The Design And Implementation Of Task Tempering For Identification System Of Mug Shot Targeting

With the development of network, the stress of data flux and computing, which are burdened by the centre parts of the network, become larger and larger. Even a best computer can hardly satisfy the needs of data exchange. The appearance of Load Balance ethnic made it possible to distribute the portfolio to multi network equipment logically, which had the same ability. Thus, to prevent the state of that one equipment is too busy while the others are almost free. With the development of the technique, aimed* at the cast and App appeared chain aggregation, high lay exchange and severs with balance plan resolving and software,hardware, and mixture forms.The problem this paper discussed is the task tempering of Identification System of Mug Shot Targeting (ISMUST). Solve the problem of logically distributing the task through the arithmetic and projects explained latter. Thus to use the resource of compare sever in the system as much as possible and at the same time, made the whole system to remain the stability of running and timely reply. The task referred by the users of ISMUST can divided to comparing task, modeling task, adding new dates task and identity validating task. Different tasks and the same task using different dates consume great different sever resource. If using conventional cycle attempter arithmetic, there would be a great waste of resource. Itmay bring many kickbacks such as dead. Aimed at this characteristic, this paper designed a dynamic arithmetic set off from the task's wastage of resource and the manage ability of sever. The main idea is to analyze the resource of comparing sever of the task may use through the experiential dates, and then distribute this task to the most free comparing sever to finish by inspect the currently resource estate of e severs.In order to prevent single point and key point disable, we designed the ability of exchange thief sever and secondary sever etc, and enhance the stability of ISMUST. And we designed a special pretreatment mechanism to avoid ineffective task use too much resource.