Regardless of the product choice that best meets your business needs, the foundation of every SpryWare solution is the core SpryWare technology. SpryWare offers a number of feed handler products with a wide array of installation options, from real-time processing of raw streaming data to indexed historical data.
Deployment ranges from fully hosted solutions on dedicated servers in our own data centers to the fully redundant SpryWare Proximity Cloud. We even offer multiple data distribution models, such as compiling and linking your client application with our embedded, In-Process feed handlers, or fan-out options like the SpryWare Gateway and the SpryWare Proxy server.
The core technology is a purpose built set of software processes written specifically for the market data industry. While many companies have a product line put together of fragmented solutions from many different origins, SpryWare’s is built based on a single set of source code, allowing us to take full advantage of the core feed handler technology across our products. Benefits include knowledge sharing across product development, more efficient updates and faster implementation of new technologies within the industry.
The core technology is a fresh alternative to existing legacy systems. Built from the ground up, it is designed to take full advantage of multi-core and parallel processing. By focusing on multi-threaded development we are able to achieve higher throughput and make full use of the processing capacity of modern server class machines and high-end network adapters. The multi-threaded development also results in linear scalability with an increase of processor cores, which is key with data messaging rates rising.
The technology is platform and hardware independent, and runs on all modern Linux and Microsoft operating systems. The core technology is extremely flexible and designed to take full advantage of the latest server platforms and equipment available.
A common API links your application into the same core technology used to build our high speed feed handlers. The familiar and easy to use API is, in essence the same no matter what kind of product and distribution model you are looking at.
The SpryWare DMS processes 350,000 messages per second per thread for a total throughput of 10 Million+ MPS in a typical server setup. The core technology is written in ANSII C++ for cross platform portability. The design is modular, and all code is tuned for low latency and high scalability. Parallelism is implemented along with lock-free database and queue mechanisms to take advantage of multi-core architectures.
At the heart of the technology are the feed handlers and database services. Access to real-time data streams and database tables are controlled by the SpryWare publisher service and are presented to client applications through the use of the SpryWare unified API. In its basic design, the core technology runs the feed handlers, database, and publisher services at the server level. Data is then processed and simultaneously delivered to the appropriate database and to the client via the publisher service.