The Challenge

Intercontinental Exchange, Inc. (ICE) mainly focused on providing trusted and reliable up-to-the-second financial data to traders, risk managers, brokers, and enable investors to help make better trading and risk management decisions, develop trading strategies, and meet compliance obligations. ICE dealt with massive amounts of volatile financial data and needed to put in place tools designed to help ensure that this volatile data is safe, accurate, timely, and easily accessible to its customers.

Keeping these requirements in mind, ICE embarked on the journey to identify a data management solution for two new business-critical applications: 1) derivatives and 2) bonds. The derivatives application processes data for various instrument types, including foreign exchange, futures, commodities options, and equities; the bonds application processes bond-specific data such as price, yield, maturity date, and coupon rates. Both applications give customers access to the information they require to help make certain data-driven decisions.

These applications must maintain a high degree of reliability and peak performance throughout the trading day and process large volumes of data in order to provide the information customers need, such as real-time trends. So ICE’s requirements for these applications around scalability, performance, and security were extremely high. Application uptime was a top consideration to avoid any impact on our customers across the globe.

The Solution

ICE’s stated business objective is to find new technologies to help its customers seamlessly access global markets and information. In line with this objective, ICE needed a data layer that handles the large volume of real-time data coming from new and varying sources for its business-critical derivatives and bonds applications.

ICE looked at different data management solutions and selected DataStax Enterprise (DSE), built on the best distribution of Apache Cassandra™, for its distributed, scalable capabilities to power both types of applications.

“For ICE, we need to provide information to the customer as to what kind of trade happened, when it happened and where it happened, and this needs to occur in real time at massive scale. For such use cases, a distributed, scalable data layer that remains always on is a very good fit,” said Ajit Singh, Director of Data Technology Platforms at ICE.

The Results

ICE currently runs Cassandra in multiple physical data centers, with 25 nodes handling a few terabytes of transactional data per day.

Even with a constantly growing number of users and data, Cassandra, thanks to its linear scalability and multi-data center replication capabilities, has been designed to give ICE the ability to deliver powerful applications that can store massive volumes of data consistently without the concern of constant downtime or performance degradation.

“It helps us easily scale our environments as we grow beyond the capabilities of a traditional cluster of database servers and to scale without being limited by our database software,” said Steve Hirsch, Chief Data Officer at ICE.

With DSE’s real-time analytics and search capabilities, ICE can quickly gain visibility into all the data such as “ticks”, pricing trends, projected pricing, and other attributes pertinent to a specific instrument. This allows customers to query and search for the changes in real time and get instant insights to apply to their trading and investment decisions.

“We compile quotes from almost every market in the world in near real-time and create synthetic products. Our ability to deliver key, reliable, real-time data products enables our customers to continuously calculate risk, accurately price assets, and power their mission critical financial platforms.”

Cassandra-powered applications have enabled ICE to provide its customers with seamless, real-time access to information that is relevant to them, regardless of the data volume and variety, thereby empowering them to make more informed trade decisions.