Monitoring performance where the action is: on the network edge
Intelligence is spreading out in organizations--moving closer to the customer, closer to customer-facing employees, and closer to the people managing business operations. Users at the edge are able to access the latest data, run sophisticated applications, and complete critical business tasks. This underlying trend drives edge computing. Remote sites are no longer very “remote.” They are just as plugged in as the core of the business.
As we distribute critical processing to the enterprise or network edge, we also have a responsibility to ensure our remote sites and branches are adequately monitored. We do not want edge computing to increase the risk of security threats or complicate the resolution of performance issues. For this reason, monitoring the network edge is an important and high-priority challenge for IT operations.
Critical incidents at the edge
As more processing moves to the edge, the consequences of a slow response can become critical. Think of a robotic surgical device or a connected car receiving input from sensors and cameras. If a service interruption impacts safety, service quality, or company reputation, CxOs will demand to know why performance issues were not caught in advance. IT can prevent surprises by making sure that monitoring at the edge is as effective as monitoring at the core.
Risks associated with edge computing
According to a survey by Dimensional Research, the majority of network professionals cite complexity as their top challenge in managing branch networks (68%), followed by increasing costs (58%) and security risk (49%). The growing importance of the network edge certainly poses challenges for IT.
Network complexity. In distributed environments, data moves over longer distances and over diverse networks, protocols, and infrastructure platforms. This additional complexity creates more potential for performance glitches. Complexity is the enemy of quick resolution, as it takes IT longer to isolate anomalies and rule out possibilities.
Larger attack surface. The use of shared infrastructure such as the internet or cloud services increases the attack surface IT must defend. The same survey found 74% of respondents used direct internet connections at their branches.. In addition, edge devices typically run software to allow remote access by IT administrators. This also increases the attack surface. Unchanged default passwords have often been used by hackers to gain unauthorized access to an enterprise.
Compliance risk. When the attack surface increases so does the scope of managing compliance. Every new branch or remote location introduces additional risk that must be mitigated and documented.
Cost of monitoring. Costs are driven, in part, by time it takes to detect and resolve an issue. Few enterprises employ networking experts at remote sites. Problems that are frequently site-specific must be detected and corrected by staff located elsewhere with minimal understanding of the remote site’s topology and risks.
Address the challenges of edge monitoring
Monitoring is the only way to ensure critical services are available and performing well at the edge. Network managers need edge strategies and solutions to deal with:
Data collection. The biggest challenge of monitoring (in any network location) is getting access to the right data—data that will let you identify and characterize performance issues. To do this, you must eliminate any blind spots in network visibility due to virtualized or cloud infrastructure. Total packet capture is important. Incomplete data can obscure the root cause of an issue and delay resolution.
Latency. Another challenge for IT is being able to prioritize and accelerate resolution of critical issues. Transporting data back to the core for monitoring is subject to delays and packet loss. As the edge grows more critical, the risk of a delayed response may be too great. Data traffic must travel over networks that are not under the direct control of the enterprise and are subject to delays and packet loss.
Cost. How can an enterprise afford to deploy best-in-class monitoring solutions at every remote or branch site? As the network expands and data volume grows, the cost of monitoring tools can outstrip the budget. The complexity of a distributed network makes it hard to predict where problems will emerge. As a result, IT needs solutions that enable comprehensive monitoring at a reasonable cost.
Adjust your network monitoring strategy
The path forward will require adjustments to an organization’s network monitoring strategy. Consider the following to maintain performance and security at the edge:
Capture data at edge: Make sure you are actively watching what happens to user experience at the network edge. To enable timely and accurate performance analysis, experts rely on packet and flow data. This information offers details that are useful for root cause identification. Some organizations store packets data lakes (for a defined period) to have them available for faster problem isolation and resolution.
Filter data to increase efficiency. Your monitoring tools will work more efficiently if you first aggregate, deduplicate, and filter the data they process. A network packet broker (NPB) performs these functions and can also decrypt secure traffic, strip away unnecessary data, generate NetFlow, and automatically balance workload across multiple tools. An NPB lets you gather data from multiple network segments and monitor with a centralized set of tools, to boost return on tool investment. NPBs are available as special-purpose devices or as software deployed on a white-box server.
Integrate data from the edge to the core. There are advantages to using the same network visibility platform to see data flow from the edge to the core. Having a single interface to set up filtering and delivery of data to monitoring tools will save you time and be easier for the network operations team to use.
Blend in proactive monitoring. You can anticipate performance issues at the edge with proactive monitoring. In this approach, you run simulated traffic through your network infrastructure and observe key performance metrics such as response time, bandwidth, and packet loss. Continuous proactive monitoring gives you a heads up when conditions deteriorate and lets you make adjustments before users are affected.
As critical business processing migrates to the network edge, you need to quickly identify, isolate, and resolve issues before users are affected. Make sure you are actively monitoring user experience and ensuring adequate service delivery. Find out more about monitoring the network edge using Ixia Hawkeye performance monitoring platform and Ixia Vision Edge network packet brokers.