Five Ways to Improve Network Performance with Active Monitoring

What will happen to application performance if you migrate to cloud infrastructure? What will happen to user experience if traffic increases? You don’t have to wait for users to let you know. Read on this article to learn how you can use active monitoring to improve network performance.

Five Ways to Improve Network Performance with Active Monitoring
Five Ways to Improve Network Performance with Active Monitoring

Anticipate and prevent performance issues, validate your hybrid infrastructure, and accelerate network troubleshooting when you use synthetic traffic simulations. Keep your network running at peak performance. This article gain insight into how active monitoring provides advantages you can’t get from the traditional reactive approach.

Well-maintained infrastructure and real-time network monitoring are not always enough to keep things running smoothly and head off trouble. You need intelligence that can help you validate performance before deploying new infrastructure or applications, observe the end-to-end performance that users experience, and anticipate how changes in your environment will impact performance and service quality.

An active or proactive monitoring platform is designed to help you answer what-if questions and identify issues before they impact users. The data you gather during network simulations lets you resolve issues in advance and reduce the risk of network and service disruptions.

Content Summary

Introduction
Active Monitoring Is Proactive, Not Reactive
Active Monitoring Is Built for ‘What-If’ Analysis
Function of Endpoints
Types of Simulations
Five Ways to Improve Network Monitoring
One: Take an End User Perspective
Two: Use Synthetic Traffic Simulations
Three: Measure Performance Across Hybrid IT Resources
Four: Validate Network Readiness
Five: Accelerate Root Cause Analysis
Conclusion

Introduction

A secure, high-performance network is critical to an organization’s success and long-term viability. Network engineers have a major responsibility, particularly as IT transformation continues and increases complexity. Well-maintained infrastructure and real-time network monitoring are not always enough to keep things running smoothly and head off trouble. You need insight into questions such as:

  • Will our current network infrastructure perform effectively if we add 100 more employees?
  • Will the performance of our newly-deployed customer-facing application meet customer expectations?
  • Can the network handle the entire salesforce joining the annual sales meeting live from their desktop computers?

Even the best network monitoring tools are not designed to answer ‘what if’ questions like these. Fortunately, there is a way to analyze potential changes in your environment using an approach called active or proactive monitoring.

‘What-if’ scenarios help you prepare your network for change and let you actively reduce the risk of downtime and application outage.

Active Monitoring Is Proactive, Not Reactive

Traditional network monitoring is reactive. Your tools gather data from your production network and alert you when performance falls short. In the best case, you identify and resolve an issue before it affects end users. However, more often than not, users notice a problem before you do. You end up scrambling to figure out what happened and get a fix in place.

Traditional monitoring also does not help you assess network readiness, which is critical to the success of new services and applications. For example, you know that migrating to the cloud will change traffic flow considerably, but how do you know for sure if you need to upgrade your network equipment? With a limited budget, it can be stressful to make that decision without having data to rely on.

Active monitoring—also known as proactive monitoring—is different because it gives you quantitative data about what could happen to your live network as conditions change. The information enables you to proactively make changes in your network to prevent performance issues from occurring. Using this approach, you work actively to reduce the risk of downtime, application outage, and dissatisfied customers.

An active monitoring platform models changes in traffic patterns to show you what will happen on your network.

Active Monitoring Is Built for ‘What-If’ Analysis

An active monitoring solution looks for potential performance issues by running synthetic network traffic through your network infrastructure and observing what happens on your network.

An active monitoring solution has two important functions:

  1. generates highly realistic, synthetic traffic and injects it into your live network in test mode
  2. monitors network functions during the simulation by collecting key performance metrics and visually displaying the results

You design and configure the simulations to answer important ‘what-if’ questions. For example, you can model traffic from a new group of users in a specific location, or you can change the mix of traffic to include a new application type or an application with specific requirements, such as video.

Active Monitoring Is Built for ‘What-If’ Analysis
Active Monitoring Is Built for ‘What-If’ Analysis

Function of Endpoints

Hardware- and/or software-based endpoints generate the synthetic traffic in active monitoring. The endpoints mimic the function of nodes in your network and you can deploy endpoints at every location where you want to measure network performance. Figure 1 below shows the most common types of endpoints.

