High-speed network packet capture
TipOff delivers high-speed, passive, network packet capture to disk with sustained data rates in excess of 40Gbps. By capturing everything, TipOff eradicates the risks associated with sampling approaches, guaranteeing that 100% of raw data received will be available for future analysis.
Utilising hardware accelerated compression technology, TipOff can store more than 20 TB of raw capture files on a 2U appliance or 80 TB of raw capture files on a 4U appliance. This approach facilitates high capacity storage with a low footprint, presenting the potential for significant cost savings.
TipOff’s high degree of compatibility, and ease of integration, makes it the perfect partner to Application Tap®, which offers a patented, high precision approach to in-application instrumentation, with minimal performance overhead.
Combining Application Tap’s ability to gain a fine-grained view of the individual components operating inside an application with TipOff’s extensive level of network visibility, blind spots are eliminated. So regardless of where an issue occurs, TipOff can pinpoint its exact source, considerably reducing the problem’s time to resolution.
High precision latency measurement
Conducting accurate time stamping to a resolution of 10 nanoseconds on a TipOff Appliance or reading the timestamps generated by the capture layer (for example those created by aggregation switches) extremely precise hop-by-hop performance measurements and analytics can be achieved.
These statistics enable TipOff users to very quickly comprehend current performance and identify latency hotspots for focused improvement.
This level of understanding enables users to set benchmarks for future performance and be alerted when thresholds are breached. This insight enables service providers to understand if the latency reported by a client is occurring within their environment or that of the client’s. It also enables exchanges to identify if the matching engine is running slower than expected or if public updates are being distributed more quickly than private.
Additionally, by identifying latency hotspots, users can accurately determine how latency reduction budgets should be apportioned to achieve the greatest return on investment (ROI).
TipOff’s precise instrumentation enables performance issues to be forensically analysed, right down to the individual network packet and nanosecond level, supporting:
- 1/10/40 Gigabyte Ethernet interfaces
- Microburst analysis at OSI layers 1 to 4
- TCP re-transmission and transport quality analytics
- TCP latency and packet layer latency analytics
All of these measures are achieved with exact timings to 10 nanosecond resolution, alongside detailed bandwidth analysis.
TipOff supports performance root cause analysis by leveraging its thorough examination of the middleware stacks that comprise each hop in a data flow. As the most significant source of latency is often the retransmission latency that results from lost packets, TipOff enables middleware-messaging activities to be independently monitored and analysed.
TipOff monitors the internal operation of reliability layers across an extensive range of middleware messaging protocols. TipOff provides protocol specific statistics, identifying retransmission requests, and generating subscriber, publisher and topic statistics for the messages themselves.
In doing so, early warning signs of operational conditions that could result in increased latency can be detected, exposing potential load balancing and capacity challenges.
Market data analysis
TipOff supports high volume, market data sequence gap detection. By instantly identifying and alerting the business to missing sequences, and recognising the feeds impacted, damage limitation methods can be implemented to avoid inaccurate trading decisions based on unknowingly partial data streams.
TipOff also enables the real-time and historical analysis of market data message volumes, which can then be used to facilitate more effective capacity planning.
Event syndication into client specific analysis systems
TipOff guarantees 100% capture and enables data to be decoded through its extensive library of high performance protocol decoders. In addition to offering an extensive range of performance analytics of its own, TipOff is also able to stream normalised data into a client’s own choice of either in-house or third party analysis tools. In doing so, the syndicated events are then able to feed client specific systems, databases and messaging tools providing vital risk and regulatory insight.
In response to client demand, TipOff has developed a range of event syndication adapters facilitating seamless integration with many of the most widely adopted third-party analysis tools.
TipOff offers equities, FX, fixed income and derivatives market data and transaction decoders for an extensive range of regional and globally adopted protocols. Decoders are currently offered for all major North American, European and Asia Pacific markets, and coverage is continuously being expanded in response to client requests.
Automatic Protocol Discovery
TipOff offers a high performance automatic protocol discovery feature, enabling users to automatically detect which market data streams they are consolidating, distributing or consuming, and to identify gaps in market data sequence numbers.
By classifying and decoding the market data protocols against TipOff’s extensive and ever expanding library of decoders, market data sequence gaps can be accurately detected and detailed message rate statistics generated. All of which can be achieved at sustained speeds of over 1.2 million packets per second, per feed.
With this functionality users can quickly get up and running and benefit from being able to:
- Detect sequence gaps in the market data that is being used to formulate trading decisions, indicating that these may be being based on incomplete information
- Gain a comprehensive understanding of which market data feeds their users are subscribing to
- Conduct point-to-point market data latency analysis
- Perform trend analysis and capacity planning by understanding the bandwidth of each feed, ticks and packets per second, and independently identify retransmission requests indicating that certain applications may be struggling to keep up with data volumes.
This feature significantly reduces the management overhead and administration required to manually identify which market data streams are present within an environment and maintain this information on an on-going basis. All of which means that users can set-up TipOff to monitor market data utilisation and detect sequence gaps extremely quickly.
To learn more about this feature, read Steve Rodger’s (Velocimetrics’ Head of Engineering’s) blog post on the addition of automatic protocol discovery to TipOff
Completely configurable on-demand and scheduled reporting
TipOff users can automatically generate intra-day, ad-hoc and scheduled reports.
Offering the flexibility to create highly configurable reports, with TipOff users can script detailed reports that specifically meet their own bespoke requirements and query the captured data using an API.
Reports can be generated for any of the aggregated data TipOff collects and analyses, providing summaries detailing latency measurements, network traffic, middleware performance or market data statistics. These reports can be used to, for example, break-down latency performance by client gateway or individual client, for time-sliced segments of the trading day. Reports can then be passed onto clients as required, shared and consumed by other business groups and archived on an on-going basis.
Available in PDF, HTML and Excel-friendly CSV formats, reports can be automatically distributed via email and RSS.
TipOff can be rapidly deployed as both an Appliance and software only solution on physical and virtual machines.
In August 2015, Velocimetrics launched a new range of TipOff Appliances. The range is deployed on 1U and 2U form factors with a variety of high performance and high capacity options that enable Velocimetrics to provide clients with the optimum solution to most effectively meet their requirements.
The raw storage capacity of the appliances ranges from 4TB(terabytes) to 24TB and provides up to 80TB (1U) or 160TB (2U) of compressed, captured trading data in a single device.
The appliances can be configured with a flexible range of capture ports at 1Gbit, 10Gbit and 40Gbit and up to 6 capture cards providing up to 20 capture ports. This allows for flexible deployment models where the appliance can integrate with aggregation switches / network packet brokers or can accept many tap or mirror port connections directly.
The top of the range V205 appliance offers extremely high capacity disks, double of RAM of its predecessor and faster CPU speeds. It also enables users to retain up to 24TB of raw captured trading data or 160TB of compressed captured trading data on a single 2U appliance.
To learn more about the deployment options TipOff offers please contact us