Corvil Unifies Application and Network Latency Management

Corvil has added an application layer to its network latency management system, claiming it is first to market with a unified performance management solution dedicated to latency. The move brings Corvil into an area of the latency management space heretofore dominated by the likes of TS Associates.

After a year of in-house development at the company’s R&D base in Dublin and two months of beta testing, Corvil’s application latency module is available today as an upgrade to CorvilNet appliances. It will typically cost between 10% and 25% of the cost of existing Corvil kit.

Donal O’Sullivan, Corvil vice president of product management, says: “There has been a gap in the performance picture, with traders being interested in orders and the speed of market data, and infrastructure people being interested in bytes and packets. If there is a problem in the system, perhaps trading slows down, it takes infrastructure people hours to investigate several sources and identify what has gone wrong.

“Our solution is unified – not integrated – and includes real-time alerting, which means problems can be fixed quickly whether they are in the network or application layer. Or, depending on the problem, traders can move to a different market or service provider. Our approach is to unify business and technical solutions.”

CorvilNet’s added application capabilities allow users to monitor, troubleshoot and report on latency compliance for market data and order execution at the level of individual messages and transactions. Support is included for application processing and transforming infrastructure components, such as feed handlers, smart order routers, FIX gateways and matching engines, while latency measurement of correlated events at the application layer is also possible.

“Part of our value proposition is in how CorvilNet monitors and analyses data,” O’Sullivan says. “Like other performance management solutions ours includes probes in the client’s infrastructure, but while some solutions remove data and analyse it remotely, which is neither quick nor easily scalable, our solution uses a distributed model that analyses data close to the point of monitoring to deliver a picture of performance management quickly.”

Corvil reckons it is first to market with a unified application and network latency management system because of its focus on latency, but acknowledges that other suppliers, such as NetQOS and NetScout, are trying to break into the application space as application management specialists look to add a network layer.

O’Sullivan says beta testing of the unified solution was popular, with more customers than usual signing up to take part. He expects many of the beta testers to deal on the product soon and expects all of the company’s customers in the US, Europe and Asia to sign up over the next six months.

Global banks, hedge funds and stock exchanges, where application performance is a top priority, are expected to lead the way in adding the application module to their network view, with service and connectivity providers following as they seek to match their customers’ performance management systems. About half of Corvil’s appliances are already collocated or in proximity to trading platforms, with O’Sullivan suggesting that early sales of the unified solution will go into collocation environments.

“In the past, customers have used CorvilNet and another supplier or in-house developed software for application management,” he says. “But it is quite expensive to buy two products and put them together. Then they both tell you different things, making it difficult to know what exactly is happening in a system.

“Our customers wanted a single solution for both network and application latency monitoring. A single solution reduces capital and operating cost, while being able to identify and resolve latency problems that multiple, separate application, network and data capture solutions cannot.”