Corvil » Solutions

Electronic Market Centers

1. Solution Overview

To attract and retain liquidity from automated trading clients, market centers aim to provide a high performance, low latency market data and trading platform. This requires managing latency and throughput across diverse systems, ranging from matching engines to the trading venue core network, to proximity services and client WAN connections. The ongoing process of performance management must simultaneously keep abreast of ever rising customer expectations and ever increasing traffic volumes.

CorvilNet is designed to ease the challenge of latency management throughout a live distributed trading system. It continuously monitors latency and traffic load at highly granular timescales, verifies that programmed service level targets are met, and deploys advanced troubleshooting and optimization capabilities when they are not. CorvilNet presents a comprehensive latency management feature set, including high-speed always-on packet capture, microsecond precision latency measurement, and advanced message and network-layer analytics.

Uniquely, CorvilNet correlates application behaviour either within or between trading sites and interconnected parties, using patented techniques that are lightweight, scalable and secure. These capabilities enable an end-to-end performance view that eliminates the important blind-spots afflicting electronic trading today.

Example applications of CorvilNet include:

2. Market Data Distribution

Manage latency for market data traffic across your internal networks and systems, using CorvilNet appliances deployed near feed aggregators and at the network edge.

• Track performance both end-to-end and segment-by-segment using precision microsecond latency measurement.
• Monitor latency in both LAN and WAN environments using CorvilNet's uniquely scalable architecture.
• Verify market data quality throughout the infrastructure with sequence gap detection and advanced TCP performance analytics.

3. Order Execution

Monitor transactional latency across the matching systems and across the network.

• Generate complete historical records with built-in latency analytics using always-on high speed packet capture.
• Track the end-user's performance experience using TCP metrics designed specifically for electronic trading.
• Diagnose issues caused by service provider performance, client connectivity problems, or trading system latency.

4. Connectivity Management

Track client's WAN connection performance and help them determine connectivity and bandwidth requirements.

• Determine bandwidth needs for specified latency and loss objectives using CorvilNet's automated bandwidth sizing capability.
• Identify bottleneck points and upgrade requirements for short timescale traffic loads revealed using microburst detection.
• Provide scalable and secure end-to-end performance visibility using CorvilNet's inter-party latency monitoring.

Market Providers

1. Solution Overview

Reliable low latency performance is the key to success for service providers inter-connecting financial traders, broker-dealers, and market centers. Implementing best-in-class latency management within the network helps service providers ensure that their systems perform at the timescales that matter for trading applications. It also creates the opportunity to offer performance transparency to trading clients, thereby enhancing the value of connectivity services and strengthening their high-performance credentials.

Deploying the CorvilNet latency management system provides full performance visibility across your network for every packet of client traffic. CorvilNet also tracks short-timescale traffic loads at network entry and exit points, revealing the latency impact of client traffic and helping to ensure their connections are appropriately sized for their performance requirements.

Uniquely, CorvilNet correlates application behavior either within or between trading sites and interconnected parties, using patented techniques that are lightweight, scalable and secure. By deploying managed CorvilNet appliances in the market centers and trading sites that you inter-connect, you can also provide a full end-to-end performance view that eliminates the important blind-spots afflicting market participants today.

Example applications of CorvilNet include:

2. Low Latency Network Monitoring

Remove any possible question mark over your network's ability to maintain highly competitive low latency SLAs for end users.

• Track microsecond latency and loss across geographically-dispersed networks, using CorvilNet's scalable latency measurement architecture. View all results using configurable, centralized dashboards.
• Continuously verify performance against low latency SLAs. If violations occur, use CorvilNet's automatically triggered smart packet captures and advanced analytics to quickly identify root cause.
• Use CorvilNet's highly granular, congestion-aware load analytics to identify bottleneck points and upgrade requirements, for network speeds up to 10G.

3. Market Data Distribution

Manage market data latency and quality from your connected market centers across your network, ensuring low latency market data delivery to end consumers.

• Identify traffic microbursts and bandwidth requirements at ingress points and throughout the network, using CorvilNet's microburst detection features.
• Monitor microsecond latency across networks, servers and feed-handlers, ensuring that each component achieves tight latency performance targets.
• Verify data quality at each stage with market data sequence gap detection and advanced TCP performance analytics.

4. Order Execution

Deliver a best-in-class order execution platform by tracking latency across each connected market center and for each of your connected clients.

• Track order entry request/response times from each CorvilNet monitoring point - at ingress points from clients, egress points towards market centers, and internal server co-lo points.
• Integrate CorvilNet's latency measurements directly into your own internal applications using the system's real-time web services interface.
• Generate complete records of order-entry performance using CorvilNet's high-speed always-on packet capture - capturing all measured latency and performance analytics alongside network packets.
• Monitor trading connection availability and end-to-end performance using advanced TCP analytics specifically designed for trading.

