Link Search Menu Expand Document

Autoscaling, Optimization, and Quality of Service

Autoscaling is a defining characteristic of serverless computing, so a great deal of research touches on it in some way. Autoscaling must balance quality of service and cost, and the work we highlight here relates directly to this tradeoff. Even with FaaS, customers are still required to configure some resources, notably the “memory size,” which serves as a proxy for instance execution resources. Sizeless [1] and COSE [2] analyze functions as they run, attempting to find optimal resource configurations. Winzinger and Wirtz [3] also provide a model for FaaS execution. There are multiple approaches to quality of service: Sequoia [4] targets policy goals, whereas Atoll [5] focuses on latency objectivs.

An eclectic mix of work rounds out the early autoscaling-focused efforts. Yussupov et al. [6] study how to reengineer existing applications for scalability, introducing the notion of “serverless parachutes” that are used only under exceptional load conditions. Spock [7] uses both server VMs and serverless functions to meet elasticity and cost goals. Anna [8] provides autoscaling tiered storage, seeking to optimize for both cost and performance goals.

  • [1]Simon Eismann, Long Bui, Johannes Grohmann, Cristina Abad, Nikolas Herbst, and Samuel Kounev. 2021. Sizeless: Predicting the Optimal Size of Serverless Functions. In Proceedings of the 22nd International Middleware Conference, 248–259.
  • [2]Nabeel Akhtar, Ali Raza, Vatche Ishakian, and Ibrahim Matta. 2020. COSE: Configuring Serverless Functions Using Statistical Learning. In IEEE INFOCOM 2020-IEEE Conference on Computer Communications, IEEE, 129–138.
  • [3]Stefan Winzinger and Guido Wirtz. 2019. Model-Based Analysis of Serverless Applications. In 2019 IEEE/ACM 11th International Workshop on Modelling in Software Engineering (MiSE), IEEE, 82–88.
  • [4]Ali Tariq, Austin Pahl, Sharat Nimmagadda, Eric Rozner, and Siddharth Lanka. 2020. Sequoia: Enabling Quality-of-Service in Serverless Computing. In Proceedings of the 11th ACM Symposium on Cloud Computing, 311–327.
  • [5]Arjun Singhvi, Arjun Balasubramanian, Kevin Houck, Mohammed Danish Shaikh, Shivaram Venkataraman, and Aditya Akella. 2021. Atoll: A Scalable Low-Latency Serverless Platform. In Proceedings of the ACM Symposium on Cloud Computing, 138–152.
  • [6]Vladimir Yussupov, Uwe Breitenbücher, Michael Hahn, and Frank Leymann. 2019. Serverless Parachutes: Preparing Chosen Functionalities for Exceptional Workloads. In 2019 IEEE 23rd International Enterprise Distributed Object Computing Conference (EDOC), IEEE, 226–235.
  • [7]Jashwant Raj Gunasekaran, Prashanth Thinakaran, Mahmut Taylan Kandemir, Bhuvan Urgaonkar, George Kesidis, and Chita Das. 2019. Spock: Exploiting Serverless Functions for Slo and Cost Aware Resource Procurement in Public Cloud. In 2019 IEEE 12th International Conference on Cloud Computing (CLOUD), IEEE, 199–208.
  • [8]Chenggang Wu, Vikram Sreekanti, and Joseph M. Hellerstein. 2019. Autoscaling Tiered Cloud Storage in Anna. Proceedings of the VLDB Endowment 12, (2019).