.shock - Fotolia


Apply network analytics to more quickly solve service issues

With little tolerance for unplanned downtime, IT is under pressure. Applying network analytics can help troubleshoot problems faster and more accurately, and prevent service issues down the line.

In today's ultracompetitive environment, businesses are reliant on their IT infrastructures to get the job done. Excessive downtime is unacceptable, and for many end users, any downtime is unacceptable. When the network goes down, the business comes to a standstill, as employee productivity plummets, communications are severed and customer service falls off a cliff. As a consequence, IT is in a constant race to get out ahead of hardware failures, routing problems, power outages, traffic congestion, security attacks, configuration errors and other issues that can knock the network offline. Simply put, with the cost of downtime pegged at $5,000 per minute and up, organizations can ill afford to have their networks go dark.

Complicating IT's challenge is that it bears the burden of keeping the network up with limited resources. As a result, IT is always searching for effective network management tools that can accommodate today's sprawling and constantly evolving networks. Applying network analytics can help.

To keep pace with the demand for tools that allow IT teams to do more with less, network management vendors are cloaking their platforms with more sophisticated capabilities that rely on automation. Increasingly, suppliers are capitalizing on advances in network analytics and cognitive computing, and incorporating these technologies into their products to help network managers detect and resolve incidents more quickly. These applications also promise more proactive network management capabilities by giving IT deeper insights into statistics, permitting engineers to tweak their networks as needed to optimize performance.

Ahead of the game

The hope is that by bringing together security analytics, topology data and real-time performance analytics, IT managers can get a better grasp of the entire enterprise environment.

The network itself has always been a rich source of the data IT needs to isolate or diagnose the source of a problem. The challenge in applying network analytics has been trying to comb through volumes of traffic statistics to find the root cause of an incident.

Analytics, cognitive computing and advanced machine learning provide mechanisms to correlate information between data sets to make determinations about the underlying source of an issue. Network analytics can put the entire network in context with data gathered from across the enterprise. This information can be invaluable in helping find a problem in near real-time and fixing it quickly. IT can also use trend analytics to determine if there is a better way to route traffic or eliminate congestion.

To that end, a number of network management vendors and other suppliers are adding analytics to their platforms to help customers minimize downtime. NTT unit Virtela Technology Services, for example, uses analytics within its LAN/WAN infrastructure management service to make troubleshooting easier and traffic routing more efficient. ExtraHop Networks, meantime, added IT operations analytics to its network management platform to provide faster incident response and to help organizations better assess traffic patterns to determine the most efficient and cost-effective way to provision the network. And Hewlett Packard Enterprise's IMC network traffic analyzer tool provides graphical views of network analytics so IT can quickly spot the source of an issue and also identify performance-related issues.

Consider the source

As valuable as network wire data is, there are also other sources like log files that can add insight into the genesis of critical events. Consolidating all of this discrete data means network management systems have to find a way to standardize and correlate all of the different types of information they collect.

Enter the Open Networking User Group (ONUG). ONUG has a group that is trying to find a method to pull all of this data together in a cohesive fashion. The hope is that by bringing together security analytics, topology data and real-time performance analytics, IT managers can get a better grasp of the entire enterprise environment. This will help organizations not only be quicker in isolating the source of a problem, but also give them the information they need to make a predictive assessment. The tools deployed for applying network analytics will provide the kind of insight that IT needs to practice true and proactive operations management.

Next Steps

Getting the most out of network analytics 2.0

Introducing security analysis tools into the enterprise

Combining security and big data to prevent security issues

This was last published in February 2016

Dig Deeper on Network management software and network analytics