CorvilNet for High Frequency Trading

1. Solution Overview

High frequency traders rely on low latency market data and order/execution infrastructure to remain competitive. But as relevant timescales shrink from seconds to milliseconds and microseconds, sources of latency within the infrastructure become increasingly difficult to spot and eliminate using traditional monitoring tools. The absence of high quality performance data from live systems makes it hard for managers to guarantee the low latency trading environment that users demand. And as the equipment cost of staying at the leading-edge escalates, it becomes more difficult to justify expensive blanket upgrades that may - or may not - improve latency.

CorvilNet's pro-active approach to latency management at highly granular timescales provides the visibility and control you need to guarantee low latency performance, while avoiding excessive equipment and operational costs. Appliances deployed at strategic points continuously track the performance of your live systems, verifying that pre-specified service level targets are met, and deploying advanced troubleshooting and optimization capabilities when they are not.

CorvilNet, uniquely, correlates application behaviour either within or between trading sites and interconnected parties, using patented techniques that are lightweight, scalable and secure. These capabilities enable an end-to-end performance view that eliminates the important blind-spots afflicting market participants today.

Example applications of CorvilNet include:

2. Order Execution

Track the performance that orders and execution reports experience across internal networks and to/from the trading venue.

• Monitor end-to-end trading connection performance using specialized "TCP for Trading" analytics and order-entry response-time measurements.
• Review and compare performance across multiple venues and infrastructure segments using CorvilNet's live centralized dashboard and service index.
• Use historical or live measured data to guide strategy development and execution, obtained direct from CorvilNet's packet capture and live XML API features.

3. Market Data Distribution

Ensure that data is distributed across internal networks and systems with minimal latency, using CorvilNet appliances deployed near inbound feeds and near key market data consumers.

• Track performance both end-to-end and segment-by-segment using precision microsecond latency measurement.
• Identify bottleneck points and upgrade requirements for short timescale traffic loads revealed using microburst detection.
• Monitor market data quality with sequence gap detection at ingress points and throughout the infrastructure.
• Generate complete historical records with built-in latency and performance data using always-on high speed packet capture.

4. Co-Location/Proximity Services

Securely monitor latency and loss performance between remote trading sites and co-located systems.

• Gain complete visibility into WAN service provider performance using CorvilNet's uniquely scalable correlation architecture.
• Make informed deployment and co-location decisions based on quantitative latency data.
• Determine bandwidth needs for specified latency and loss objectives using CorvilNet's automated bandwidth sizing capability.

High Performance Wans

1. Solution Overview

The corporate WAN is playing an increasingly important role in providing remote users with fast access to business critical applications. To meet the business needs in this evolving environment, maintaining the performance of the WAN is becoming more challenging due to number of trends including:

• Consolidation of server and applications hosted in one or more data centres
• Accelerating deployment of web based applications
• Greater diversity in applications including Virtualization, VoIP and Video
• Greater use of cloud and shared infrastructure services to reduce cost
• Increasing LAN speeds…but access bandwidth not keeping pace (still too expensive)

CorvilNet's real-time approach to latency, jitter and loss measurements, and the delivery of all network and application analytics at highly granular timescales provides the visibility and control you need to guarantee a high performance WAN. CorvilNet appliances deployed at strategic points continuously track the performance of your live systems, verify compliance to pre-specified service level targets, and enable advanced troubleshooting and optimization capabilities.

Applications include:

Network Application Performance Monitoring

• Real-time applications such as VoIP and Video have strict latency, jitter and loss targets that must be achieved to ensure acceptable end-user performance. Remote desktop and web applications require fast network and server response-times to be used effectively. CorvilNet tracks both network and application performance with unsurpassed granularity, allowing you to quickly determine whether requirements are being met and to determine the root cause of any divergence from expectations.

Full Visibility and Reporting of WAN Usage

• All of CorvilNet measurements are based on monitoring every single application packet and reporting utilization and latency at microsecond timescales. Combined with CorvilNet layer-7 DPI application recognition engine, this provides complete visibility of all traffic and users of the WAN.

Network Service Level Compliance Reporting

• For corporate WANs that have chosen a managed WAN service, CorvilNet can provide an independent measure of the service levels delivered by the WAN. For WANs managed by the corporate IT staff, the network engineering team can utilize CorvilNet to show service level compliance from periods ranging from microseconds up to the last year.

Network Optimization

• Service providers strive to ensure adequate performance across their core network but often view WAN link performance as an issue for their customers to deal with. CorvilNet identifies congestion in the service provider versus WAN link congestion. For WAN link issues, CorvilNet delivers concrete recommendations to ensure the WAN is properly provisioned to deliver the required application service levels.

