Whether you're outsourcing your monitoring or doing it in-house, there are many factors to consider when choosing a network performance monitoring solution. One that may seem trivial, but will have a significant impact on your environment now and in the future, has to do with agent versus agentless monitoring.
What is an agent?
Non-native software provided by a third party which must be installed locally on the device to be monitored.
What is meant by agentless?
Agentless monitoring requires no separate installation of a software agent and is enabled by protocols embedded in the software of the device by the manufacturer (SNMP, WMI) or exposed via an API by the OEM software developer. Agentless NMS platforms leverage built-in monitoring capabilities.
Points to Consider with Agents
Agents are often sold separately along with maintenance agreements covering support and future releases. You may have to purchase multiple agents to monitor the diverse platforms deployed in your network.
While agents may expose additional monitoring capabilities, they’re usually proprietary to the specific NMS platform you’re using. If you wish to switch NMS platforms, the new NMS will not be able to utilize the proprietary agents you invested in. In addition, if you want to switch hardware or software platforms, there may not be an equivalent agent available, or you may need to purchase additional agents.
Any time you introduce 3rd party software in your network, it must be vetted for proper security. This takes time, expertise and testing to ensure agents are secure.
If you upgrade your hardware, operating system or application, will a compatible, tested agent be available when you need it? It may or may not, and if it is, it needs to be tested and vetted by your staff and then deployed on each of the affected devices. There may also be a cost or delay in getting an updated agent.
Your staff will need to deploy the appropriate agent(s) on every device you wish to monitor. Even with automation tools, this can be a time-consuming and daunting task.
How agents interact with other applications running on the system should be evaluated. The agent may consume resources, such as CPU and memory, needed by the primary application.
Points to Consider with Agentless Monitoring
Agentless monitoring capabilities are already embedded in the operating system or application at no additional charge.
Agentless monitoring uses standard protocols like SNMP, WMI, REST or API calls. You likely have equipment and applications from multiple vendors in your network. If you switch platforms, operating systems or applications, an agentless solution can monitor them.
Since the manufacturer or developer owns the source code of the embedded capabilities, security is built-in to the platform. There is no additional cost to examine and vet those capabilities for security. In addition, while agents have direct and implicit access to the platforms they’re running on, agentless monitoring allows access to be restricted to only what is needed.
When you upgrade your hardware or software, the manufacturer provides necessary upgrades to the embedded capabilities at the same time. There will rarely be a disconnect between releases and availability of appropriate monitoring.
Since there is nothing to install on the devices you wish to monitor, there is no need to deploy anything additional.
An agentless monitoring platform can leverage multiple capabilities, depending upon what you wish to monitor. For example, it might use SNMP for metrics like disk utilization or memory, synthetic transactions to validate batch jobs, and API calls to monitor specific application data.
Ultimately, you have to decide what’s best for your environment. iGLASS provides 24x7 agentless monitoring and offers an unparalleled degree of flexibility, no matter what monitoring challenges you’re facing. Contact us to learn how we can look after the health and operation of your network.