According to recent research, network managers catch only 60% of network problems before end-users are affected and report them. Clearly there is a need for NetOps teams to put greater consideration into the network management solution being used to monitor networks and alert NetOps of problems before they affect end users. But evaluating which products and vendors can meet today's modern and complex IT business requirements is a challenge.

To help, I'd like to explore 10 key questions every IT admin should be asking when evaluating or working with network performance tools.

1. Does the solution offer complete end-to-end visibility?

Supporting a seamless, high-performance digital experience is a requirement of a modern network management solution. Yet seeing only part of the network doesn't provide the full picture. Appropriate tools need to gather network-performance metrics from infrastructure devices — including routers, firewalls, load balancers and switches — using this application-performance enriched flow data to create a comprehensive application-impact analysis. This includes traffic information from SD-WAN, cloud and remote sites. The tools should also support integrated application visualizations, including application-path analytics, by having the ability to alert on application-performance issues caused by network-device issues. When it comes to performance problems, these solutions should offer streamlined analysis features that can help accelerate the identification of root causes.

2. Can you really see into SD-WAN?

Organizations are increasingly looking to SD-WAN deployments for improved performance and reduced cost. In fact, WAN environments are made more dynamic and secure with SD-WAN automation. For example, it can provide a direct internet connection from a branch in Tulsa to an office in Seattle, enabling teams to balance between multiple service provider and transport types more easily while making intelligent adjustments to application paths for better performance. But without visibility into these traffic flows, it will be difficult to quickly address performance issues. The appropriate tool offers advanced analytics capabilities to gain insights into SD-WAN performance, QoS policies, path routing, and traffic management complexities.

3. Is cloud monitoring supported?

The rise of cloud and hybrid IT gives administrators more options when it comes to finding the right network monitoring solution for their business. IT teams can manage solutions on-premises or in the cloud, or a third party can manage network monitoring at their site. But for true application performance visibility in the public cloud, you'll need to see traffic to and from that cloud infrastructure. If not, the cloud will effectively become a "black box," leaving you unable to isolate performance issues. This cloud visibility is also critical for planning and optimizing as you migrate more services to the cloud.

4. Can you conduct comprehensive application monitoring and optimization?

Application performance is critical to business success. Given that, network teams need to ensure that the network is optimized to support the desired performance of the applications that are traversing that network. But network health and performance characteristics will influence application performance in different, sometimes subtle ways. Understanding the nuances is important, meaning the right network analytics solution must combine application context with network infrastructure metrics and traffic.

5. Does the solution provide insights into voice and video applications?

Voice and video are especially sensitive to network latency. Organizations need to understand, hop-by-hop, how applications are impacted by network infrastructure and routing. Unfortunately, the machine-to-machine, east-west traffic within data centers — the type of traffic driven by increased digital transformation — often stays invisible to IT teams. These blind spots are common and can be expensive. Without granular insights, identifying, troubleshooting, and resolving voice and video traffic issues is difficult.

6. Does the solution leverage machine-learning for advanced anomaly detection and correlation?

Your network monitoring and management solution should incorporate machine-learning techniques to continuously learn and apply knowledge based on big-data performance trends. This includes the ability to create dynamic baselines and identify anomalous behavior from multiple sources of raw data. Critical performance corrections, including determining which voice traffic to prioritize, when to throttle bandwidth and whether a user's access should be blocked, is something that should be supported by machine-learning algorithms. Moreover, you should be able to create automatic baseline trends to ensure that capacity issues don't contribute to performance issues or downtime.

7. Does the solution offer advanced analytics?

Network operations need to apply more sophisticated analytics to network data to derive meaningful insights into complex issues. The right solution should not only allow users to report on N dimensions (application, user, site, device, segment, etc.) and easily pivot reports to focus on key network performance intelligence, but it should also enable custom reporting for baselining and trend analysis. Additionally, it should correlate data across multiple network domains such as WAN, LAN, Data Center, Cloud, etc., to provide a cohesive big-picture view of performance metrics throughout the entire network. 

8. How does it handle capacity planning?

For optimal application performance, capacity planning is critical. Inadequate resource allocation leads to congestion—resulting in bad user experience, loss of productivity and a negative business impact. To avoid inadequate capacity, most organizations resort to over-planning. However, over-planning can be almost as bad as under-allocating, resulting in excess capital spend and a hit to the bottom line. Whether you're ensuring that there is enough bandwidth through a service provider, or verifying the load on network devices, having full awareness in a single view is of utmost importance.

9. Does the solution incorporate AIOps?

The more that NetOps teams can automate, the faster they'll have intelligent, actionable insights at their fingertips to continuously improve network performance — saving your organization time and resources in the process. The benefit of AIOps is that it can learn patterns and correlations, allowing teams to identify, address and resolve slow-downs and outages faster, and with fewer errors, than if they had to sift manually through alerts from multiple IT tools. Even better, AIOps can allow teams to automate corrective action to prevent problems before they arise. Benefits include reduced MTTR, modernizing IT departments and teams, and being able to shift to predictive management as opposed to reactive.

10. Can the solution provide scalable, enterprise support?

Finding solutions that can support the extensive number of devices in your network is important in determining suitable network monitoring tools for large-scale enterprises. If your network is going to expand, you need to keep this in mind as you decide on a monitoring solution. Whatever solution you use needs to be able to analyze devices and environments at scale without latency, and grow into monitoring new computing environments, including SD-WAN, multi-vendor WAN, and public and private cloud environments.

These are some of the top things to consider when picking and evaluating the network performance management solution that's right for your business. It's essential to understand the complexity of enterprise networks and the technology needed to manage them to ensure your business runs smoothly.