Tolly Talks Network Monitoring and Reliability of Source Data
The Tolly Group reveals key performance benchmarks as high-speed networks underscore need for greater integrity in network monitoring.
In today’s networks where 40 Gb links are becoming more and more common, simply “capturing” network data is half the battle. It must also be effectively analyzed to provide actionable insights. The caveat is that network performance monitoring insights can only be truly accurate when the underlying data is also accurate and complete. For those of us focused on networking, network analysis, and network security forensics, an endless stream of packets large and small constitute the data that must be captured.
This is important not only to build the big picture through intuitive dashboards and visual aids like baselining but to enable a detailed drilldown to evaluate, if needed, every single packet, especially in cases where network security forensics might be required. That’s why metadata for high-level reporting and individual network conversations, through each and every packet, is so important. Getting accurate data from today’s networks requires a very high-performance system and that is what we found when benchmarking the Viavi Observer GigaStor solution.
Putting the Tools to the Test
The GigaStor 288T was tested using 8×10 GbE and 2×40 GbE network connections and was able to sustain lossless capture to disk at a rate of 41.6 Gbps across either network configuration. Sustained capture rate was benchmarked on a packet level to be 14.3 million packets per second. Our test also benchmarked the highest burst-data rate that the GigaStor 288T could capture for an interval of 30 seconds. That test verified a rate of 60 Gbps.
Captured traffic will likely contain sensitive and confidential information that must be protected from unauthorized use. Viavi offers AES-256 encryption of stored data to provide this protection. The performance tests were run with disk encryption enabled and verified that the performance results remained unchanged. The tests used a sophisticated Ixia network test system to generate the traffic. The claim of “no dropped packets” was validated by comparing transmit/receive statistics of the Ixia solution with the packet counters on the Viavi Observer Performance Management Platform. Tolly engineers verified that the two sets of counters matched.
All of the data and detailed results can be found in the Tolly Report available from Viavi Solutions as well as Tolly document #217122 from Tolly.com. To learn more about network performance management strategies for effective enterprise service delivery and end-user experience, register now for the Network Performance Management Trilogy webinar where the heroes of IT will be armed with the tactical knowledge they need to manage today’s networks effectively.