LSTM-based Forecasting for Non-linear Workloads in On-premises Software Systems
Resumo
Cloud computing has become a commonly adopted solution to increase business operations efficiency and scalability. However, costs related to cloud infrastructure are increasingly prohibitive for some organizations in long-term deployment scenarios. This expense barrier has led to the emergence of a cloud repatriation trend, where existing software systems are moved back to on-premises environments. Although this initiative enhances cost control management and autonomy, it reintroduces technical challenges. This work investigates, through an empirical study, the applicability of Long Short-Term Memory (LSTM) networks for workload prediction to support auto-scaling decisions in on-premises environments. Initially, we contrasted the prediction accuracy of LSTM with classical methods, utilizing both standard benchmark datasets and access logs collected from an industrial-grade software system. In the second phase, we examined how varying temporal resolutions affect precision and computational expense. Results show the LSTM consistently outperforms classical methods, reducing MAPE by up to 18.02% and RMSE by 8.77% on benchmark datasets. Considering the in-field dataset, it outperforms in long-term predictions, especially regarding MAE and MAPE. Throughout all analyzed temporal resolutions, the 10-minute resolution optimally balanced accuracy with efficiency. Moreover,we noted that LSTM requires less training time than classical methods, emphasizing its applicability for real-world scenarios. Therefore, the results indicate that LSTM emerges as a viable solution to enable proactive auto-scaling within on-premises environments, potentially reducing costs in the context of cloud repatriation.
Palavras-chave:
Auto-scaling, Time Series Forecasting, On-premises Software Systems, Cloud Repatriation
Referências
CP Amajuoyi, LK Nwobodo, and MD Adegbola. 2024. Transforming business scalability and operational flexibility with advanced cloud computing technologies. Computer Science & IT Research Journal 5, 6 (2024), 1469–1487.
anon. 2025. LSTM-based Forecasting for Non-linear Workloads in On-premises Software Systems. DOI: 10.5281/zenodo.15285346 Computational notebook.
Kuciuk Artiom. 2025. Cloud Migration Framework: Transitioning from On-Premises to Azure Cloud for Improved System Reliability and Scalability. The American Journal of Applied sciences 7, 02 (2025), 5–11.
Dariusz Rafal Augustyn. 2017. Improvements of the Reactive Auto Scaling Method for Cloud Platform. In Computer Networks, Piotr Gaj, Andrzej Kwiecień, and Michał Sawicki (Eds.). Springer International Publishing, Cham, 422–431.
V Bandari. 2020. Cloud Workload Forecasting with Holt-Winters, State Space Model, and GRU. Journal of Artificial Intelligence and Machine Learning in Management 4, 1 (2020), 27–41.
Leo Breiman. 2001. Random Forests. Machine Learning 45, 1 (2001), 5–32.
Hui Chang. 2022. The Differences and Advantages between Cloud Services and Traditional Services. In 2022 8th Annual International Conference on Network and Information Systems for Computers (ICNISC). 497–499.
C. Chatfield. 2016. The Analysis of Time Series: An Introduction (6th ed.). Vol. 11. Chapman and Hall/CRC. 2834–2839 pages.
Tianqi Chen and Carlos Guestrin. 2016. XGBoost: A Scalable Tree Boosting System. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM, 785–794.
Kyunghyun Cho, Bart van Merrienboer, Yoshua Bengio, and Holger Schwenk. 2014. On the Properties of Neural Machine Translation: Encoder-Decoder Approaches. In Proceedings of EMNLP 2014. [link]
ClarkNet. 1995. ClarkNet HTTP Dataset. [link].
C. Cortes and V. Vapnik. 1995. Support-Vector Networks. Machine Learning 20, 3 (1995), 273–297.
Fatoumata Dama and Christine Sinoquet. 2021. Analysis and modeling to forecast in time series: a systematic review. CoRR abs/2104.00164 (2021). arXiv:2104.00164 [link]
D. Dickey and Wayne Fuller. 1979. Distribution of the Estimators for Autoregressive Time Series With a Unit Root. JASA. Journal of the American Statistical Association 74 (06 1979). DOI: 10.2307/2286348
Javad Dogani, Reza Namvar, and Farshad Khunjush. 2023. Auto-scaling techniques in container-based cloud and edge/fog computing: Taxonomy and survey. Computer Communications 209 (2023), 120–150.
