Bloomberg uses Apache Cassandra to support the exponential growth of its fixed income index production platforms. The platforms handle roughly 200 million data points over around 200,000 bonds covering all markets globally, with an extremely low tolerance for errors and the need to adhere to tight regional SLAs.
To ensure that millions of data points are computed correctly and published on time as fast as possible for markets across the globe, Bloomberg built an Index Construction Platform with an Apache Solr-backed search platform to handle thousands of searches per minute, an Apache Cassandra back-end database to store millions of data points per day, and a distributed computational engine to handle millions of computations daily.
The suite of software allows their index production team to monitor the daily index production processes, react and execute any necessary corrective actions, and validate the index calculations to make sure any data they publish are of top quality.