Concept content delivery network (CDN) system and load balancing mechanisms were considered in this paper. The main task of the CND network — providing the qualitative information delivery to the end user. Content Delivery Network (CDN) — is a geographically distributed network, that contains a number of content servers and routers. As a rule, it consists of a main node (Origin), and caching nodes (Edges) — points of presence, which can be located in various parts of the world. All content are stored and updated on the Origin server. The main advantages of the CDN are the next:
- increasing the content delivery speed;
- reducing the load on the Origin server;
- reducing the load on the primary server.
We propose architecture, for select the optimum route and server for a user request, based on the objective functions. The main criterions of the objective function were taken delay time, server load and probability of packet lost. All user requests are sent to the nearest or lowest-load server to quicken the response.
As part of this paper experimental research video content in CDN network was conducted. The main attention was focused on the primary server to broadcast video content and one of Edge servers. The intensity of the packets flow to Origin server and Edge server were presented. With the increasing load on the Edge server, server resources increase is needed. The load balancing technology can ensure that the user request points to the nearest edge server with minimum load in the network, so that network content is efficiently distributed. Load balancing system performs the following tasks: monitoring of servers and network equipment; choosing the server that will respond to the client’s request; traffic control between client and server.
Dependencies delay time of the intensity receipt of user requests on a Edge server was presented. As we can see from the presented dependencies, delay time and jitter increases with the intensity of the input stream. Load balancing mechanism will allow more efficiently redirect user requests to the Edge servers. This makes it possible to provide services of better quality, ie less latency and jitter, which is critical for real-time services.
1. D. Awduche, A. Chiu. A Elwalid, I. Widjaja Overview and principles of internet traffic engineering. IETF, RFC3272, 2002. 2. Sivasubramanian S., Szymaniak M., Pierre G., “Replication for web hosting systems,” ACM Computing Surveys, vol. 36, no. 3, no. 3, pp. 291–334, 2004. 3. Mykhailo Klymash, Maryan Kyryk, Nazar Pleskanka, Volodymyr Yanyshyn Data Buffering Multilevel Model at a Multiservice Traffic Service Node // Smart Computing Review. Korea — Vol. 4. No. 4. August 31, 2014, p. 294–306. 4. OLNN.CN.CDN, The Content distributed network technology, server, May 28, 2007, http://olnn.cn/html/server/6/20070528/3660.html. 5. G. Dmitriev, B. Margolis, M. Muzanna. Solution of the optimal routing problem on the criterion of network congestion // Software products and systems. — 2013. — No. 4. — P. 17–19. 6. V. Parfenov, S. Zolotarev. An algorithm for solving the optimal routing problem by the criterion of average delay // Vestnik VSU: Ser. Physics. Mathematics. — 2007. — No. 2. — P. 28–32. 7. M. Schwartz. Communication Networks: Protocols, Modeling and Analysis: Trans. from English .: 2 h. — M .: Nauka, 1992. — Part 1 — 336, 2. — Р. 28–32. 8. L. Kleinrock Queueing theory. Trans. with Eng. / Ed. I. Grushko; Ed. V. Neiman. — M .: Engineering, 1979. — Р. 292–320. 9. Pallis George, Konstantinos Stamos, Athena Vakali “and other”. Replication based on Objects Load under a Content Distribution Network // Proceedings of the 22nd (In conjunction with ICDE’06). — Atlanta, 2006. — 53 p. 10. Tim Wauters, Jan Coppens, Bart Dhoedt, Piet Demeester. Load balancing through efficient distributed content placement // In proceeding of: Next Generation Internet Networks. — NY, 2005. — P. 99–105.