Agent-based monitoring

OpManager has taken the next step to better suit your network monitoring needs. The long awaited agent-based monitoring feature is finally here! Now you can install a fully-functional agent in your critical devices, and OpManager will help you comprehensively monitor important performance metrics in real time.

With the introduction of this agent based monitoring, the operating load on your OpManager server reduces drastically, allowing you to improve server performance. Download a free trial for 30 days right away, and start experiencing the capabilities of OpManager's agent-based monitoring feature!

Agent monitoring feature in OpManager is available in build 125312 and above.

How is it different from the previous versions of OpManager we've used?

Until now, agent-based and agent less monitoring have mostly been viewed as two entirely different concepts. OpManager agent-based monitoring allows you to explore the capabilities of agent-based monitoring, while retaining the original functionalities of agent-less monitoring.

From the user experience perspective, OpManager has almost no changes and the user interface would be just the same. All changes that have been done are at the code level, and this leaves the user experience unaffected while also improving the efficiency and performance of the OpManager server.

How does this agent-based monitoring work?

Unlike OpManager's usual polling mechanism that involves the server constantly contacting the devices, a small executable file called the agent is installed on any device you wish to monitor. This file is extremely lightweight ( ~ 1MB in size ), and also consumes negligible amounts of CPU, memory and bandwidth. (Note that this agent can be installed only in Windows-based devices.)

Once this agent is installed and mapped with OpManager in a very quick process, details about the device-specific monitors (monitoring interval, monitors to be polled) are sent to the agents, and the agent takes over from here. It then regularly fetches the requested performance metrics and sends it to the OpManager server at regular intervals. Users will not have to worry about any security concerns, since all communication between the agent and the server happens over an SSL connection. To ensure the optimum performance of OpManager without carrying out load configuration changes, it is recommended that the agent be installed in not more than 1000 devices.

How do I choose between these two? What would be the best method to use?

Agent-less and agent-based monitoring have their own user bases, and both have their own pros and cons. The preferences are based on the requirements of the user. Agent-less monitoring is an easy to set up, quick and non-intrusive method to set up network monitoring in your infrastructure. But on the downside, it needs more resources to function as polling will be done entirely from OpManager's end.

On the other hand, agent-based monitoring needs a little additional work during setup, but is more robust and reduces load on the OpManager server. It optimizes the polling process to a great extent and allows OpManager to monitor devices more efficiently using lesser resources. So it all comes down to the user's preferences and requirements. For a detailed comparison between agent-based and agentless monitoring, click here.

OpManager provides a lot more flexibility when it comes to choosing between these methods. You can choose either of these methods to implement in your network based on your monitoring requirements, or even use a combination of these in your network to efficiently perform real-time monitoring of your network devices.

OpManager Agent Deployment in LAN environment

Working of OpManager's Agent-based monitoring in LAN architecture

  1. Server component:

    The OpManager server is installed in the customer's network and is used to perform network and server monitoring activities, set performance thresholds, configure workflows for devices and generate reports for devices in your network. All these actions can be performed from the OpManager web console with just a few clicks.

  2. Agent component:

    OpManager agent is a lightweight software that gets installed on the monitoring machine. It performs as an arm of the OpManager server that takes over all the polling and monitoring activities of the device. All monitor related configuration information are periodically synced to the agent, and the agent fetches these information from the device and updates it to the OpManager server.

  3. Web Console:
    • Provides a unified portal for the user to access all network monitoring functions.
    • Can be accessed from anywhere: LAN, Remote Offices, and from home through Internet/VPN.
    • No separate client installations are required to access the web console.
  4. Notification server (if Push Mode is enabled in any of the agents):

    The Notification Server is a utility available as a part of OpManager. When an agent is installed, it will initiate a secure connection to the Notification Server to get real-time notifications when a monitor is added/modified, or when actions like Test Monitor or any Inventory update is performed on that device. When a notification is sent to the corresponding device, the agent contacts the OpManager server immediately and collects the updated info about the device's monitors.