N.R. Draper and H. Smith. 1998. Applied Regression Analysis. Wiley.
Cameron Fisher. 2018. Cloud versus on-premise computing. American Journal of Industrial and Business Management 8, 9 (2018), 1991–2006.
A Shaji George. 2024. The Cloud Comedown: Understanding the Emerging Trend of Cloud Exit Strategies. Partners Universal International Innovation Journal 2, 5 (2024), 1–32.
Stephen Haben, Martin Voss, and William Holderbaum. 2023. Previsão de Séries Temporais: Conceitos e Definições Essenciais. In Conceitos Básicos e Métodos em Previsão de Carga. Springer. DOI: 10.1007/978-3-031-27852-5_5
Nikolas Roman Herbst, Samuel Kounev, and Ralf Reussner. 2013. Elasticity in cloud computing: What it is, and what it is not. In 10th international conference on autonomic computing (ICAC 13). 23–27.
Sepp Hochreiter and Jürgen Schmidhuber. 1997. Long Short-Term Memory. Neural Computation 9, 8 (1997), 1735–1780.
Rob J. Hyndman and George Athanasopoulos. 2021. Forecasting: Principles and Practice (3rd ed.). OTexts, Melbourne, Australia. [link]
Deepak Janardhanan and Enda Barrett. 2017. CPU workload forecasting of machines in data centers using LSTM recurrent neural networks and ARIMA models. In 2017 12th International Conference for Internet Technology and Secured Transactions (ICITST). 55–60.
Kiran Jewargi. 2023. Public Cloud to Cloud Repatriation Trend. Sch J Eng Tech 1 (2023), 1–3.
Abul Khayer, Md. Shamim Talukder, Yukun Bao, and Md. Nahin Hossain. 2020. Cloud computing adoption and its impact on SMEs’ performance for cloud supported operations: A dual-stage analytical approach. Technology in Society 60 (2020), 101225.
Jitendra Kumar and Ashutosh Kumar Singh. 2018. Workload prediction in cloud using artificial neural network and adaptive differential evolution. Future Generation Computer Systems 81 (2018), 41–52.
Krishan Kumar, K. Gangadhara Rao, Suneetha Bulla, and D. Venkateswarulu. 2021. Forecasting of Cloud Computing Services Workload using Machine Learning. Turkish Journal of Computer and Mathematics Education (TURCOMAT) 12, 11 (2021), 4841–4846.
Denis Kwiatkowski, Peter C.B. Phillips, Peter Schmidt, and Yongcheol Shin. 1992. Testing the null hypothesis of stationarity against the alternative of a unit root: How sure are we that economic time series have a unit root? Journal of Econometrics 54, 1 (1992), 159–178. DOI: 10.1016/0304-4076(92)90104-Y
HS Lim and Wahidah Husain. 2013. A study on cloud computing adoption in e-business. Jurnal Sistem Informasi 9, 1 (2013), 13–17.
A. I. Maiyza, N. O. Korany, K. Banawan, et al. 2023. VTGAN: hybrid generative adversarial networks for cloud workload prediction. Journal of Cloud Computing 12 (2023), 97.
Mohammad Masdari and Afsane Khoshnevis. 2020. A survey and classification of the workload forecasting methods in cloud computing. Cluster Computing 23, 4 (2020), 2399–2424.
S. Meisenbacher, M. Turowski, K. Phipps, M. Rätz, D. Müller, V. Hagenmeyer, and R. Mikut. 2022. Review of automated time series forecasting pipelines. WIREs Data Mining and Knowledge Discovery 12 (2022). Issue 6.
Valter Rogério Messias, Julio Cezar Estrella, Ricardo Ehlers, Marcos José Santana, Regina Carlucci Santana, and Stephan Reiff-Marganiec. 2016. Combining time series prediction models using genetic algorithm to autoscalingWeb applications hosted in the cloud infrastructure. Neural Computing and Applications 27, 8 (1 Nov 2016), 2383–2406.
Abdulghafour Mohammad and Yasir Abbas. 2024. Key Challenges of Cloud Computing Resource Allocation in Small and Medium Enterprises. Digital 4, 2 (2024), 372–388.
