Five Criteria for Choosing an Enterprise Network Performance Management (NPM) Solution

Enterprises understand that their network is no longer a collection of technology assets but a critical component of their business strategy. In order for IT to provide strategic value to the business, you need tools that proactively ensure the reliability of network service while simultaneously boosting user experience. This requires a new approach to network management, including selecting a networkmanagement vendor that can meet modern IT and business requirements.

To select network performance management (NPM) solutions, enterprises must push beyond the traditional selection criteria, which focus on threshold-based alerts and alarms.

If your team is considering an NPM platform—including LiveAction, SolarWinds, Riverbed, or NetScout— before deciding on your vendor, consider these five criteria.

1. Comprehensive end-to-end network visibility and performance management – Is the solution able to monitor the entire network, including SD-WAN, LAN, cloud, wireless, campus, and datacenter, or only specific categories and environments?

2. Network traffic analysis – Does the platform have deep packet capture and analytics to monitor realtime network traffic and application performance?

3. Application visibility and performance monitoring – Can the platform analyze and correlate data from network devices, applications, and cloud environments to manage the entire digital experience?

4. Enterprise-scale – Can the solution monitor enterprise networks in organizations with more than 20,000 devices without performance issues or latency?

5. AIOps, analytics, and visualizations – Does the solution incorporate AIOps for advanced anomaly detection and correlation and visualizations that enable easy understanding of monitored environments?

1. Is the solution able to monitor and manage the entire network or only specific categories or environments?

Does the platform manage the entire digital experience and monitor and manage the cloud, wireless, SD-WAN, LAN, application, campus, and datacenter environments?

The modern IT infrastructure is more complex than ever and encompasses a range of components that work individually and collectively to improve network performance, support productivity, and ensure a positive digital experience. While local technology stacks have increased in density, companies have also expanded their infrastructure into third-party SaaS applications that make it a challenge to maintain comprehensive visibility. Let’s suppose a single component is suffering from performance issues. The effects can be wide-reaching, and the catalyst of the resulting problems can be challenging to pinpoint across such a wide range of moving parts. This is why modern NPM solutions must be able to not only monitor these individual components and environments in a single solution but also correlate collected performance data to quickly identify root causes and understand their potential effects on individual environments and the overall digital experience.

An effective modern NPM solution needs to collect and correlate performance data from the entire network, often from highly complex hybrid environments. This includes monitoring all types of network devices used, wireless components, SD-WAN, LAN, cloud environments, customer and enterprise applications, VoIP devices, and the datacenter. The solution needs to collect and analyze data not only for root-cause and performance analytics but also proactive health metrics. Key health analytics include top network users, availability, common traffic patterns that contribute to performance issues, application jitter, latency, and loss. Finally, NPM solutions should automatically create baseline and trending metrics to ensure that capacity issues do not contribute to downtime or performance issues.

2. Does the vendor’s platform support granular, correlated network-traffic insights?

Does the vendor’s platform have the ability to correlate traffic insights with application performance and user experience?

Whether users are accessing applications hosted internally or in the cloud, an NPM tool should correlate traffic data in real time with application performance and end-user experience. Doing this enables network operations teams to be much more efficient. Rather than analyzing every fault that the monitoring tools detect, engineers can focus their attention on problems that affect end users. This correlation can also reduce false positives and alarm storms.

The platform you choose should deliver insights into application performance and end-user experience through deep real-time processing and packet-by-packet analysis. A tool should present network transactions with performance indicators, including the ability to support views into complex, multi-tiered applications. These transactions should be correlated with end-user sessions and applications. Ideally, the solution will show the engineer how critical applications perform on the network and how users are experiencing individual sessions. This visibility should enable network operations teams to quickly isolate problems, particularly if the tool combines insight into end-user experience with awareness of the network domain.

3. Does the solution enable complete application visibility and performance monitoring?

Does the solution analyze and correlate data from network devices, applications, and cloud environments in a single platform? Does the solution support monitoring the complete digital experience, including the synthetic user experience?

Today’s solutions are required to support a seamless, high-performance digital experience. Solutions need to gather network-performance metrics from infrastructure devices, including routers, firewalls, load balancers, switches, and application-performance enriched flow data to create a comprehensive application impact analysis. The solution should support integrated application visualizations, including application-path analytics, by having the ability to alert on application-performance issues caused by network-device issues. The NPM solution should also be able to:

  • Visualize packet-by-packet to perform detailed application-performance analysis.
  • Analyze packet data to identify application errors in packet payloads.
  • Pinpoint the source of latency as network versus application.
  • Provide application usage and performance data at the site level.
  • Report top application performance highlighting average application and network delays.

4. Can the solution monitor enterprise networks in organizations with more than 20,000 devices without performance issues or latency?

Does the solution complete large-scale performance management across numerous types of devices and environments without latency and in a highly secure manner? Can the solution analyze capacity to help plan for future requirements to ensure that the network is not over- or under-provisioned?

Modern NPM platforms need 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. NPM should also support capacity planning to avoid both over- and under-provisioning of resources and predict whether a network can support growing business-critical traffic.

5. Does the solution incorporate AIOps to enable advanced anomaly detection and correlation and visualizations that aid easy understanding of monitored environments?

Because scale-related performance is critical, modern NPM solutions should incorporate machine-learning techniques to enable the platform to continuously learn and apply knowledge based on big-data trends. This includes the ability to create dynamic baselines and identify anomalous behavior from multiple sources of raw data. Machine-learning algorithms should support critical performance corrections, including determining which voice traffic to prioritize, when to throttle bandwidth, and whether a user’s access should be blocked.