Savvius has released a new troubleshooting appliance called Spotlight, which analyzes network performance in real time based on traffic flows.
Spotlight is designed to help network administrators identify network and application latency and track performance of TCP and VoIP in real time. To speed up problem identification, Spotlight’s UI is configured to report on the worst-performing parts of the network.
Spotlight can be deployed as a 1U standalone appliance for flow-based analysis, or run on top of Savvius’s Omnipliance Ultra packet capture devices, including the T300, M200, and C100 models. This combination provides the real-time analysis in addition to full packet capture at up to 20Gbps, depending on the appliance model.
If a customer wants to run a deeper analysis using full packet data, they can run Savvius’s Omnipeek software against packets captured on the Ominipliance Ultra.
By combining flow-based analytics and packet capture, Savvius positions Spotlight as a tool for both real-time troubleshooting and post-event forensic analysis, with the goal of reducing the need to turn to full packet analysis to resolve problems.
“We’re reducing the number of times you might have to do forensic search,” said Jay Botelh, Director of Products at Savvius. “You can see from a flow perspective if you need more detail or if you can draw conclusions in real time.”
Customers can also export Spotlight data to a database for long-term analysis and baselining.
It makes sense for Savvius to expand its portfolio to help customers address performance issues where full packet analysis may be overkill. The UI I saw in a demo looks clean and sensibly arranged, and customers can easily drill in from the main screen to get more details on a particular network segment, source/destination addresses, protocols, and other key information.
Spotlight also enables Savvius to push back against competitors such as ExtraHop, which touts its real-time analytics based on packet metadata as a key differentiator.
Spotlight is available and shipping.