Rafael Moreno-Vozmediano, Rubén S Montero, and Ignacio M Llorente. 2012. Iaas cloud architecture: From virtualized datacenters to federated cloud infrastructures. Computer 45, 12 (2012), 65–72.
P.A. Morettin and C.M.C. Toloi. 2018. Análise de séries temporais: modelos lineares univariados. BLUCHER.
A. Nielsen. 2019. Practical Time Series Analysis: Prediction with Statistics and Machine Learning. O’Reilly Media.
E. Patel and D. S. Kushwaha. 2022. A hybrid CNN-LSTM model for predicting server load in cloud computing. The Journal of Supercomputing 78 (2022), 1–30.
Damiano Perri, Marco Simonetti, Sergio Tasso, Federico Ragni, and Osvaldo Gervasi. 2021. Implementing a Scalable and Elastic Computing Environment Based on Cloud Containers. In Computational Science and Its Applications – ICCSA 2021, Osvaldo Gervasi, Beniamino Murgante, Sanjay Misra, Chiara Garau, Ivan Blečić, David Taniar, Bernady O. Apduhan, Ana Maria A. C. Rocha, Eufemia Tarantino, and Carmelo Maria Torre (Eds.). Springer International Publishing, Cham, 676–689.
Vladimir Podolskiy, Anshul Jindal, and Michael Gerndt. 2018. IaaS Reactive Autoscaling Performance Challenges. In 2018 IEEE 11th International Conference on Cloud Computing (CLOUD). 954–957.
E.G. Radhika and G. Sudha Sadasivam. 2021. A review on prediction based autoscaling techniques for heterogeneous applications in cloud environment. Materials Today: Proceedings 45 (2021), 2793–2800.
Krzysztof Rzadca, Pawel Findeisen, Jacek Swiderski, Przemyslaw Zych, Przemyslaw Broniek, Jarek Kusmierek, Pawel Nowak, Beata Strack, Piotr Witusowski, Steven Hand, et al. 2020. Autopilot: workload autoscaling at google. In Proceedings of the Fifteenth European Conference on Computer Systems. 1–16.
P. Singh, P. Gupta, and K. Jyoti. 2018. Tasm: Technocrat ARIMA and SVR Model for Workload Prediction of Web Applications in Cloud. Cluster Computing 22, 2 (2018), 619–633.
Parminder Singh, Pooja Gupta, and Kiran Jyoti. 2019. TASM: technocrat ARIMA and SVR model for workload prediction of web applications in cloud. Cluster Computing 22, 2 (1 Jun 2019), 619–633.
Emmanuel Tachu. 2022. A quantitative study of the relationship between cloud flexibility and on-premise flexibility. Issues in Information Systems 23, 1 (2022).
R.S. Tsay and R. Chen. 2018. Nonlinear Time Series Analysis. Wiley.
A. Ullah, J. Li, Y. Shen, et al. 2018. A control theoretical view of cloud elasticity: taxonomy, survey and challenges. Cluster Computing 21 (2018), 1735–1764.
FeiWang and LiGang Zhao. 2019. Complexity Analysis of Air Traffic Flow Based on Sample Entropy. In Chinese Control And Decision Conference. 5368–5371.
Ang Xuan, Mengmeng Yin, Yupei Li, Xiyu Chen, and Zhenliang Ma. 2022. A comprehensive evaluation of statistical, machine learning and deep learning models for time series prediction. In 2022 7th International Conference on Data Science and Machine Learning Applications (CDMA). 55–60.
Ming Yan, XiaoMeng Liang, ZhiHui Lu, Jie Wu, and Wei Zhang. 2021. HANSEL: Adaptive horizontal scaling of microservices using Bi-LSTM. Applied Soft Computing 105 (2021), 107216.
Natalya Yezhkova. 2024. Assessing the Scale of Workload Repatriation: Insights from IDC’s Server and Storage Workloads Surveys, 1H23 and 2H23. Technical Report US50903124. International Data Corporation (IDC). [link] IDC Survey Report.
Nathalie Zeghmouli. 2025. The cloud computing dilemma for financial services institutions. Journal of Securities Operations & Custody 17, 2 (2025), 166–177.
