Big data and analytics are hot topics of conversation for almost anyone in IT today, including network operations. This is one of the reasons Gigamon has been on a tear over the past couple of years, especially since its IPO last year.
Last week, Gigamon announced an upcoming application to generate and export NetFlow records from its visibility fabric. The NetFlow Generation application will create NetFlow records and then send that information to one of the many NetFlow collectors and analyzers available on the market today.
Historically, Gigamon has traditionally focused on developing features and applications to help optimize the performance of network tools. This application, though, will help optimize the performance of network infrastructure, such as routers and switches. Generating NetFlow traffic can be very processor-intensive and offloading this to the visibility fabric can reduce the burden on network hardware.
NetFlow is a feature that was originally introduced by Cisco and is a simple way of collecting basic information about the network. By analyzing the data, NetFlow analysis tools can determine things such as source and destination of the traffic, class of service, usage patterns, top talkers and causes of congestion. Over time, other vendors have added support for NetFlow, although many have created their unique branding, such as Juniper’s J-Flow and Brocade’s S-Flow.
Gigamon customers should benefit greatly from the NetFlow Generation application because it not only offloads this process from the network, but also centralizes the generation of the traffic and can generate traffic even for high-throughput rates, which some network devices struggle with. Additionally, because the NetFlow generation is integrated into the Gigamon Visibility Fabric, customers can apply many of the Gigamon features, such as filtering and replication, providing greater control over the data.
Gigamon NetFlow Generation should help improve the quality of data as well. Because of the overhead on network devices, some network devices could actually drop production traffic. As a result, many of the products or inline devices use a sampling methodology to generate the data. A low sampling rate, such as 1 in 100 or, in some cases, 1 in 1000, can result in missing out on events that happen on the network.
In a sense, this application now lets network managers have their cake and eat it too, as the Gigamon infrastructure can be used for both packet and flow-level visibility, improving the performance and efficiency of the network monitoring tools that customers have already deployed.
NetFlow is a great, low-cost, high-value way of monitoring and analyzing the network and applications. However, many organizations shy away from turning on NetFlow because of the performance tax on the network devices. Gigamon’s NetFlow Generation application largely solves this problem and enables customers to get more value out of the management dollars already spent.