Successful insight into the performance of a company's networks starts with effective network performance management (NPM) tools. However, with the plethora of options it can be overwhelming for IT teams to choose the right one. This blog continues the 10 essential questions to ask before selecting an NPM tool.

Start with: 10 Questions to Ask When Evaluating Network Performance Management Solutions - Part 1

Question #6: Does the NPM solution support machine-learning, advanced anomaly detection and correlation?

Most solutions make broad claims in these areas, without much to show for it. Networks, and the demands on those networks, are highly unique between companies, so it's extremely difficult even with today's computing technologies to apply generalizations across network performance monitoring. But what is becoming more practical is the ability of NPM solutions to learn and apply knowledge based on machine learning of data trends over time, to create baselines and identify anomalous behavior without having to pre-configure limits or behavior characteristics.

Legacy systems require a great deal of a prior knowledge, and then significant configuration, for anomaly detection to work effectively. ML and AI are beginning to change that, but it's important to really validate the claims of any NPM solution.

Question #7: Is the solution utilizing advanced analytics and reporting?

To derive meaningful insights into complex issues, analytics platforms must provide reports and analyses on most, if not all, of a network's performance. This includes offering custom reporting for baselining and trend analysis and the ability to easily pivot reports to focus on key network performance intelligence.

Additionally, a modern NPM solution should correlate data across multiple network domains offering a cohesive, big-picture view of performance metrics and providing intelligent alerting, giving back valuable time to strapped IT teams.

Question #8: Does the solution assist with capacity planning?

Under-provisioning network resources can lead to congestion, bad user experience, and loss of productivity — overall, a negative business impact. Over-provisioning can lead to excess spending and a hit to the bottom line. Therefore, capacity planning is critical in helping to avoid performance problems and negative impacts.

When looking at an NPM solution, it is critical that it supports capacity planning through these features:

■ Service Level Agreement (SLA) management

■ Network and application analysis

■ Baselining and trending

■ Exception management

■ QoS management

Question #9: Does the solution facilitate root-cause analysis?

Most NPM solutions focus on visualization and reporting based on flow data (NetFlow, sFlow, IPFIX, etc.). These solutions, and the flow data that feed them, provide enough detail to troubleshoot many network and application issues. But at times flow data are simply not enough to get to the root cause of a problem. When more detailed data are needed, a recording of the network traffic itself, at the packet level, provides the detailed data needed for root-cause analysis. And when this packet data is analyzed with appropriate software, the software itself can identify many of these detailed network and application issues.

An NPM solution that can quickly pivot from flow data for visualization and reporting to packet data for analysis provides the most comprehensive solution and will significantly reduce the mean time to repair (MTTR).

Question #10: Can the solution provide scalable, enterprise support?

As the number of devices in many organizations continues to grow, it's important to implement tools that support this growth, particularly for large-scale organizations. A modern NPM platform must be able to analyze devices and environments at scale without latency and extend into additional environments such as multi-vendor WAN, public and private clouds and more. It also must support capacity planning and predict if a network can support an increase in business-critical traffic.

As organizations continue to grow and disperse, it is more evident than ever that ensuring optimal network performance is critical to business efficiency. When choosing a network performance monitoring solution, considering the questions above and implementing a unified platform will help organizations eliminate the cost and complexity of point solutions, reduce downtime, and successfully address the challenges of a modern network system.