A model for ensuring fault-tolerance of container virtual services at data centers

Mathematical support and software for computers, complexes and networks


Аuthors

Rybalko A. A.1*, Naumov A. V.2**

1. EMC Informational Systems CIS LLC, EMC Informational Systems CIS LLC, 3, Begovaya str., building 1, 125284, Russia
2. Moscow Aviation Institute (National Research University), 4, Volokolamskoe shosse, Moscow, А-80, GSP-3, 125993, Russia

*e-mail: aar@mai.ru
**e-mail: naumovav@mail.ru

Abstract

The following work is focused on the topic of enhancing the reliability of distributed services using container virtualization technology. An infrastructure based on container virtualization is presented and analyzed in the environment of changing workloads. Means of deployment, monitoring and lifecycle management are included in the infrastructure. The solution also includes a thorough description of networking exchange between services and service-monitoring system interchange of data in a multi-tenant environment. In modern datacenters the efficiency of resources usage is based on the number of business applications per host unit, while maintaining the isolation of applications from each other, as well as providing the stability, backup and disaster recovery, lifecycle management and network interconnections control.

The article compares the benefits of full and container virtualization, as well as analysis of service-oriented architecture on the base of container virtualization. The instruments analyzed from full virtualization perspective are VMware ESXi, KVM, Microsoft Hyper-V, Citrix XenServer, while OpenVZ and LXC/Docker technologies were reviewed from the side of container virtualization. One of the most important criteria analyzed is the scalability of the resulting solution built using the virtualization instruments presented above. Using container virtual machines, the number of objects on each host greatly grows, which is a more effective usage of resources from one hand, but poses a problem of monitoring and lifecycle management of a much larger number of objects on the other hand. The article analyses the network infrastructure and the monitoring toolset designed to handle much larger numbers of controlled objects than what is currently available on the market.

A hybrid architecture of Token Ring and Partial Mesh domains are presented as the more scalable and effective solution in the article for network communications between container hosts and between containers themselves. This hybrid approach is also applicable in different infrastructures build for high loads and big data processing.

Keywords:

container virtualization, system fault-tolerance, hypervisor, distributed networking

References

  1. Rybalko A.A. Vestnik Moskovskogo Aviatsionnogo Instituta, vol. 16, no. 6, pp. 143-149.

  2. IBM Research Report. An Upd ated Performance Comparison of Virtual Machines and Linux Containers / Wes Felter, Alexandre Ferreira, Ram Rajamony, Juan Rubio. IBM Research Division, Austin Research Laboratory, 11501 Burnet Road, Austin, TX 78758, USA. RC25482 (AUS1407-001) July 21, 2014. Computer Science.

  3. James Lewis, Martin Fowler. Microservices — a definition of this new architectural term, available at: https://martinfowler.com/articles/microservices.html

  4. Rybalko A.A. Materialy XX Mezhdunarodnoi konferentsii po vychislitel’noi mekhanike i sovremennym prikladnym programmnym sistemam (VMSPPS’2017), 25-31 maya 2017, Alushta, pp. 159 – 160.

  5. Alfred V. Aho, Jeffrey D. Ullman, John E. Hopcroft. Data Structures and Algorithms, ISBN-10: 0201000237, ISBN-13: 978-0201000238. Pearson; 1st edition (January 11, 1983), 427 p.

  6. Rybalko A.A. Materialy VIII Mezhdunarodnoi konferentsii po neravnovesnym protsessam v soplakh i struyakh (NPNJ’2010), 24-31 maya 2010, Alushta, pp. 595 – 597.

  7. Feature comparison of different virtualization solutions, available at: https://openvz.org/Comparison

  8. Performance Evaluation of Virtualization Technologies for Server Consolidation / Pradeep Padala, Xiaoyun Zhu, Zhikui Wang, Sharad Singhal, Kang G. Shin. HP Laboratories, HPL-2007-59R1, 2007, available at: http://www.hpl.hp.com/techreports/2007/HPL-2007-59R1.pdf

  9. VMware documentation guide on vSphere 6.5 maximums, available at: https://www.vmware.com/pdf/vsphere6/r65/vsphere-65-configuration-maximums.pdf

  10. OpenzVZ project documentation page on UBC primary parameters, available at: https://openvz.org/UBC_primary_parameters

  11. What Is Open Virtual Network (OVN)? How It Works? – Open Virtual Network project documentation page, available at: https://www.sdxcentral.com/sdn/network-virtualization/definitions/what-is-open-virtual-network-ovn-h...

  12. John Dias. VMware corporate blog. Posted March 23, 2017, available at: https://blogs.vmware.com/management/2017/03/vr-ops-6-5-whats-new-improved-scalability.html

  13. Elvis Pranskevichus, Yury Selivanov. Python documentation. What’s New In Python 3.6, available at: https://docs.python.org/3/whatsnew/3.6.html

  14. Thomas Mangin. ExaBGP project documentation, available at: https://github.com/Exa-Networks/exabgp

  15. Faucet project documentation. «What is Faucet?» available at: https://faucetsdn.github.io/

  16. Zakharov V.N. Sistemy i sredstva informatiki, 2006, vol. 16, no. 3, pp. 279 – 298.

  17. James Smith, Ravi Nair. The Architecture of Virtual Machines, IEEE Computer Society, 2005, 38 (5), pp. 32 – 38.

  18. Gideon Gerzon. Intel® Virtualization Technology, Processor Virtualization Extensions and Intel® Trusted Execution Technology. 2007, available at: https://software.intel.com/sites/default/files/m/0/2/1/b/b/1024-Virtualization.pdf

  19. Baranov A.V., Nikolaev D.S. Programmnye sistemy: teoriya i prilozheniya, 2016, vol. 7, no. 1, pp. 117 – 134.

  20. Virtualizatsiya: tekhnologicheskie podkhody, available at: http://ru.pcmag.com/osnovy/3919/help/virtualizatsiia-tekhnologicheskie-podkhody

  21. Kondrashin M.A., Arsenov O.Yu., Kozlov I.V. Trudy MAI, 2016, no. 89, available at: http://trudymai.ru/eng/published.php?ID=73411

  22. Naumov A.V., Sai Kkhin Aung Tint Trudy MAI, 2011, no. 42, available : http://trudymai.ru/published.php?ID=24321

  23. Naumov A.V., Mkhitaryan G.A., Rybalko A.A. Software se t of intellectual support and security of LMS MAI CLASS.NET. Vestnik YuUrGU. Matematicheskoe modelirovanie i programmirovanie, 2016, vol. 9, no. 4, pp. 129-140.


Download

mai.ru — informational site MAI

Copyright © 2000-2021 by MAI

Вход