Figure 1. The function of endpoints in active monitoring
Figure 1. The function of endpoints in active monitoring

Types of Simulations

The active monitoring solution injects the synthetic traffic into your network according to a profile you create. Each profile represents a ‘what-if’ scenario and the data collected during the simulation demonstrates the potential response of your network under the profile’s specific conditions.

Simulations provide insight that enable you to:

  • validate performance before going live with new applications or migrating a workload to infrastructure in a different location, such as the cloud
  • monitor the end-to-end performance over a diverse and distributed network, such as a LAN with Wi-Fi access points
  • ensure the quality of voice and video transmissions at scale
  • confirm service providers are meeting service level agreements (SLAs)
  • track performance at the network edge as use of powerful applications at branch and remote locations increases

Five Ways to Improve Network Monitoring

One: Take an End User Perspective

Traditionally, operations engineers collect data from network devices and management systems to assess uptime. Downticks may indicate the need to upgrade equipment or signal conditions that will cause performance to degrade over time. While uptime monitoring is useful, it only reflects the availability of the element being polled, not the overall user experience.

Take an active approach: When you shift to an active approach, you observe performance from the end user’s perspective, which involves performance across the entire chain of networked elements as shown in Figure 2.

Figure 2. The customer experience depends on a complex chain of networked elements
Figure 2. The customer experience depends on a complex chain of networked elements

The synthetic transactions of active monitoring let you monitor availability from the perspective of the end user.

In active monitoring, you probe network elements using regular network and application layer pings that form a “heartbeat” to track responsiveness. The pings create synthetic transactions which are used to monitor availability from an end user perspective, rather than from an equipment perspective. You can position proactive monitoring probes in the same location as end users to more accurately detect issues that impact them. This type of result cannot be observed by simply pinging individual components.

The ‘what-if’ scenarios you test let you see what happens to customer experience when the volume, type, mix, or geographical location of your network traffic changes. Active monitoring platforms track how traffic moves through the entire network topology. You can monitor key performance metrics on connectivity, network infrastructure, network services, calls out to the internet, and communications with cloud-based application and service providers.

Benefit: You see availability and uptime from the end user perspective.

Two: Use Synthetic Traffic Simulations

The typical IT team uses multiple network performance monitoring (NPM) and application performance monitoring (APM) tools to analyze performance and generate alerts when conditions exceed certain thresholds. The tools use copies of actual user traffic collected from the network via taps or SPAN ports. Analysis generally takes place as transactions occur. The problem with this approach is that it focuses on detecting issues that are already happening. It does not help you anticipate problems and correct vulnerabilities before conditions deteriorate and impact the business.

Take an active approach: An active monitoring platform enables you to expand how you use your existing NPM and APM tools and to perform “what-if” analyses. You can generate highly realistic, synthetic traffic to run through your network infrastructure and simulate the behavior of actual users. The platform provides you with information on throughput, latency, packet loss, and quality of service in the scenarios you build.

The active approach, which uses traffic simulation, lets you build models that closely mimic what you expect to see moving across your network. You can easily observe what happens to performance if you introduce a new application or expand your use of an existing one. The platform also provides you with an extensive library of common application signatures to quickly generate simulated traffic without spending time on manual programming.

An active monitoring solution transforms your NPM and APM tools into powerful what-if analysis engines.

The process for effectively using an active monitoring platform:

  • Survey your environment
  • Model expected usage patterns
  • Simulate anticipated traffic
  • Analyze network characteristics

The proactive approach also allows you to gather performance metrics without having any actual users on the network. You have the freedom to run a simulation whenever you choose and at regular intervals, thereby helping make continuous monitoring a best practice.

Benefit: You can anticipate and prevent performance issues using what-if scenarios.

Three: Measure Performance Across Hybrid IT Resources

Monitoring performance in a hybrid environment introduces additional complexity and cost. For one, you will need a mix of physical, virtual, and cloud-based taps to access traffic and gather metadata. You also must ensure the security of data that is transferred back to the data center for analysis. Another challenge is correlating data from diverse network segments. Finally, deploying passive probes in remote locations is typically costly and there is no way to test remote access before going live.

