Abstract
Autoscaling serverless architectures utilizing Function as a Service (FaaS) is an established model. While there is virtually no limit to scalability in theory, in practice, a trade-off between price and performance determines the cost-efficient scalability of cloud deployments. Finding the correct specifications becomes even harder when the computational demands depend highly on the functions’ inputs. Consequently, a single configuration is often not cost-efficient enough.
To solve this problem, our paper proposes a deployment model for multiple specifications to cover inputs with differing computational demands. By defining categories for the functions’ inputs, requests can be routed to particular deployments to increase the overall cost-performance ratio. Applied filters to the functions’ triggers alleviate the complexity of multiple deployments, and deployments can actively select inputs within their assigned category.
We evaluated our approach with multiple use cases and programming languages on Amazon Web Services (AWS) and Azure. Multiple deployments can generally be justified, if cost is higher for shorter duration. The efficiency of our approach depends on (i) the assignment of correct categories, (ii) the number of requests in each category, and (iii) the configuration granularity of the cloud service provider. While different languages do not influence the effectiveness of this approach, it is hindered by limited configuration possibilities on Azure. Thus, it is easier to find the best cost-performance ratio on AWS.
This work has been supported by the Doctoral College Resilient Embedded Systems, which is run jointly by the TU Wien’s Faculty of Informatics and the UAS Technikum Wien.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Alencar, D., Both, C., Antunes, R., Oliveira, H., Cerqueira, E., Rosário, D.: Dynamic microservice allocation for virtual reality distribution with GoE support. IEEE Trans. Netw. Serv. Manag. 19(1), 729–740 (2022)
vom Brocke, J., Hevner, A., Maedche, A.: Introduction to design science research. In: vom Brocke, J., Hevner, A., Maedche, A. (eds.) Design Science Research. Cases. PI, pp. 1–13. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-46781-4_1
Dashbird: Performance monitoring for aws lambda. Web, March 2023. https://dashbird.io/blog/performance-monitoring-for-aws-lambda/
Eismann, S., Bui, L., Grohmann, J., Abad, C.L., Herbst, N.R., Kounev, S.: Sizeless: predicting the optimal size of serverless functions. In: Proceedings of the 22nd International Middleware Conference (2021)
Eismann, S., Grohmann, J., van Eyk, E., Herbst, N., Kounev, S.: Predicting the costs of serverless workflows. In: Proceedings of the ACM/SPEC International Conference on Performance Engineering, pp. 265–276. ICPE ’20, Association for Computing Machinery, New York, NY, USA (2020)
Elgamal, T., Sandur, A., Nahrstedt, K., Agha, G.: Costless: optimizing cost of serverless computing through function fusion and placement. In: 2018 IEEE/ACM Symposium on Edge Computing (SEC), pp. 300–312 (2018)
Hellerstein, J.M., et al.: Serverless computing: one step forward, two steps back. ArXiv abs/1812.03651 (2019)
Ishakian, V., Muthusamy, V., Slominski, A.: Serving deep learning models in a serverless platform. In: 2018 IEEE International Conference on Cloud Engineering (IC2E), pp. 257–262 (2018)
Kim, Y., Lin, J.: Serverless data analytics with flint. In: 2018 IEEE 11th International Conference on Cloud Computing (CLOUD), pp. 451–455 (2018)
Kusnierz, J., et al.: A serverless engine for high energy physics distributed analysis. In: 2022 22nd IEEE International Symposium on Cluster, Cloud and Internet Computing (CCGrid). IEEE, May 2022
McGrath, G., Brenner, P.R.: Serverless computing: design, implementation, and performance. In: 2017 IEEE 37th International Conference on Distributed Computing Systems Workshops (ICDCSW), pp. 405–410 (2017)
Qu, C., Calheiros, R.N., Buyya, R.: Auto-scaling web applications in clouds: a taxonomy and survey. ACM Comput. Surv. 51(4), 1 (2018)
Savazzi, S., Nicoli, M., Rampa, V.: Federated learning with cooperating devices: a consensus approach for massive IoT networks. IEEE Internet Things J. 7(5), 4641–4654 (2020)
Son, M., et al.: Splice: an automated framework for cost-and performance-aware blending of cloud services. In: 2022 22nd IEEE International Symposium on Cluster, Cloud and Internet Computing (CCGrid), pp. 119–128 (2022)
Spillner, J., Mateos, C., Monge, D.A.: FaaSter, better, cheaper: the prospect of serverless scientific computing and HPC. In: Mocskos, E., Nesmachnow, S. (eds.) CARLA 2017. CCIS, vol. 796, pp. 154–168. Springer, Cham (2018). https://doi.org/10.1007/978-3-319-73353-1_11
Wang, H., Niu, D., Li, B.: Distributed machine learning with a serverless architecture. In: IEEE INFOCOM 2019 - IEEE Conference on Computer Communications, pp. 1288–1296 (2019)
Wang, J., Pan, J., Esposito, F., Calyam, P., Yang, Z., Mohapatra, P.: Edge cloud offloading algorithms: issues, methods, and perspectives. ACM Comput. Surv. 52(1), 17–18 (2019)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 IFIP International Federation for Information Processing
About this paper
Cite this paper
Patsch, R., Göschka, K.M. (2023). Optimizing the Cost-Performance Ratio of FaaS Deployments. In: Papadopoulos, G.A., Rademacher, F., Soldani, J. (eds) Service-Oriented and Cloud Computing. ESOCC 2023. Lecture Notes in Computer Science, vol 14183. Springer, Cham. https://doi.org/10.1007/978-3-031-46235-1_18
Download citation
DOI: https://doi.org/10.1007/978-3-031-46235-1_18
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-46234-4
Online ISBN: 978-3-031-46235-1
eBook Packages: Computer ScienceComputer Science (R0)
