logo

SCIENTIA SINICA Informationis, Volume 50 , Issue 11 : 1645(2020) https://doi.org/10.1360/SSI-2020-0126

Flexible configuration update delivery for large clusters

More info
  • ReceivedMay 7, 2020
  • AcceptedOct 13, 2020
  • PublishedNov 9, 2020

Abstract

Efficient configuration management plays an important role in managing large-scale clusters in public cloud services. Because a cluster may contain millions of containers, it is challenging to guarantee that configuration updates will be delivered to interested containers reliably and in time to meet the requirements of different scenarios. Existing solutions to this problem have limitations: consensus algorithms limit the scalability of the cluster and may not work for large clusters; epidemic algorithms face the challenge of long-tail latency, which means the overall response time is too long to deliver critical configuration updates effectively. To overcome the limitations of existing solutions, in this paper, we present a novel flexible approach for delivering configuration updates for large-scale clusters. This pub/sub approach uses a configurable, complete $N$-ary tree as the overlay and introduces flexible, two-phase configuration update delivery. This method of update delivery uses a portion of subscribers' resources to improve its performance. Furthermore, it is fault-tolerant when it encounters node failures and network partitions. The strategies and the parameters of the overlay can be changed to meet performance and reliability requirements for different scenarios. Evaluations show that our approach significantly reduces the latency of update delivery compared to existing solutions. It also performs well in cases of node failures and network partitions.


Funded by

国家重点研发计划(2017YFB1001804)

阿里巴巴创新研究计划(Alibaba Innovative Research,AIR)


References

[1] Lamport L. The part-time parliament. ACM Trans Comput Syst, 1998, 16: 133-169 CrossRef Google Scholar

[2] Ongaro D, Ousterhout J. In search of an understandable consensus algorithm. In: Proceedings of USENIX Annual Technical Conference (USENIX ATC '14), Philadelphia, 2014. 305--319. Google Scholar

[3] Demers A, Greene D, Hauser C, et al. Epidemic algorithms for replicated database maintenance. In: Proceedings of the 6th Annual ACM Symposium on Principles of Distributed Computing (PODC '87), New York, 1987. Google Scholar

[4] Birman K P, Hayden M, Ozkasap O. Bimodal multicast. ACM Trans Comput Syst, 1999, 17: 41-88 CrossRef Google Scholar

[5] Sharma Y, Ajoux P, Ang P, et al. Wormhole: reliable pub-sub to support geo-replicated internet services. In: Proceedings of the 12th USENIX Conference on Networked Systems Design and Implementation (NSDI '15), Berkeley, 2015. 351--366. Google Scholar

[6] Ji S P, Ye C Y, Wei J, et al. MERC: match at edge and route intra-cluster for content-based publish/subscribe systems. In: Proceedings of the 16th Annual Middleware Conference (Middleware '15), New York, 2015. 13--24. Google Scholar

[7] Alibaba. Nacos: an easy-to-use dynamic service discovery, configuration and service management platform for building cloud native applications. 2019. https://nacos.io/en-us/. Google Scholar

[8] Kulkarni S S, Demirbas M, Madappa D, et al. Logical physical clocks. In: Proceedings of International Conference on Principles of Distributed Systems (OPODIS '14), Cham, 2014. 17--32. Google Scholar

[9] Demirbas M, Leone M, Avva B, et al. Logical physical clocks and consistent snapshots in globally distributed databases. https://cse.buffalo.edu/tech-reports/2014-04.pdf. Google Scholar

[10] Lin M J, Marzullo K. Directional gossip: gossip in a wide area network. In: Proceedings of the 3rd European Dependable Computing Conference on Dependable Computing (EDCC-3), Berlin, 1999. 364--379. Google Scholar

[11] Eugster P T, Guerraoui R, Handurukande S B, et al. Lightweight probabilistic broadcast. In: Proceedings of International Conference on Dependable Systems and Networks (DSN '01), Washington DC, 2001. 443--452. Google Scholar

[12] Eugster P T, Guerraoui R, Handurukande S B. Lightweight probabilistic broadcast. ACM Trans Comput Syst, 2003, 21: 341-374 CrossRef Google Scholar

[13] Demers A, Greene D, Houser C, et al. Epidemic algorithms for replicated database maintenance. ACM SIGOPS Operating Syst Rev, 1988, 22: 8--32. Google Scholar

[14] Apache. Apache Cassandra. 2019. https://cassandra.apache.org/. Google Scholar

[15] HashiCorp. Consul by HashiCorp. 2019. https://www.consul.io/. Google Scholar

[16] Facebook. Location-aware distribution: configuring servers at scale, 2019. https://engineering.fb.com/data-infrastructure/location-aware-distribution-configuring-servers-at-scale. Google Scholar

[17] Facebook. Systems at scale 2018: scaling data distribution at Facebook using LAD. 2019. https://atscaleconference.com/videos/systems-scale-2018-scaling-data-distribution-at-facebook-using-lad/. Google Scholar

[18] Costa P, Migliavacca M, Picco G P, et al. Introducing reliability in content-based publish-subscribe through epidemic algorithms. In: Proceedings of the 2nd International Workshop on Distributed Event-based Systems (DEBS '03), New York, 2003. Google Scholar

[19] Riabov A, Liu Z, Wolf J L, et al. Clustering algorithms for content-based publication-subscription systems. In: Proceedings of the 22nd International Conference on Distributed Computing Systems (ICDCS '02), Vienna, 2002. 133--142. Google Scholar

[20] Opyrchal L, Astley M, Auerbach J, et al. Exploiting IP multicast in content-based publish-subscribe systems. In: Proceedings of IFIP/ACM International Conference on Distributed Systems Platforms (Middleware '00), Heidelberg, 2000. 185--207. Google Scholar

[21] Li G L, Hou S, Jacobsen H-A. A unified approach to routing, covering and merging in publish/subscribe systems based on modified binary decision diagrams. In: Proceedings of the 25th IEEE International Conference on Distributed Computing Systems (ICDCS '05), Washington DC, 2005. 447--457. Google Scholar

[22] Li G L, Muthusamy V, Jacobsen H-A. Adaptive content-based routing in general overlay topologies. In: Proceedings of the 9th ACM/IFIP/USENIX International Conference on Middleware (Middleware '08), Berlin, 2008. Google Scholar

[23] Mahambre S P, Kumar S.D. M, Bellur U. A Taxonomy of QoS-Aware, Adaptive Event-Dissemination Middleware. IEEE Internet Comput, 2007, 11: 35-44 CrossRef Google Scholar

[24] D'Angelo G, Ferretti S, Marzolla M. Adaptive event dissemination for peer-to-peer multi-player online games. In: Proceedings of the 4th International ICST Conference on Simulation Tools and Techniques (SIMUTools '11), Brussels, 2011. 312--319. Google Scholar