Font Size: a A A

DDS-based Financial Data Subscription And Distribution System Design

Posted on:2015-11-04Degree:MasterType:Thesis
Country:ChinaCandidate:Q ZhouFull Text:PDF
GTID:2308330464463221Subject:Software engineering
Abstract/Summary:PDF Full Text Request
In the financial transaction of investment bank, real-time data plays important role. Rode Company committed the support of financial transaction system in the background. Lots of applications in subsystems request the latest real-time data to support the action of business. Different systems have different requirements to the data. Message center sends the data to TIBCO EMS topics which supports JMS to get it be consumed by downstream client. In such case, lots of instances as durables created for each topic of each client which using for consuming data separately. If the data could not be consumed immediately, the data will be pending in the durables. If the data keep increasing, heavy load issue will happens on JMS side. If support team prefer to clean up the durable which is full, the data inside will be removed and get missed. If downstream client requests republish these data, all downstream clients will get duplicate data. How to resolve the heavy load problem and make sure the data quality of downstream clients turns to be a quite critical issue.First of all, the issue of data synchronizes between upstream message center and a downstream client is exposed base on the requirement analysis. Enable the middleware system of DDS is necessary to resolve this kind of problem. Next, the functions of data distribution and subscription are discussed, the whole process includes data distribution, data cache and data subscription is analyzed. Then, the architecture of DDS based data subscription and distribution system is analyzed. The items as how to set up data caches database, how to implement the data distribution effective, how to design a visible user interface for monitoring the status are concerned and included into the discussion. The key point of how to design an effective system to handler big data distribution with cache provides multiple solutions to downstream client for subscription is discussed. Finally, the population prospect of this implementation in investment bank will also be analyzed.
Keywords/Search Tags:Financial Transaction, Data Distribution, Real-time, Database, Cache
PDF Full Text Request
Related items