systems

The following list of pre-requisites should be fulfilled before you start monitoring the Linux systems.

  • For internal monitoring of the individual Linux system, a light-weight eG VM Agent has to be installed on each of the Linux systems. The steps for downloading installing the VM agent are explained in Downloading and Installing eG VM Agent topic.

  • When monitoring the Linux systems, the remote agent needs to be in the same subnet as that of the eG VM Agent or accessible on the local network of the eG VM Agent.

  • Ensure that the eG VM Agent communicates with the IP address of the remote agent directly and that the communication is not via a NAT or port forward. To know how does the eG VM agent pushes the performance metrics to the remote agent, refer the following section.

How does the eG VM Agent communicate with the eG Remote Agent?

An eG remote agent on a remote Windows/Linux host is used for monitoring. The remote agent does not initiate any communication with the eG VM Agent on the individual system. This is because, this remote agent has no knowledge of the Linux Systems Group on which the systems are operating. Therefore, it can neither discover the IP address/host name of the Linux Systems Group, nor can it communicate with them to collect the 'inside view' metrics.

This is why, the eG VM Agent on a Linux system has been specifically engineered to initiate all communication. To establish this communication, the eG VM agent should first know which remote agent it needs to connect to and which Linux Systems Group it has to monitor. In order to feed this information to the eG VM agent, you can do one of the following:

  • Configure the eG VM Agent with the IP address of the eG remote agent and the nick name of the Linux Systems Group component it needs to monitor. This is the recommended approach.

  • Alternatively, you can configure the eG VM Agent with the IP address/port number of the eG manager with which it should communicate. If you choose this approach, then the eG VM agent will download the details of the remote agent and Linux Systems Group component from the eG manager itself. Take this route, only under the following circumstances:

    • If you be will changing the remote agent for a VM agent dynamically - i.e., using the eG manager itself, and/or;

    • If you want to auto-upgrade the eG VM Agent from the eG manager.

To know how to achieve the above, refer to the Configuring the eG VM Agent to Communicate with the eG Manager/eG Remote Agent topic.

Once the eG VM agent has all the information required for communicating with the eG remote agent, it does the following:

  • Collects 'inside view' metrics from the Linux system it is deployed on;

  • Caches the metrics locally;

  • Connects to the eG remote agent that has been assigned to the 'Linux Systems Group' component it monitors

  • Pushes the cached metrics to that remote agent via TCP; note that a TCP port has to be opened on the remote agent to facilitate this communication. If the remote agent is behind a firewall, then make sure that the firewall is configured to allow one-way communication from the eG VM agent to the remote agent.

Upon receipt of the 'inside view' metrics, the remote agent sends the metrics to the eG manager.

eG VM Agent and eG Remote Agent Communication

Figure 1 : How does the eG VM Agent communicate with the eG Remote Agent?