Java specialist Azul Systems and messaging provider C24 Technologies have combined their expertise to deliver an in-memory storage solution compromising C24’s Simple Data Objects (SDO) message compaction technology integrated with Azul’s Zing Java virtual machine (JVM). The solution supports the Financial Products Markup Language (FpML), as well as FIX and ISO 20022 corporate actions messaging formats, and is designed, in the case of FpML to reduce the cost and improve the performance of OTC derivatives data storage. On a broader scale, it also addresses financial firms’ concerns about Java memory size and predictable performance.
Azul and C24 have worked together for some years using C24’s Complex Data Object storage technology on the Zing Java platform, which mitigates JVM issues such as jitter and garbage collection pauses. They are now moving forward with C24’s latest SDO message compaction technology that has a smaller footprint and can significantly reduce the hardware infrastructure cost of OTC derivates data storage. Performance improvement is achieved by giving Zing through simplified, yet fast access to the in-memory data.
Scott Sellers, Azul Systems co-founder and CEO, says: “C24’s SDO technology amplifies the benefits of Zing. By combining the technologies, the fears that many banks and investment houses might have concerning Java memory size and predictable performance are eradicated. C24 addresses big data memory use efficiency and Azul mitigates JVM-induced production issues.”
John Davies, chief technology officer and co-founder of C24, explains: “C24 SDO technology compacts traditional complex XML messages down to a tightly packed binary and wraps them with an ultra-efficient Java applications programming interface, in many cases reducing the memory footprint for stored data by well over 10-times. The data can be searched or queried with very little object creation, but the applications built on this technology run in Java and the only way to make that work with the sort of service level agreements our clients need is with a JVM like Zing.”
Azul and C24 expect both ultra-low latency and low latency traders to consider the solution for uses cases such as real-time big data analytics and real-time reporting. To date, two large banks have tested the solution and are moving it into production.