Minqi Zhou, Rong Zhang, Dadan Zeng, and Weining Qian. 2010. Services in the cloud computing era: A survey. In 2010 4th international universal communication symposium. IEEE, 40–46.
Ding Zou, Wei Lu, Zhibo Zhu, Xingyu Lu, Jun Zhou, Xiaojin Wang, Kangyu Liu, Kefan Wang, Renen Sun, and Haiqing Wang. 2024. OptScaler: A Collaborative Framework for Robust Autoscaling in the Cloud. Proc. VLDB Endow. 17, 12 (2024), 4090–4103.
anon. 2025. LSTM-based Forecasting for Non-linear Workloads in On-premises Software Systems. DOI: 10.5281/zenodo.15285346 Computational notebook.
Kuciuk Artiom. 2025. Cloud Migration Framework: Transitioning from On-Premises to Azure Cloud for Improved System Reliability and Scalability. The American Journal of Applied sciences 7, 02 (2025), 5–11.
Dariusz Rafal Augustyn. 2017. Improvements of the Reactive Auto Scaling Method for Cloud Platform. In Computer Networks, Piotr Gaj, Andrzej Kwiecień, and Michał Sawicki (Eds.). Springer International Publishing, Cham, 422–431.
V Bandari. 2020. Cloud Workload Forecasting with Holt-Winters, State Space Model, and GRU. Journal of Artificial Intelligence and Machine Learning in Management 4, 1 (2020), 27–41.
Leo Breiman. 2001. Random Forests. Machine Learning 45, 1 (2001), 5–32.
Hui Chang. 2022. The Differences and Advantages between Cloud Services and Traditional Services. In 2022 8th Annual International Conference on Network and Information Systems for Computers (ICNISC). 497–499.
C. Chatfield. 2016. The Analysis of Time Series: An Introduction (6th ed.). Vol. 11. Chapman and Hall/CRC. 2834–2839 pages.
Tianqi Chen and Carlos Guestrin. 2016. XGBoost: A Scalable Tree Boosting System. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM, 785–794.
Kyunghyun Cho, Bart van Merrienboer, Yoshua Bengio, and Holger Schwenk. 2014. On the Properties of Neural Machine Translation: Encoder-Decoder Approaches. In Proceedings of EMNLP 2014. [link]
ClarkNet. 1995. ClarkNet HTTP Dataset. [link].
C. Cortes and V. Vapnik. 1995. Support-Vector Networks. Machine Learning 20, 3 (1995), 273–297.
Fatoumata Dama and Christine Sinoquet. 2021. Analysis and modeling to forecast in time series: a systematic review. CoRR abs/2104.00164 (2021). arXiv:2104.00164 [link]
D. Dickey and Wayne Fuller. 1979. Distribution of the Estimators for Autoregressive Time Series With a Unit Root. JASA. Journal of the American Statistical Association 74 (06 1979). DOI: 10.2307/2286348
Javad Dogani, Reza Namvar, and Farshad Khunjush. 2023. Auto-scaling techniques in container-based cloud and edge/fog computing: Taxonomy and survey. Computer Communications 209 (2023), 120–150.
N.R. Draper and H. Smith. 1998. Applied Regression Analysis. Wiley.
Cameron Fisher. 2018. Cloud versus on-premise computing. American Journal of Industrial and Business Management 8, 9 (2018), 1991–2006.
A Shaji George. 2024. The Cloud Comedown: Understanding the Emerging Trend of Cloud Exit Strategies. Partners Universal International Innovation Journal 2, 5 (2024), 1–32.
Stephen Haben, Martin Voss, and William Holderbaum. 2023. Previsão de Séries Temporais: Conceitos e Definições Essenciais. In Conceitos Básicos e Métodos em Previsão de Carga. Springer. DOI: 10.1007/978-3-031-27852-5_5
Nikolas Roman Herbst, Samuel Kounev, and Ralf Reussner. 2013. Elasticity in cloud computing: What it is, and what it is not. In 10th international conference on autonomic computing (ICAC 13). 23–27.
Sepp Hochreiter and Jürgen Schmidhuber. 1997. Long Short-Term Memory. Neural Computation 9, 8 (1997), 1735–1780.
