What to Look for in a Network Monitoring Application
Network monitoring is necessary to keep a data center functioning at maximum performance and efficiency. Monitoring applications are designed to detect, observe, and analyze every facet of network operation, scrutinizing applications and devices in real time and allowing IT to respond to a wide range of operational issues quickly.
The most important monitoring application capabilities are the ones that provide situational awareness, explained Dustin Hillard, CTO of Versive, a machine learning startup. "A good monitoring application should enable prioritizing the hygiene of devices and traffic patterns that represent a risk to business operations," he added. The goal is to bring visibility to various events and to enable the network team to act with confidence and contextual understanding.
A monitoring application also needs to accommodate multiple types of users, each with different needs and levels of technical understanding. "A network operations center user will want good aggregation and correlation capabilities to highlight problem areas and direct attention to network health, while a network engineer will be more interested in deep diagnostic capabilities, such as protocol tracing and performance management," said Michael Cantor, CIO, Park Place Technologies, a storage and server support provider. "Even end-users are a potential target for network monitoring applications; they may be interested in a heavily-utilized web site’s traffic flow in order to understand its usage patterns."
A monitoring application should provide the ability to combine network data with business data. "This gives everyone within the company the ability to understand what is going on in the network as it relates to their own job functions," noted Justin Ryburn, head of solutions engineering at network analytics solutions provider Kentik. A marketing team, for instance, might use monitoring insights to identify fresh sales opportunities or prospects based on network usage. Meanwhile, enterprise leaders can follow trends, such as application usage, that reveal important insights into employee behavior. "For example, the data can show how much employees are leveraging a BYOD policy versus relying on company-owned tech," Ryburn said.
While metrics are a fundamental part of network monitoring, it's important to remember that metrics alone cannot improve network, IT, or business health. "Network monitoring applications can provide important visibility to understand when the network is suffering from unintended or malicious behaviors, but only when problems are surfaced can they be understood and immediately acted on," Hillard observed.
Although monitoring applications have become more open and intuitive in recent years, there are still features that remain difficult for people outside of network operations to understand. "As more companies become reliant on the network for their business, more stakeholders within the organization need to understand what is crossing the network and how [the network] is performing," Ryburn stated. "Confusing features make it hard for these newer stakeholders to really understand the information in the application."
Overly-complicated reporting and monitoring features are a particular headache. "Too much insight can mean that problems are hidden amongst noise," warned Chris Payne, managing director of Advanced Cyber Solutions, a UK-based IT security technology provider.
Mike Puglia, chief strategy officer at network monitoring software provider Kaseya, urges managers not to be carried away by flashy, yet often irrelevant features. "Everyone is impressed with visually stunning NOC (network operations center) views and maps," he observed. But a birds-eye status view is usually only important for managers. "Most technicians never use these features to troubleshoot ... and we tend to apply an inordinate amount of value to these features which aren't used in the day-to-day operation," he said.
One of the biggest mistakes managers make is choosing a monitoring tool that can't scale to meet evolving network requirements. "A lot of solutions aggregate data or provide roll-ups," Ryburn explained. "This leaves you blind, lacking the real details of your network traffic." Another mistake is choosing an application that lacks APIs. "This leaves your data locked in a single, siloed application, and more often than not users are unable to correlate it in a meaningful way," he noted.
Evolution and automation
The most important ongoing evolution in monitoring applications is the growing adoption of machine learning, artificial intelligence, and robotic process automation. "The bane of the network management tool for many years has been the false alert—endless loops of email alerts that actually don’t signify anything," Cantor said. The use of AI and ML technologies promises to decrease the manual effort necessary to sort through false positives, and RPA (robotic process automation) technology will allow the system to take sophisticated actions without human labor, he predicted.
The scale of enterprise networks and the diversity of applications and network loads is making it impossible for humans to keep track of everything that takes place on the network, Hillard said "Automation is a critical capability to enable understaffed IT teams, focusing machines on what they are good at and humans on what they are," he added. "This can help focus human effort on fixing the problems that impact the business, rather than on tracking down rampant alerts that don’t end up being real problems."
Recommended For You
Low-Power WANs offer an alternative to 5G for connecting a fast-growing array of basic devices and sensors that transmit small amounts of data.
An effective network visibility strategy requires understanding the technical, financial, political, and legal aspects impacting your network operations.
Emerging organizational structures for IT include placement of IT pros in user areas and departments forming their own "micro IT's."
Comparing a good and bad trace helps identify performance issues. Dynamic baselining can be used when you do not have a good trace to reference.
Combining commodity server platforms and FPGA-based SmartNICs will allow network applications to operate at hundreds of gigabits of throughput with support for millions of simultaneous flows.
SD-WAN implementations are on the rise thanks to the potential cost savings, increased network resiliency, and better application performance they deliver.