An On-Going Challenge
As market data rates continue to skyrocket, legacy data delivery solutions have struggled to keep pace. During times of high data volumes, systems have bogged down and the delivery of data updates have been delayed. This increase in market data latency creates a disadvantage for trading applications and has prompted consumers of market data to turn towards direct feed solutions where communication lines are run from the client’s site to exchanges and ECNs, and customers manage their own ticker plants within the enterprise.
While this approach has succeeded in insulating consumers from latency introduced by market data vendors, customers now face the same problems that have dogged the market data vendors, how to deliver increasingly larger amounts of information without adding latency to the data.
Ticker Plant Capacity Management
Proper system sizing is a vital ingredient for providing timely access to market data. Direct feed implementations can increase the number of inbound data sources ten- fold, and each of these sources brings its own variability and sudden “spikes” to the capacity management equation. The surge in data rates ripple throughout the enterprise taxing feed handlers, networks, distribution servers, and client desktop applications.
Managing an in house ticker plant introduces a new set of challenges for the market data managers and their support staff. Understanding how and when to increase system capacity has evolved into a discreet discipline requiring constant vigilance and quantitative measurement.
WHSS Core Competency – Systems Engineering and Capacity Management
Since its inception in 1998, West Highland Support Services have been wholly dedicated to managing clients’ market data distribution platforms. Countless capacity related infrastructure upgrades have been executed under their watch and without exception; these upgrades have gone off without a hitch.
With the launch of its professional services practice, West Highland has augmented the depth of its offerings adding staff previously responsible for managing full depth direct feed products that provided low latency market data to electronic trading applications. Within the past three years, market data rates have increased by more than 400%, requiring that capacity management be treated as an on-going discipline as opposed to a periodic statistical review.
West Highland’s Capacity Management Services
West Highland professional services help customers address their capacity concerns by drawing upon their wealth of experience in dealing with the explosive growth in market data rates. The West Highland team works with the customers’ systems engineering staff to ensure the following:
|– Capacity of the current delivery platform is understood|
|– Breaking points for each infrastructure component are known|
|– Capacity requirements for end user applications have been quantified|
|– The impact of “slow consumer” end user applications is known|
|– Environments and tools exist to stress test delivery infrastructure and applications in order to determine their breaking point|
|– Scale out models are established and verified for each component|
|– Processing load is optimally distributed on the network|
|– Component monitoring has been implemented and alert levels have been established|
|– Latency attributable to each system component is known and measured|
|– Support group has access to capacity metrics|
|– Information is reviewed on a scheduled basis|
|– Methodology for predicting growth in market data rates has been adopted|
Working with West Highland’s professional services group will help customers establish the rigorous discipline needed to avoid capacity related problems.