Take an active approach: Active monitoring gives you the flexibility to measure performance of applications running on hybrid network infrastructure. Endpoints that you deploy throughout your network generate synthetic traffic and reproduce the behavior of end users accessing cloud resources. Simulations let you gather hop-by-hop and overall performance metrics on a new service or application before service turn-up.

An active platform also enables you to run regularly scheduled monitoring simulations between any two locations or endpoints in your network. The results provide you with time-based data on uptime, throughput, delay, packet loss, and other indicators, so you can easily see trends and spot deviations. Continuous monitoring gives you the insight you need to fine-tune your hybrid network and keep things running smoothly.

Figure 4. Active monitoring generates synthetic traffic within various clouds and virtual machines
Figure 4. Active monitoring generates synthetic traffic within various clouds and virtual machines

Active monitoring also helps you manage performance as critical business traffic moves to the network edge. Simulations measure performance between remote endpoints and the cloud. Network planners can also compare the performance impact of different network designs and topologies.

Benefit: Active endpoints measure performance between any two elements in a hybrid network.

Four: Validate Network Readiness

Traditional monitoring takes place on your live production network and focuses on quickly identifying and solving performance issues. If you want to reduce the risk that a performance issue will result from an anticipated change in infrastructure or traffic patterns, you need to take a different approach.

Take an active approach: Active monitoring lets you be more proactive and validate if your network is sufficiently prepared to handle upcoming changes. You construct a scenario reflecting the anticipated changes, run the test, and the platform provides you with the resulting performance metrics.

For example, you can see the impact to your network of deploying SD-WAN or migrating an application to the cloud. When the results fail to meet your expectations, you can take preventative action to change the results before the actual deployment. Active monitoring lets you substantially reduce the risk of network disruption or application downtime by validating network performance under new conditions. Validation assessments are also useful when you are planning infrastructure upgrades or choosing the appropriate service level from a third-party provider.

Benefit: You can validate the readiness of network infrastructure before production rollout.

Five: Accelerate Root Cause Analysis

Network monitoring is a critical tool for troubleshooting, but a reactive approach has several drawbacks. One is the time it takes engineers to understand how traffic actually moves through their network. The use of cloud resources, third-party SaaS applications, and mobile devices means packets enter and leave the network from a wide variety of ingress and egress points. The standard path discovery tool used by network engineers is ‘traceroute.’ But using ‘traceroute’ takes time, skill, and is often insufficient for identifying the root cause of an issue.

A second issue is that post-event analysis to isolate the source of an issue requires significant expertise, which often delays resolution.

Take an active approach: An active monitoring platform can reduce the time it takes to understand how packets move through your network. The path discovery feature automatically produces maps like Figure 5 below, which shows the actual path traffic takes through the network.

Engineers can reproduce network issues after a fix, in order to verify a resolution and prevent issues from recurring.

Figure 5. Example of path discovery resulting from active monitoring
Figure 5. Example of path discovery resulting from active monitoring

This hop-by-hop visual display eliminates the need to manually prepare and maintain topology maps, and also provides a readily-available starting point for troubleshooting. Path maps are always up-to-date and accurate when active monitoring happens on a continuous basis.

An active monitoring platform also lets you reproduce the traffic patterns and conditions that existed just before an issue or outage occurred and replay them on-demand. This function lets you change the settings on your NPM and APM tools and replay what happened to see if your tools react differently.

Deep expertise is not needed to interpret the results of a ‘what-if’ scenario. You simply specify the performance threshold you want to achieve, run the simulation, and the platform indicates ‘pass’ or ‘fail.’ Problem resolution moves forward more quickly.

Support engineers also can use the replay function to validate solutions they implement in the live network and increase confidence in the fix. Avoiding the recurrence of issues makes troubleshooting more efficient.

Benefit: You can accelerate issue identification and verify issue resolution.

Conclusion

When your business depends on peak performance, active monitoring delivers additional advantages you can’t get from a traditional reactive approach. Give your team a flexible, cost-efficient platform to run “what-if” scenarios and traffic simulations to identify vulnerabilities before they impact end users. Whether you need to assess end user experience, prepare your network for workload changes, validate your hybrid infrastructure, or increase the efficiency of network troubleshooting, active monitoring lets you develop deeper insight into network performance.

The result is stronger networks, more consistent application performance, and an excellent user experience.

Source: Ixia