ANALYSIS AND DESIGN OF DISTRIBUTED SYSTEMS BASED ON CLUSTER TECHNOLOGIES
DOI:
https://doi.org/10.28925/2663-4023.2021.14.186191Keywords:
distributed systems, cluster technologies, CAP theorem, analysis and design of distributed systems, horizontal and vertical scaling.Abstract
This paper reviews some modern solutions that provide analysis and design of distributed systems based on cluster technologies. These solutions are based on the use of the CAP theorem, which states that for a distributed computer system it is impossible to simultaneously satisfy more than two of the three properties: consistency, availability and partition tolerance. In relation to this theory, a formalization of the definition of parameters for information and distributed systems is given.
The use of scalability mechanisms in two directions is considered: vertical scaling (in which the computing power of one server is increased) and horizontal scaling (in which a large number of servers are built within one cluster). Also, the stage of the analysis of distributed systems is described, where the focus is on horizontal scaling.
The paper gives an example of creating a software environment for experimental research in the field of evaluation of distributed systems. The built distributed system is based on a cluster that combines a corporate network with 3 local subnets and up to 40 servers. Each server can have several virtual machines with different operating systems. Examples of program code are given that create pods on which the nginx web server is installed, horizontal scaling of the created pod, and checks the current state of the created objects and triggering events.
This work provides an opportunity to obtain and analyze experimental data at subsequent stages of the study of cluster systems using the load effects on web servers and databases.
Downloads
References
Cooper, B. et al. (2008). PNUTS: Yahoo!’s Hosted Data Serving Platform. VLDB Endowment (VLDB 08), 1277-1288.
Sobel, J. (2008). Scaling Out. Facebook Engineering Notes. www.facebook.com/note.php?note_id=23844338919&id=9445547199.
Fox, A., Gribble, S. D., Chawathe, Y., Brewer, E. A., Gauthier, P. (1997). Cluster-based scalable network services. ACM SIGOPS Operating Systems Review, 31(5), 78–91. https://doi.org/10.1145/269005.266662.
Brewer, E. (2012). CAP twelve years later: How the "rules" have changed. Computer, 45(2), 23–29. https://doi.org/10.1109/mc.2012.37.
Brewer, E. A. (2001). Lessons from giant-scale services. IEEE Internet Computing, 5(4), 46–55. https://doi.org/10.1109/4236.939450.
Abadi, D. (2010). Problems with CAP, and Yahoo’s Little Known NoSQL System. DBMS Musings, blog. http: //dbmsmusings. blogspot.com/2010/04/problems-with-cap-and-yahoos-little.html.
Brewer, E. A. (2000). Towards robust distributed systems. In the nineteenth annual ACM symposium. ACM Press. https://doi.org/10.1145/343477.343502
Mahajan, P.. Alvisi, L., Dahlin, M. (2011). Consistency, Availability, and Convergence, tech. report UTCS TR-11-22. Univ. of Texas at Austin.
Shapiro, M. (2011). Convergent and Commutative Replicated Data Types. Bulletin of the EATCS, 104, 67-88.
Gilbert, S., Lynch, N. (2002). Brewer's conjecture and the feasibility of consistent, available, partition-tolerant web services. ACM SIGACT News, 33(2), 51–59. https://doi.org/10.1145/564585.564601.
DeCandia, G. (2007). Dynamo: Amazon's Highly Available Key-Value Store. In 21st ACM SIGOPS Symp. Operating Systems Principles (SOSP 07) ACM (p. 205-220).
Hale, C. (2010). You Can’t Sacrifice Partition Tolerance. http:// codahale.com/you-cant-sacrifice-partition-tolerance
Du, B., Brewer, E. A. (2008). DTWiki: A Disconnection and Intermittency Tolerant Wiki. In 17th Int’ l Conf. World Wide Web [Online]. ACM (p. 945-952)
