Predicting FaaS Runtime with the Orama Framework Using Machine Learning
Resumo
One of the significant challenges in Function-as-a-Service (FaaS) is the unpredictability of function runtimes, which can lead to cost overruns and performance degradation when deploying applications across multiple cloud providers. This paper presents a Machine Learning (ML)-based predictor integrated into the Orama Framework, which combines static code metrics (Halstead complexity measures) with empirical performance data to estimate function runtimes directly from the source code. Three neural network architectures (Dense, LSTM, and BLSTM) were evaluated, with the BLSTM achieving the highest accuracy (R2 = 0.91) and a 20% lower MSE compared to the baselines. The predictor is available via API and through a graphical interface within the Orama Framework, supporting deployment planning, multi-cloud comparisons, and cost-performance optimization for serverless applications.Referências
Akmal, F., Nurcahya, A., Alexandra, A., Yulita, I. N., Kristanto, D., and Dharmawan, I. A. (2024). Application of machine learning for estimating the physical parameters of three-dimensional fractures. Applied Sciences, 14(24):12037.
Almuqati, M. T., Sidi, F., Mohd Rum, S. N., Zolkepli, M., and Ishak, I. (2024). Challenges in supervised and unsupervised learning: A comprehensive overview. International Journal on Advanced Science, Engineering & Information Technology, 14(4).
Barbato, G., Barini, E., Genta, G., and Levi, R. (2011). Features and performance of some outlier detection methods. Journal of Applied Statistics, 38(10):2133–2149.
Barry, B. et al. (1981). Software engineering economics. New York, 197(1981):40.
Bishop, C. M. and Nasrabadi, N. M. (2006). Pattern recognition and machine learning, volume 4. Springer.
Carvalho, L. R. d., Kamienski, B., and Araujo, A. (2024). Main FaaS Providers Behavior Under High Concurrency: An Evaluation with Orama Framework Distributed Architecture. SN Computer Science, 5(5):541.
Chai, T., Draxler, R. R., et al. (2014). Root mean square error (rmse) or mean absolute error (mae). Geoscientific model development discussions, 7(1):1525–1534.
Dai, X., Wei, T.-C., Yoo, S., and Chen, S. Y.-C. (2024). Quantum machine learning architecture search via deep reinforcement learning. In 2024 IEEE International Conference on Quantum Computing and Engineering (QCE), volume 1, pages 1525–1534. IEEE.
Dakkak, A., Li, C., Garcia de Gonzalo, S., Xiong, J., and Hwu, W.-m. (2019). Trims: Transparent and isolated model sharing for low latency deep learning inference in function-as-a-service. In 2019 IEEE 12th International Conference on Cloud Computing (CLOUD), pages 372–382.
de Carvalho, L. R., Kamienski, B., and Araújo, A. P. (2023). Faas benchmarking over orama framework’s distributed architecture. In CLOSER, pages 67–77.
Filippini, F., Cavenaghi, L., Calmi, N., Savi, M., and Ciavotta, M. (2025). Ml-based performance modeling in edge faas systems. In European Conference on Service- Oriented and Cloud Computing, pages 112–127. Springer.
Franzese, M., Iuliano, A., et al. (2018). Correlation analysis. In Encyclopedia of bioinformatics and computational biology: ABC of bioinformatics, volume 1, pages 706–721. Elsevier.
García, S., Luengo, J., Herrera, F., et al. (2015). Data preprocessing in data mining, volume 72. Springer.
Halstead, M. H. (1977). Elements of Software Science (Operating and programming systems series). Elsevier Science Inc.
Horovitz, S., Amos, R., Baruch, O., Cohen, T., Oyar, T., and Deri, A. (2019). Faastest - machine learning based cost and performance faas optimization. In Coppola, M., Carlini, E., D’Agostino, D., Altmann, J., and Bañares, J. Á., editors, Economics of Grids, Clouds, Systems, and Services, pages 171–186, Cham. Springer International Publishing.
Jiang, C., Jiang, C., Chen, D., and Hu, F. (2022). Densely connected neural networks for nonlinear regression. Entropy, 24(7):876.
Khan, B. and Nadeem, A. (2023). Evaluating the effectiveness of decomposed halstead metrics in software fault prediction. PeerJ Computer Science, 9:e1647.
Kurniawan, O., Alkhalifi, Y., Fitriana, L. A., Firdaus, M. R., Rais, A. N., and Hadi, S. W. (2024). Hyperparameter tuning optimization on machine learning models to predict software defects. In 2024 International Conference on Advanced Information Scientific Development (ICAISD), pages 35–40. IEEE.
McCabe, T. (1976). A complexity measure. IEEE Transactions on Software Engineering, SE-2(4):308–320.
Mitchell, T. M. (1997). Does machine learning really work? AI magazine, 18(3):11–11.
Pandey, M. and Kwon, Y.-W. (2024). Funcmem: reducing cold start latency in serverless computing through memory prediction and adaptive task execution. In Proceedings of the 39th ACM/SIGAPP symposium on applied computing, pages 131–138.
Radman, A. and Suandi, S. A. (2021). Bilstm regression model for face sketch synthesis using sequential patterns. Neural Computing and Applications, 33(19):12689–12702.
Schleier-Smith, J., Sreekanti, V., Khandelwal, A., Carreira, J., Yadwadkar, N. J., Popa, R. A., Gonzalez, J. E., Stoica, I., and Patterson, D. A. (2021). What serverless computing is and should become: The next phase of cloud computing. ACM, 64(5):76–84.
Tan, F. (2021). Regression analysis and prediction using lstm model and machine learning methods. In Journal of Physics: Conference Series, volume 1982, page 012013. IOP Publishing.