    Note that this component is included in the process of agent-based monitoring only if any of the agents have Push Mode enabled. Click here to know how push mode can enabled in an agent-monitored device, and the prerequisites for using the Notification server for Push Mode.

OpManager Agent Deployment in WAN environment

OpManager Enterprise allows you to monitor and manage devices in different geographical locations. In this architecture, several Probe servers are installed in different geographical location, under a Central server that handles the monitoring data from all the Probe servers and the devices under them.

Working of OpManager's Agent-based monitoring in LAN architecture

  1. Central server component:

    OpManager has to be installed in the LAN of the head office, and the device must be configured so as to provide internet connectivity to the server. You might require to adopt some security standards to harden the OS where this OpManager Central server is installed.

  2. Probe server component:

    The Probe server component is installed in the server of the branch office, and it takes care of the monitoring activities of all devices in that branch. This Probe server communicates with the Central server to fetch all device details and monitoring configurations for that branch.

    1. Communication with agent is very resource-efficient, as it consumes very less bandwidth
    2. Fetches all monitoring info from the Central server and performs real-time monitoring independently
    3. Communicates over a secure channel (SSL/HTTPS) with the Central server
    4. Probe server upgrades will be pushed from the Central server automatically or when triggered by user
  3. Agent component:

    OpManager agent is a lightweight software that gets installed on the monitored device, under the control of any of the Probe servers. It works as an arm of the Probe server that performs the monitoring activities on that particular device.

    1. Lightweight software component
    2. Can be installed manually, via Active Directory or from the agent installation page in the web console
    3. Agent installation is one-time and upgrades, if any, will be performed automatically
  4. Web Console:
    • Provides a unified portal for the user to access all network monitoring functions.
    • Can be accessed from anywhere: LAN, Remote Offices, and from home through Internet/VPN.
    • No separate client installations are required to access the web console.
  5. Notification server (if Push Mode is enabled in any of the agents):

    The Notification Server is a utility available as a part of OpManager. When an agent is installed, it will initiate a secure connection to the Notification Server to get real-time notifications when a monitor is added/modified, or when actions like Test Monitor or any Inventory update is performed on that device. When a notification is sent to the corresponding device, the agent contacts the OpManager server immediately and collects the updated info about the device's monitors.

    Note that this component is included in the process of agent-based monitoring only if any of the agents have Push Mode enabled. Click here to know how push mode can enabled in an agent-monitored device, and the prerequisites for using the Notification server for Push Mode.

Key features of agent-based monitoring:

  1. Quick installation and setup: Installing the agent in your devices is as simple as a few clicks! You can choose from three methods of installation: you can install it manually, or either via Active Directory or by manually pushing it to devices from OpManager from the Agent Installation window in the web console.

  2. Highly optimized polling mechanisms for real-time monitoring: High polling frequencies for a large number of devices would essentially drain your network resources. Instead, opting for agent-based monitoring in suitable devices will save you precious computing power.

    Instead of the OpManager server performing the polls every time, the agent fetches all the polling details needed periodically and only sends the polled data to OpManager. This allows the user to configure high frequency polls without consuming too much of resources.

    The performance metrics are monitored and updated in real-time, enabling admins to know the status of the devices with much ease. Also, when a device goes down, the status of the device is updated instantly allowing admins to track downtime more efficiently.

  3. Improved security: Agent-based monitoring has quite a few pros with respect to security of the network. When you install an agent in a device, you don't have to provide credentials for that device in OpManager, and no special privileges are needed for the agent to be installed.

    The agent also communicates with the server only over SSL, and that assures that the communication between the agent and server is always secure. Also, polling requests are usually sent inward to the device from OpManager. But with agent-based monitoring, all the data is only being sent outward, and there is almost no direct communication for OpManager with the device (except with the agent). This ensures that there are no security concerns due to monitoring.

Want to know more on agent-based monitoring? Sign up for a free, personalised demo of ManageEngine OpManager and let our technical expert guide you on how it can prove to be highly efficient for your network!

 
 Pricing  Get Quote