Rob J. Hyndman and George Athanasopoulos. 2021. Forecasting: Principles and Practice (3rd ed.). OTexts, Melbourne, Australia. [link]
Deepak Janardhanan and Enda Barrett. 2017. CPU workload forecasting of machines in data centers using LSTM recurrent neural networks and ARIMA models. In 2017 12th International Conference for Internet Technology and Secured Transactions (ICITST). 55–60.
Kiran Jewargi. 2023. Public Cloud to Cloud Repatriation Trend. Sch J Eng Tech 1 (2023), 1–3.
Abul Khayer, Md. Shamim Talukder, Yukun Bao, and Md. Nahin Hossain. 2020. Cloud computing adoption and its impact on SMEs’ performance for cloud supported operations: A dual-stage analytical approach. Technology in Society 60 (2020), 101225.
Jitendra Kumar and Ashutosh Kumar Singh. 2018. Workload prediction in cloud using artificial neural network and adaptive differential evolution. Future Generation Computer Systems 81 (2018), 41–52.
Krishan Kumar, K. Gangadhara Rao, Suneetha Bulla, and D. Venkateswarulu. 2021. Forecasting of Cloud Computing Services Workload using Machine Learning. Turkish Journal of Computer and Mathematics Education (TURCOMAT) 12, 11 (2021), 4841–4846.
Denis Kwiatkowski, Peter C.B. Phillips, Peter Schmidt, and Yongcheol Shin. 1992. Testing the null hypothesis of stationarity against the alternative of a unit root: How sure are we that economic time series have a unit root? Journal of Econometrics 54, 1 (1992), 159–178. DOI: 10.1016/0304-4076(92)90104-Y
HS Lim and Wahidah Husain. 2013. A study on cloud computing adoption in e-business. Jurnal Sistem Informasi 9, 1 (2013), 13–17.
A. I. Maiyza, N. O. Korany, K. Banawan, et al. 2023. VTGAN: hybrid generative adversarial networks for cloud workload prediction. Journal of Cloud Computing 12 (2023), 97.
Mohammad Masdari and Afsane Khoshnevis. 2020. A survey and classification of the workload forecasting methods in cloud computing. Cluster Computing 23, 4 (2020), 2399–2424.
S. Meisenbacher, M. Turowski, K. Phipps, M. Rätz, D. Müller, V. Hagenmeyer, and R. Mikut. 2022. Review of automated time series forecasting pipelines. WIREs Data Mining and Knowledge Discovery 12 (2022). Issue 6.
Valter Rogério Messias, Julio Cezar Estrella, Ricardo Ehlers, Marcos José Santana, Regina Carlucci Santana, and Stephan Reiff-Marganiec. 2016. Combining time series prediction models using genetic algorithm to autoscalingWeb applications hosted in the cloud infrastructure. Neural Computing and Applications 27, 8 (1 Nov 2016), 2383–2406.
Abdulghafour Mohammad and Yasir Abbas. 2024. Key Challenges of Cloud Computing Resource Allocation in Small and Medium Enterprises. Digital 4, 2 (2024), 372–388.
Rafael Moreno-Vozmediano, Rubén S Montero, and Ignacio M Llorente. 2012. Iaas cloud architecture: From virtualized datacenters to federated cloud infrastructures. Computer 45, 12 (2012), 65–72.
P.A. Morettin and C.M.C. Toloi. 2018. Análise de séries temporais: modelos lineares univariados. BLUCHER.
A. Nielsen. 2019. Practical Time Series Analysis: Prediction with Statistics and Machine Learning. O’Reilly Media.
E. Patel and D. S. Kushwaha. 2022. A hybrid CNN-LSTM model for predicting server load in cloud computing. The Journal of Supercomputing 78 (2022), 1–30.
Damiano Perri, Marco Simonetti, Sergio Tasso, Federico Ragni, and Osvaldo Gervasi. 2021. Implementing a Scalable and Elastic Computing Environment Based on Cloud Containers. In Computational Science and Its Applications – ICCSA 2021, Osvaldo Gervasi, Beniamino Murgante, Sanjay Misra, Chiara Garau, Ivan Blečić, David Taniar, Bernady O. Apduhan, Ana Maria A. C. Rocha, Eufemia Tarantino, and Carmelo Maria Torre (Eds.). Springer International Publishing, Cham, 676–689.
