Оптимізація інформаційної мережі вузу

Abstract

UK: Представлено модель кластерної платформи інформаційної мережі вузу з метою її оптимізації. Дослідження присвячено вирішенню актуального завдання підвищення ефективності та одночасного доступу до сукупності web-сервісів і інформаційних баз даних у процесі навчання. RU: Представлена модель кластерной платформы информационной сети вуза с целью ее оптимизации. Исследование посвящено решению актуальной задачи повышения эффективности и одновременного доступа к совокупности web-сервисов и информационных баз данных в процессе обучения. EN: Offered the model of cluster platform information network of the university to its optimization. Research is sent to the decision of actual task of increase of efficiency and simultaneous access to totality of web- services and information databases in the process of educating. Physically, the cluster is organized as a group of servers, combined logically. They are able to handle the requests coming from users and used as a single resource. The implementation of the cluster is mainly using university LAN. Total processing time of the requests to the server set by its maximum load at simultaneous user requests to the file server from a set of databases. It provides access to resources, handles a lot of traffic, a large volume stores files and archives. Problem solution of remote access to systems of information provision depends of access speed to it. Packages generation were held by connecting to various department servers to switches that generate traffic to the desired address and specified port. Using multiple work processes, on the one hand, reduces the load on each specific workflow, on the other hand, running the aggregate workflow allows for more efficient use of hardware resources working server. Besides running multiple workflows allowed to increase server reliability, by isolating groups of clients, working with different information bases. In a cluster, where possible run multiple workflows, each of them can be automatically restarted after the time specified by the administrator of the cluster.