适用于交易前合规的开源框架/技术 [英] Open source frameworks/technologies suitable for PreTrade Compliance

问题描述

我们期待在开源框架/技术上建立交易前合规性，以便最终将旧代码替换为某些有用的框架.我们已经有基于遗留代码的OMS.

我需要帮助来找出合适的零件/框架，以便我们将精力投入到正确的工具集上.请参考建议的架构图...

选择新技术/框架的目的



1)必须是开源的，因为不想投资付费框架.

2)可扩展

3)容错

4)负载平衡功能

5)如果可能的话，具有网格计算能力

6)快速计算

7)具有时态逻辑功能会很棒.



问题/疑虑/疑问



最担心的是选择计算引擎



1)我已经看到了很多关于计算引擎的讨论.有人说cep很好，而另一些人则倾向于使用RETE规则引擎来实现交易前合规.

2)Drools提供了规则引擎以及cep产品，因此drools cep似乎能够利用某些优点，例如在网页/决策表中定义规则.因此，业务用户可以定义规则，而不是由技术人员在ESper中编写EQL/模式，这很棒.

3)我知道可以使用Esper中的模式来应用规则.

4)我在Esper中没有看到一些bug.在生产级别使用免费的Esper是否安全?我知道也有付费选择.

5)据说Esper比流口水快.这是真的吗?

6)是否可以将其他任何引擎(例如mongodb，storm，Hadoop和map/reduce等)一起使用?

7)RabbitMQ是处理大量数据流的正确选择吗?



非常感谢您完成这么长的篇幅.我将不得不了解您的想法，以改善整体设计.如果您可以建议任何参考体系结构，那就太好了.如果某些问题与某个特定框架没有直接关系，但可以帮助您定义总体流程，请原谅.



期待您的帮助...

Hi,

We are looking forward to create Pre-Trade compliance on open source frameworks/technologies so that legacy code can eventually be replaced by some of the useful frameworks. We already have our OMS based on legacy code.

I need help to find out the right pieces/frameworks so that we apply our effort on right set of tools. Please refer the proposed architecture diagram…

Objective of choosing new technologies/frameworks



1) Has to be open source, as don’t want to invest in paid frameworks.

2) Scalable

3) Fault Tolerant

4) Load balancing capabilities

5) If possible Grid computing abilities

6) Lightening fast calculations

7) Would be great to have temporal logic capabilities.



Questions/Concerns/Doubts



The biggest concern is in choosing Calculation Engine



1) I have seen lots of discussions on calculation engines. Some say cep is good while others prefer to use RETE rule engines for pre-trade compliance.

2) Drools provide rule engine as well as cep product so it seems that drools cep would be able to utilize some of the goodies like defining rules in webpage/decision tables. So a business user can define the rule rather than technical person writing EQL/Patterns in ESper, which is wonderful.

3) I understand that rules could be applied using the patterns in Esper.

4) I have seen few bugs in Esper. Is it safe to use free Esper till production level? I know that there is paid option as well.

5) Esper is said to be faster than drools. Is this true?

6) Is it viable to use any other engine like mongodb, storm, Hadoop with map/reduce etc.

7) Is RabbitMQ, the right choice for so much of data flow?



I deeply thank you to go through such a long post. I would be obliged to know your thoughts to improve the overall design. It would be great if you can suggest any reference architecture to look at. Please excuse if some of the questions are not directly associated with a particular framework but somewhere it helps to define the overall flow.



Looking forward to your help…

推荐答案

这个问题超出了此类论坛的讨论范围，但是请关注LMAX干扰模式和实现.

This question is out of the scope of a forum like this, however look into the LMAX disruptor pattern and implementation.

这篇关于适用于交易前合规的开源框架/技术的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！