2. Key Functionality

• Monitor one-way end-to-end latency, jitter and loss for all packets of every application in the network including VoIP and Video
• Report all usage of the WAN with real-time microsecond granularity and historical usage for the past year
• Measure network service quality to report against service level objectives for every class of service in the network. Highlight all service level violations
• Quickly determine if the WAN (versus the application server, firewall, other infrastructure, etc.) is the cause of application performance issues
• Clearly identify service provider issues versus congestion on the WAN link when the WAN is identified as contributing to poor application performance
• Determine the required WAN link speeds and quality of service policy settings based on programmed service level targets. CorvilNet directly computes how much bandwidth your traffic needs to achieve your stated service level objectives
• Monitor application response times versus network response times to determine the cause of end-user perceived performance issues

3. Diagram

4. Key Benefits

• All applications are delivered to the remote users with the required service levels to meet the business needs
• Retain a complete historical record of service level compliance and be alerted whenever targets are violated
• Avoid unnecessary bandwidth upgrades and the associated costs
• Quality of service policies are easily configured without any specialized expertise or offline modelling required
• The root cause of any breaches in service level objectives can be quickly determined
• Locate, analyze and resolve service level problems faster and more reliably using CorvilNet's advanced event troubleshooting features.

Low Latency Data Center

1. Solution Overview

Data centers are evolving to meet rapidly escalating demands for higher levels of performance and resource virtualization. Low latency data centers typically involve a number of infrastructure components including:

• LAN switching fabrics for user access and general networking
• Inter-processor communications (IPC) fabric for low latency messaging between compute clusters
• Storage connections for access to shared storage/file resources

Each of these infrastructure components are potential sources of additional latency within the data center. Relevant timescales for measuring latency in these data center infrastructures are typically in terms of microseconds. At this timescale, these latencies are difficult to spot and eliminate using traditional monitoring tools. Equipment costs are escalating for leading-edge hardware required, making it difficult to justify a blanket upgrade that may - or may not - improve latency.

CorvilNet provides sub-microsecond precision one-way latency measurements for every single packet in the instrumented data center infrastructure. Microsecond latency measurements can be measured and reported for any segment of the data center including individual elements such as firewalls, switches, servers, etc. All latency measurements are calculated in real time with each appliance scaling to instrument up to four segments in the infrastructure. CorvilNet appliances continuously track the performance of your infrastructure, verify compliance to pre-specified service level targets, and enable advanced troubleshooting capabilities.

Applications in Low Latency Data Centers include:

Network Infrastructure Performance Monitoring

• Every element in the data center infrastructure has the potential to contribute additional unwanted latency in terms of milliseconds or microseconds. CorvilNet provides one-way latency measurements with sub-microsecond accuracy across any single element in the data center infrastructure. This provides unsurpassed granular measurements, allowing you to quickly determine whether requirements are being met, and to determine the root cause of any divergence from expectations.

Full Visibility and Reporting of Infrastructure Usage

• All of CorvilNet measurements are based on monitoring every single application packet and reporting utilization and latency at microsecond timescales. Combined with CorvilNet layer-7 DPI application recognition engine, this provides complete visibility of all traffic and users of the data center.

Infrastructure Service Level Compliance Reporting

• CorvilNet enables the specification of microsecond service level latency objectives. These can be measured across each instrumented element or segment of the data center infrastructure. Current performance versus these latency objectives can be reviewed at-a-glance using CorvilNet’s Service Index metric, while any SLA violations trigger automatic alerts and advanced troubleshooting analytics.

2. Key Functionality

• Monitor one-way latency, jitter and loss with sub-microsecond accuracy for all packets in the data center infrastructure
• Report all usage of the data center infrastructure with real-time microsecond granularity and layer 7 deep packet inspection
• Measure latency across individual elements or segments of the data center infrastructure
• Monitor every packet against precise latency and loss SLAs for each element or segment of the data center infrastructure
• Utilize SLA-Aware event-triggered smart packet captures that enable users to quickly identify the cause of programmed SLA violations
• Detect short timescale traffic spikes that cause excessive latency and quickly identify the application or traffic source

3. Diagram

4. Key Benefits

• Pro-actively manage latency at microsecond timescales for individual elements or segments of the data center infrastructure
• Control equipment costs by targeting upgrades where excessive latency has been quantitatively identified
• Resolve latency issues more quickly and more accurately
• Retain a complete historical record of service level compliance and be alerted whenever targets are violated
• Locate, analyze and resolve service level problems more quickly and more reliably using CorvilNet's advanced event troubleshooting features.

Unified Communications

1. Solution Overview

Unified Communications (UC) integrates multiple types of communications and applications across the network infrastructure with the goal of allowing people, teams, and organizations to communicate more effectively. This goal can only be achieved if the complete UC deployment is meeting user requirements for all types of communications. This requires the network infrastructure to be managed, provisioned and optimized to deliver the diverse set of requirements demanded by the UC environment.

