Bloomberg’s Compute Architecture group is tasked with scaling data storage and compute services across clusters of machines, enabling developers to easily build distributed applications.

Distributed Key-Value Stores

Distributed Key-Value Stores What if your application needs to consume market data as it happens, in true real time? Though it’s possible to look up data quickly, there is still a delay, and we’re aiming to reduce that wait to as close to zero as we can get it. Though Bloomberg has a variety of…


Data Pipelines

Data Pipelines As data gets distributed from disparate publishers to consumers throughout the organization, there is a need to create a pipeline to connect these sources and sinks of data without mandating bespoke pairwise integration. The pipeline must allow for a standardized schema-driven publishing mechanism with listeners able to decode messages off the wire, transform…


Search Infrastructure

Search Infrastructure Bloomberg’s Foundational Search team provides subject matter expertise and defines best practices for application development teams. Its goal is to provide an easy to use, flexible and powerful search-as-a-service to allow application teams to focus on building applications instead of worrying about managing search infrastructure. The team has developed an Apache Solr analytics…


Cluster and Workload Management

Cluster and Workload Management Bloomberg employs an extensive service-oriented architecture with a large footprint of long-running services which work together to offer up functionality to its users. The Compute Architecture Group investigates methods to deploy these services using resource managers like YARN and Mesos alongside deployment technologies such as Apache Slider and Docker. The company…