Vladimir Podolskiy, Anshul Jindal, and Michael Gerndt. 2018. IaaS Reactive Autoscaling Performance Challenges. In 2018 IEEE 11th International Conference on Cloud Computing (CLOUD). 954–957.
E.G. Radhika and G. Sudha Sadasivam. 2021. A review on prediction based autoscaling techniques for heterogeneous applications in cloud environment. Materials Today: Proceedings 45 (2021), 2793–2800.
Krzysztof Rzadca, Pawel Findeisen, Jacek Swiderski, Przemyslaw Zych, Przemyslaw Broniek, Jarek Kusmierek, Pawel Nowak, Beata Strack, Piotr Witusowski, Steven Hand, et al. 2020. Autopilot: workload autoscaling at google. In Proceedings of the Fifteenth European Conference on Computer Systems. 1–16.
P. Singh, P. Gupta, and K. Jyoti. 2018. Tasm: Technocrat ARIMA and SVR Model for Workload Prediction of Web Applications in Cloud. Cluster Computing 22, 2 (2018), 619–633.
Parminder Singh, Pooja Gupta, and Kiran Jyoti. 2019. TASM: technocrat ARIMA and SVR model for workload prediction of web applications in cloud. Cluster Computing 22, 2 (1 Jun 2019), 619–633.
Emmanuel Tachu. 2022. A quantitative study of the relationship between cloud flexibility and on-premise flexibility. Issues in Information Systems 23, 1 (2022).
R.S. Tsay and R. Chen. 2018. Nonlinear Time Series Analysis. Wiley.
A. Ullah, J. Li, Y. Shen, et al. 2018. A control theoretical view of cloud elasticity: taxonomy, survey and challenges. Cluster Computing 21 (2018), 1735–1764.
FeiWang and LiGang Zhao. 2019. Complexity Analysis of Air Traffic Flow Based on Sample Entropy. In Chinese Control And Decision Conference. 5368–5371.
Ang Xuan, Mengmeng Yin, Yupei Li, Xiyu Chen, and Zhenliang Ma. 2022. A comprehensive evaluation of statistical, machine learning and deep learning models for time series prediction. In 2022 7th International Conference on Data Science and Machine Learning Applications (CDMA). 55–60.
Ming Yan, XiaoMeng Liang, ZhiHui Lu, Jie Wu, and Wei Zhang. 2021. HANSEL: Adaptive horizontal scaling of microservices using Bi-LSTM. Applied Soft Computing 105 (2021), 107216.
Natalya Yezhkova. 2024. Assessing the Scale of Workload Repatriation: Insights from IDC’s Server and Storage Workloads Surveys, 1H23 and 2H23. Technical Report US50903124. International Data Corporation (IDC). [link] IDC Survey Report.
Nathalie Zeghmouli. 2025. The cloud computing dilemma for financial services institutions. Journal of Securities Operations & Custody 17, 2 (2025), 166–177.
Minqi Zhou, Rong Zhang, Dadan Zeng, and Weining Qian. 2010. Services in the cloud computing era: A survey. In 2010 4th international universal communication symposium. IEEE, 40–46.
Ding Zou, Wei Lu, Zhibo Zhu, Xingyu Lu, Jun Zhou, Xiaojin Wang, Kangyu Liu, Kefan Wang, Renen Sun, and Haiqing Wang. 2024. OptScaler: A Collaborative Framework for Robust Autoscaling in the Cloud. Proc. VLDB Endow. 17, 12 (2024), 4090–4103.
Publicado
22/09/2025
Como Citar
XAVIER, Rafael; CAFEO, Bruno; FONSECA, Baldoino; BAIA, Davy; CIRILO, Elder.
LSTM-based Forecasting for Non-linear Workloads in On-premises Software Systems. In: SIMPÓSIO BRASILEIRO DE ENGENHARIA DE SOFTWARE (SBES), 39. , 2025, Recife/PE.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2025
.
p. 260-270.
ISSN 2833-0633.
DOI: https://doi.org/10.5753/sbes.2025.9917.
