Abstract
Serverless workflows consist of multiple chained functions that pass intermediate data through function invocations. Existing platforms implement data passing via remote storage services that incur significant performance and maintenance overhead. Some recent works optimize data passing with local shared memory but require specialized function scheduling support and are available only in constrained settings. In this paper, we observed that a simplified, peer-to-peer form of distributed shared memory (DSM) is sufficient and efficient to pass data in serverless workflows. With the observation, we propose FaaSPipe, a serverless workflow runtime built on the simplified DSM. FaaSPipe provides PipeFunc, a user-friendly shared memory-based serverless workflow programming model. To support PipeFunc and take full advantage of the simplified DSM, FaaSPipe designs the intra-workflow memory sharing scheme for address space coordination and builds full-duplex memory transfer channels to enable fast, non-blocking peer-to-peer data passing. Evaluation results on real-world workflow applications show that FaaSPipe reduces workflow latency by up to 61% and consumes up to 2.07\(\times \) less network traffic compared to state-of-art serverless platforms.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
AWS Step Functions. https://aws.amazon.com/step-functions/?step-functions
Google Cloud Composer. https://cloud.google.com/composer
Huawei Cloud Function Graph. https://www.huaweicloud.com/intl/en-us/product/functiongraph.html
LightGBM’s documentation. https://lightgbm.readthedocs.io/en/latest/index.html
Akkus, I.E., et al.: SAND: towards high-performance serverless computing. In: 2018 USENIX Annual Technical Conference, Boston, MA, USA, 11–13 July 2018, pp. 923–935
Baldi, P., Sadowski, P., Whiteson, D.: Searching for exotic particles in high-energy physics with deep learning. Nat. Commun. 5, 4308 (2014)
Cadden, J., Unger, T., Awad, Y., Dong, H., Krieger, O., Appavoo, J.: SEUSS: skip redundant paths to make serverless fast. In: EuroSys 2020: Fifteenth EuroSys Conference 2020, Heraklion, Greece, 27–30 April 2020, pp. 32:1–32:15 (2020)
Du, D., et al.: Catalyzer: sub-millisecond startup for serverless computing with initialization-less booting. In: ASPLOS 2020: Architectural Support for Programming Languages and Operating Systems, Lausanne, Switzerland, 16–20 March 2020, pp. 467–481 (2020)
Jia, Z., Witchel, E.: Nightcore: efficient and scalable serverless computing for latency-sensitive, interactive microservices. In: ASPLOS 2021: 26th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Virtual Event, USA, 19–23 April 2021, pp. 152–166 (2021)
Khandelwal, A., Tang, Y., Agarwal, R., Akella, A., Stoica, I.: Jiffy: elastic far-memory for stateful serverless analytics. In: EuroSys 2022: Seventeenth European Conference on Computer Systems, Rennes, France, 5–8 April 2022, pp. 697–713 (2022)
Klimovic, A., Wang, Y., Stuedi, P., Trivedi, A., Pfefferle, J., Kozyrakis, C.: Pocket: elastic ephemeral storage for serverless analytics. In: 13th USENIX Symposium on Operating Systems Design and Implementation, OSDI 2018, Carlsbad, CA, USA, 8–10 October 2018, pp. 427–444 (2018)
Oakes, E., et al.: SOCK: rapid task provisioning with serverless-optimized containers. In: 2018 USENIX Annual Technical Conference, Boston, USA, 11–13 July 2018 (2018)
Shillaker, S., Pietzuch, P.R.: FaaSM: lightweight isolation for efficient stateful serverless computing. In: 2020 USENIX Annual Technical Conference, USENIX ATC 2020, 15–17 July 2020, pp. 419–433 (2020)
Sreekanti, V., et al.: Cloudburst: stateful functions-as-a-service. Proc. VLDB Endow. 13(11), 2438–2452 (2020)
Ustiugov, D., Petrov, P., Kogias, M., Bugnion, E., Grot, B.: Benchmarking, analysis, and optimization of serverless function snapshots. In: ASPLOS 2021: 26th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Virtual Event, USA, 19–23 April 2021, pp. 559–572 (2021)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 IFIP International Federation for Information Processing
About this paper
Cite this paper
Tong, R. (2022). FaaSPipe: Fast Serverless Workflows on Distributed Shared Memory. In: Liu, S., Wei, X. (eds) Network and Parallel Computing. NPC 2022. Lecture Notes in Computer Science, vol 13615. Springer, Cham. https://doi.org/10.1007/978-3-031-21395-3_8
Download citation
DOI: https://doi.org/10.1007/978-3-031-21395-3_8
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-21394-6
Online ISBN: 978-3-031-21395-3
eBook Packages: Computer ScienceComputer Science (R0)