A complete UC deployment consists of a number of modes of communications and applications each with unique networking challenges due to the varying traffic characteristics and user expectations. For example, the following UC applications all have different traffic characteristics and different performance requirements:

• VoIP and Video must be delivered within strict latency and jitter requirements to meet user expectations.
• Presence status can contain large amounts of data (ie: calendar and contact information) and updates must happen in near real time. This volume of data needs to be managed to mitigate network issues.
• Instant Messaging users are accustomed to quick response times that the network must be provisioned to deliver.
• Web-Based Conferencing and Collaboration demands near real time interactivity and strict latency and jitter targets for any corresponding audio traffic.

CorvilNet's real-time approach to latency, jitter and loss measurements, and the delivery of all network and application analytics at highly granular timescales provides the visibility and control you need to guarantee application delivery in a UC environment. CorvilNet appliances deployed at strategic points continuously track the performance of your live systems, verify compliance to pre-specified service level targets, and enable advanced troubleshooting and optimization capabilities.

Applications of CorvilNet in a UC environment include:

Network Application Performance Monitoring

• Real-time UC applications VoIP and Video have strict latency, jitter and loss targets that must be achieved to ensure acceptable end-user performance. Presence, Instant Messaging and Web-Based Conferencing/Collaboration require fast network and server response-times to be used effectively. CorvilNet tracks both network and application performance with unsurpassed granularity, allowing you to quickly determine whether requirements are being met and to determine the root cause of any divergence from expectations.

Full Visibility and Reporting of Network Usage

• All of CorvilNet measurements are based on monitoring every single application packet and reporting utilization and latency at microsecond timescales. Combined with CorvilNet layer-7 DPI application recognition engine, this provides complete visibility of all traffic and users of the network.

Network Optimization

• Deployment of UC applications require quality of service policies to be deployed throughout the network infrastructure. These are required to ensure the network delivers the diverse set of requirements demanded by the UC environment. CorvilNet delivers concrete recommendations to ensure each quality of service policy is properly provisioned to deliver the required application service levels.

2. Key Functionality

• Monitor one-way end-to-end latency, jitter and loss for all packets of every application in the UC environment including VoIP and Video
• Report all usage of the network with real-time microsecond granularity and historical usage for the past year
• Measure network service quality to report against service level objectives for every class of service in the network. Highlight all service level violations
• Quickly determine the cause of poorly performing UC applications using programmable event-triggered smart packet capture, analysis, decode and export
• Determine the required quality of service policy settings based on programmed service level targets. CorvilNet directly computes how much bandwidth your traffic needs to achieve your stated service level objectives
• Monitor application response times versus network response times to determine the cause of end-user perceived performance issues

3. Diagram

4. Key Benefits

• All UC applications are delivered to the remote users with the required service levels to meet the business needs
• A complete historical record of service level compliance is retained and network managers are alerted whenever targets are violated
• Comprehensive views are provided from a single platform for all UC applications, their usage, and their end-to-end performance
• Quality of service policies are easily configured without any specialized expertise or offline modelling required
• The root cause of any breaches in service level objectives can be quickly determined
• UC service level problems can be located, analyzed and resolved faster and more reliably using CorvilNet's advanced event troubleshooting features

CorvilClear

In the complex maze of market data and order execution, a latency bump in the road could surface anytime, any place. Latency bumps can cost you bundles.

Imagine being able to see those bumps in the road, whenever and wherever they occur, be it within the four walls of your own system, or better yet, outside them. Now you can, with CorvilClear.

The Transparency Revolution

Thanks to CorvilClear, that elephant in the electronic trading room - latency management - can finally be tamed by market centres and their trading partners. At Corvil, we see this as being nothing short of a revolution in the world of electronic trading. So too do people like Yann L’Huillier, Chief Information Officer at Torquoise, the pan-European trading platform.

“A lack of transparency in the execution process can undermine trading strategies. We want to know exactly where and when latency is induced, inside or outside our networks. CorvilClear gives us this ability and Turquoise sees it as an opportunity to develop the first latency transparent market on our [MTF].”

Using peer-to-peer latency sharing technology, CorvilClear exchanges microsecond precision performance data between trading partners who set the level of transparency and sharing policy without exposing any proprietary information.

The industry impact of collaborative troubleshooting with shared sets of data and diagnostic tools has been outlined in detail in our White Paper ‘Inter-Party Latency Management.’ Allow us to paraphrase by saying that CorvilClear means reduced costs, minimized risk, and greater certainty, all adding up to competitive advantage.

Join the transparency revolution - Get CorvilClear.

For the technical nuts and bolts of latency transparency, request our CorvilClear Datasheet.