希望在明天

 

Regionalized concept

 Introduce

  we all known that the routine process in the companies are closely depended on the information system process and may be great impact if information system is down. Currently, more and more companies extend their business oversea. Different cultures and different regulations may lead to some changes of the routine process and information systems are requested to change correspondingly. The multi version of information systems in regional diversities is derived from the single system. However, the information system in different regions will be updated separately according to local business request and evolve into a series of non-related systems. It’s very hard for IT to apply to the patches to all areas when an application bug is found and solved in one area but the others already unsolved.  What’s more, any enhancement for systems needs to be separately done many times for the regions. The resources are greatly wasted.

 Regionalized concept is introduced into the system design to avoid the above case occurred. Each area will share the same common source but run their own environment separate from each other. As a result of regionalization, it’s necessary for these areas to share a generic environment so that future core upgrades and individual amendment release could be affected once and across all these systems. Such would facilitate an economy of scale both on the time to effect for the changes as well as the support costs for the separate and secrete implementations.

 Scope

  In system, program libraries could be divided into “Core” and “Local”. Core libraries contain the standard version of the program source and objects. The contests in these libraries are constant for the same version of the multi sites. On the other hand, “Local” libraries contain the additional local amendments / enhancements which are solely applicable to the area itself. As such, the “Core” library is the information that could be shared among the regionalized areas with the same core version.

 Among various libraries, such as database, program source, object, etc, the core program source and object libraries are preferable to be shared as they are transparent to user and less impact to production operation. After regionalization, the database for each HUB site is till separated but the core source, object and data dictionary libraries are shared among the regionalized areas under the same core version. On the other hand, each regionalized area could keep their local features individually in its local source and object libraries.

 

Local(B)	object (TAI)	object(CN)	object(HK)
	source(TAI)	source(CN)	source(HK)
Core(A)	object
	source

  A.     Core libraries contain program source and objects which are consistent for the same core version of all area.  The above definition also allows different groups to run in the same Core version.

 B. In order to fulfil the local regulatory and/or business requirements, different nature of local amendments has been developed in different sites during system implementation.  Such local amendments were developed either on the core programs which are known as ‘local to core’, or developed as new local objects which are called ‘pure local’.  Over the past decade, more and more local amendments have been built and so none of the production is identical even though they are under the same core version.  To have the regionalised all areas sharing a common object environment, one of the exercises, therefore, is to minimise the local amendments.

 Deployment

 This concept can be introduced in Java Platform and AS Platform.

 公司的日常业务越来越越离不开IT，IT作为后勤部门地位越来越高，但这始终都改变不了IT作为非盈利性部门的地位。而且如果公司业务扩展了，并扩展到国外越来越多的国家，那么IT的工作就越来越多，一个地方的业务有所改变，系统要做相应的更改，其他地方不需要该，但如果一个地方原有的业务系统有漏洞，打完这地方的补丁后，其他地方也要相应的改，改与不改，要改动的地方是否一致，一系列的问题出现了。IT人员每天就要忙这找出要打的补丁,有效的时间浪费了,人力资源也大大消耗.冗余性的代码,重复性的工作,IT 投入了大量的精力却收效甚微.

 区域化的设计理念如果从一开始设计就带入,将会减少和避免上述的情况.每个区域共享一套相同的代码,还有一套自我特色的代码. 如果一个区域的共同代码变了,那么其他区域的共同代码也要变.如果一个区域自我特色的代码边了,那么其他区域的自我特色代码不需要变.这样一来,就有利于业务的扩展和支持费用的减低.

 在系统里,有两层程序代码”Core” 和”Local”.  在”Core” 的库里有程序代码和程序,这些对各个区域的代码都是固定的,唯一的.而各个区域的”local”程序是不同的.

 

Local(B)	object (TAI)	object(CN)	object(HK)
	source(TAI)	source(CN)	source(HK)
Core(A)	object
	source

 

A 每个区域都有的程序代码和程序集

B 程序代码

1.在A 的基础上的更新,但系统将会调用B的代码,由于在排列顺序里，B运行先A, 如果B找不到运行的程序,才到A中找. 当然随着年月的增加，B中会存在越来越多改写A的代码，而相应的A中的代码就不再被调用。

2. B 程序代码有可能A中是没有，完全是新加的

 

这种区域代码控制，在Java平台和AS400平台都可以实现。软件内可以设定程序集的运行顺序。

在IT界也摸怕滚打了几年，其中的辛酸苦辣，欢乐，兴奋已化做一段段碎片般的记忆．这些也算是自己积累的经验，为了防止日益生疏而忘却，还是把它总结归档吧：－）

 本人实在太懒.所以就先列个提纲

1．Regional concept

2．Message queue

3．Test directory

4．Trigger

 

从接AOC implementation project,自己就在昏天暗地地忙呼着,这是对自己一个意志和体力的考验,心细,胆大,忍耐,坚持,我知道这是我到HSBC最大的challenge,也是我工作以来的所面对的最大压力(或许是自己给自己的负担),项目总是羞答答的,搞不懂真面目,带着这项目到上海,心里总不塌实,给自己的一个准则就是尽力而为,见机行事.

上海大千世界,住在南京路边,每天固定的从南京西路到南京东路,再从南京东路到南京西路,踏着色彩斑斓的霓虹灯之夜,看着穿梭于人群中的俊男美女,掠过响誉海内外的十里洋场,本应是富丽堂皇,繁华得让人惊叹,在我眼中却是嘈杂,无序,每天掰着指头算着何时才能离开.大上海,大上海,这么多人为她俯首称臣,却让我几欲逃离.我明白这是自己心境的问题,最好的并不是最适合自己的.