Tomaras, D., Tsenos, M., and Kalogeraki, V. (2023). Prediction-driven resource provisioning for serverless container runtimes. In 2023 IEEE International Conference on Autonomic Computing and Self-Organizing Systems (ACSOS), pages 1–6.
Almuqati, M. T., Sidi, F., Mohd Rum, S. N., Zolkepli, M., and Ishak, I. (2024). Challenges in supervised and unsupervised learning: A comprehensive overview. International Journal on Advanced Science, Engineering & Information Technology, 14(4).
Barbato, G., Barini, E., Genta, G., and Levi, R. (2011). Features and performance of some outlier detection methods. Journal of Applied Statistics, 38(10):2133–2149.
Barry, B. et al. (1981). Software engineering economics. New York, 197(1981):40.
Bishop, C. M. and Nasrabadi, N. M. (2006). Pattern recognition and machine learning, volume 4. Springer.
Carvalho, L. R. d., Kamienski, B., and Araujo, A. (2024). Main FaaS Providers Behavior Under High Concurrency: An Evaluation with Orama Framework Distributed Architecture. SN Computer Science, 5(5):541.
Chai, T., Draxler, R. R., et al. (2014). Root mean square error (rmse) or mean absolute error (mae). Geoscientific model development discussions, 7(1):1525–1534.
Dai, X., Wei, T.-C., Yoo, S., and Chen, S. Y.-C. (2024). Quantum machine learning architecture search via deep reinforcement learning. In 2024 IEEE International Conference on Quantum Computing and Engineering (QCE), volume 1, pages 1525–1534. IEEE.
Dakkak, A., Li, C., Garcia de Gonzalo, S., Xiong, J., and Hwu, W.-m. (2019). Trims: Transparent and isolated model sharing for low latency deep learning inference in function-as-a-service. In 2019 IEEE 12th International Conference on Cloud Computing (CLOUD), pages 372–382.
de Carvalho, L. R., Kamienski, B., and Araújo, A. P. (2023). Faas benchmarking over orama framework’s distributed architecture. In CLOSER, pages 67–77.
Filippini, F., Cavenaghi, L., Calmi, N., Savi, M., and Ciavotta, M. (2025). Ml-based performance modeling in edge faas systems. In European Conference on Service- Oriented and Cloud Computing, pages 112–127. Springer.
Franzese, M., Iuliano, A., et al. (2018). Correlation analysis. In Encyclopedia of bioinformatics and computational biology: ABC of bioinformatics, volume 1, pages 706–721. Elsevier.
García, S., Luengo, J., Herrera, F., et al. (2015). Data preprocessing in data mining, volume 72. Springer.
Halstead, M. H. (1977). Elements of Software Science (Operating and programming systems series). Elsevier Science Inc.
Horovitz, S., Amos, R., Baruch, O., Cohen, T., Oyar, T., and Deri, A. (2019). Faastest - machine learning based cost and performance faas optimization. In Coppola, M., Carlini, E., D’Agostino, D., Altmann, J., and Bañares, J. Á., editors, Economics of Grids, Clouds, Systems, and Services, pages 171–186, Cham. Springer International Publishing.
Jiang, C., Jiang, C., Chen, D., and Hu, F. (2022). Densely connected neural networks for nonlinear regression. Entropy, 24(7):876.
Khan, B. and Nadeem, A. (2023). Evaluating the effectiveness of decomposed halstead metrics in software fault prediction. PeerJ Computer Science, 9:e1647.
Kurniawan, O., Alkhalifi, Y., Fitriana, L. A., Firdaus, M. R., Rais, A. N., and Hadi, S. W. (2024). Hyperparameter tuning optimization on machine learning models to predict software defects. In 2024 International Conference on Advanced Information Scientific Development (ICAISD), pages 35–40. IEEE.
McCabe, T. (1976). A complexity measure. IEEE Transactions on Software Engineering, SE-2(4):308–320.
Mitchell, T. M. (1997). Does machine learning really work? AI magazine, 18(3):11–11.
Pandey, M. and Kwon, Y.-W. (2024). Funcmem: reducing cold start latency in serverless computing through memory prediction and adaptive task execution. In Proceedings of the 39th ACM/SIGAPP symposium on applied computing, pages 131–138.
Radman, A. and Suandi, S. A. (2021). Bilstm regression model for face sketch synthesis using sequential patterns. Neural Computing and Applications, 33(19):12689–12702.
Schleier-Smith, J., Sreekanti, V., Khandelwal, A., Carreira, J., Yadwadkar, N. J., Popa, R. A., Gonzalez, J. E., Stoica, I., and Patterson, D. A. (2021). What serverless computing is and should become: The next phase of cloud computing. ACM, 64(5):76–84.
Tan, F. (2021). Regression analysis and prediction using lstm model and machine learning methods. In Journal of Physics: Conference Series, volume 1982, page 012013. IOP Publishing.
Tomaras, D., Tsenos, M., and Kalogeraki, V. (2023). Prediction-driven resource provisioning for serverless container runtimes. In 2023 IEEE International Conference on Autonomic Computing and Self-Organizing Systems (ACSOS), pages 1–6.
Publicado
28/10/2025
Como Citar
CARVALHO, Leonardo Rebouças de; ROCHA FILHO, Geraldo Pereira; ARAUJO, Aleteia.
Predicting FaaS Runtime with the Orama Framework Using Machine Learning. In: SIMPÓSIO EM SISTEMAS COMPUTACIONAIS DE ALTO DESEMPENHO (SSCAD), 26. , 2025, Bonito/MS.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2025
.
p. 49-60.
DOI: https://doi.org/10.5753/sscad.2025.15723.
