International Journal of Advanced Technology and Engineering Exploration ISSN (Print): 2394-5443    ISSN (Online): 2394-7454 Volume-13 Issue-138 May-2026
  1. 4774
    Citations
  2. 2.8
    CiteScore
CAKS: a real-time carbon-aware Kubernetes scheduler for heterogeneous cloud environments

Kiran Kumari1 and Dhananjay Dakhane2

Research Scholar, Department of Computer Engineering,Ramrao Adik Institute of Technology, D.Y. Patil Deemed to be University, Nerul, Navi Mumbai,Maharashtra,India1
Professor, Department of Computer Science and Engineering,Ramrao Adik Institute of Technology, D.Y. Patil Deemed to be University, Nerul, Navi Mumbai,Maharashtra,India2
Corresponding Author : Kiran Kumari

Recieved : 26-October-2025; Revised : 14-May-2026; Accepted : 18-May-2026

Abstract

Kubernetes is a widely adopted platform for container orchestration, driven by the growing popularity of container-based services. It is a preferred choice in industry, and its adoption in public cloud environments has drawn significant attention to the potential impact of carbon emissions (CE). Considerable research has been conducted to minimize the CE of jobs and applications. Existing approaches primarily shift workloads temporally and geographically to regions with lower carbon intensity or reduce the resource utilization of Kubernetes nodes. However, hardware heterogeneity among nodes, challenges in cloud environments, and cluster-level CE are often overlooked. To address these issues, a real-time carbon-aware Kubernetes scheduler (CAKS) and a DaemonSet are proposed to minimize CE in a Kubernetes cluster (KC). The carbon-greedy CAKS schedules incoming jobs to nodes with lower CE, whereas the DaemonSet running on each heterogeneous node collects real-time CE data. CAKS is implemented using the Kubernetes scheduling framework to preserve its lightweight design and is deployed on the Amazon Elastic Kubernetes Service (EKS), a service within Amazon Web Services (AWS). Experimental results demonstrate a reduction of up to 2% in cluster-level CE compared to the Kubernetes Default Scheduler (KDS).

Keywords

Kubernetes scheduling, Carbon-aware computing, Container orchestration, Carbon emissions reduction, Amazon elastic Kubernetes service, Heterogeneous cloud environments.

Cite this article

Kumari K, Dakhane D. CAKS: a real-time carbon-aware Kubernetes scheduler for heterogeneous cloud environments. International Journal of Advanced Technology and Engineering Exploration. 2026;13(138):731-748. DOI : 10.19101/IJATEE.2025.121221412

References
[1]
Thaqi R, Kadriu M, Krasniqi B, Rexha B. On-premises versus cloud computing: a comparative analysis of energy consumption. In international conference on renewable energies and smart technologies (REST) 2024 (pp. 1-5). IEEE.
[Crossref] [Google Scholar]
[2]
Hewage TB, Ilager S, Rodriguez MA, Buyya R. A framework for carbon-aware real-time workload management in clouds using renewables-driven cores. IEEE Transactions on Computers. 2025; 74(8):2757-71.
[Crossref] [Google Scholar]
[3]
Beena BM, Ranga PC, Holimath V, Sridhar S, Kamble SS, Shendre SP, et al. Adaptive energy optimization in cloud computing through containerization. IEEE Access. 2025; 13:159548-65.
[Crossref] [Google Scholar]
[4]
Seo E, Elmroth E. Pioneering eco-efficiency in cloud computing: the carbon-conscious federated reinforcement learning (CCFRL) approach. IEEE Internet of Things Journal. 2025; 12(7):8958 -79.
[Crossref] [Google Scholar]
[5]
Menouer T, Cérin C, Darmon P. Koptim: kubernetes optimization framework. In international parallel and distributed processing symposium workshops (IPDPSW) 2024 (pp. 900-8). IEEE.
[Crossref] [Google Scholar]
[6]
She R, Wu Y, Cui E. Energy-aware scheduling for reliability-oriented real-time parallel applications allocation on heterogeneous computing systems. Future Generation Computer Systems. 2025; 168:107738.
[Crossref] [Google Scholar]
[7]
Ma H, Ding A. Method for evaluation on energy consumption of cloud computing data center based on deep reinforcement learning. Electric Power Systems Research. 2022; 208:107899.
[Crossref] [Google Scholar]
[8]
Sanjalawe Y, Al-e’mari S, Fraihat S, Makhadmeh S. AI-driven job scheduling in cloud computing: a comprehensive review. Artificial Intelligence Review. 2025; 58(7):1-113.
[Crossref] [Google Scholar]
[9]
Oukil K, Chiba Z, Moussaid K, Abghour N. Green cloud computing: a literature review on the current state of research. In international symposium on networks, computers and communications (ISNCC) 2024 (pp. 1-6). IEEE.
[Crossref] [Google Scholar]
[10]
Rao W, Li H. Energy-aware scheduling algorithm for microservices in kubernetes clouds. Journal of Grid Computing. 2025; 23(1):1-22.
[Crossref] [Google Scholar]
[11]
Marchese A, Tomarchio O. SLO-aware container orchestration on Kubernetes clusters. In 18th international conference on cloud computing (CLOUD) 2025 (pp. 318-27). IEEE.
[Crossref] [Google Scholar]
[12]
Gouaouri MD, Bagaa M, Ouahouah S, Ouameur MA, Ksentini A. Enabling power-awareness for Kubernetes scheduling through multi-criteria optimization. In ICC international conference on communications 2025 (pp. 1500-5). IEEE.
[Crossref] [Google Scholar]
[13]
Didas M, Chali FH, Noe E. Facets related to challenges and prospects of cloud computing adoption in resource-constrained settings: systematic review analysis. International Journal of Advanced Computer Research. 2024;14(66):18-41.
[Crossref] [Google Scholar]
[14]
Chiorescu R, Djemame K. Scheduling energy-aware multi-function serverless workloads in OpenFaaS. In international conference on the economics of grids, clouds, systems, and services 2024 (pp. 137-49). Cham: Springer Nature Switzerland.
[Crossref] [Google Scholar]
[15]
Sukprasert T, Souza A, Bashir N, Irwin D, Shenoy P. On the limitations of carbon-aware temporal and spatial workload shifting in the cloud. In proceedings of the nineteenth European conference on computer systems 2024 (pp. 924-41).
[Crossref] [Google Scholar]
[16]
Xiao D, Shen W, Yi H, Zhang W. Load-balanced scheduling optimization strategy for high-communication tasks in Kubernetes with RDMA. Computer Communications. 2025; 241:108260.
[Crossref] [Google Scholar]
[17]
https://kubernetes.io/docs/concepts/scheduling-eviction/scheduling-framework/. Accessed 24 February 2026.
[18]
Centofanti C, Santos J, Gudepu V, Kondepu K. Impact of power consumption in containerized clouds: a comprehensive analysis of open-source power measurement tools. Computer Networks. 2024; 245:1-15.
[Crossref] [Google Scholar]
[19]
https://docs.watttime.org/#tag/Introduction. Accessed 24 February 2026.
[20]
https://www.electricitymaps.com/data/methodology. Accessed 24 February 2026.
[21]
Haghshenas K, Taheri S, Goudarzi M, Mohammadi S. Infrastructure aware heterogeneous-workloads scheduling for data center energy cost minimization. IEEE Transactions on Cloud Computing. 2020; 10(2):972-83.
[Crossref] [Google Scholar]
[22]
Den TJ, Braakman P, Durieux T. SCALE: a CO2-aware scheduler for openshift at ING. In companion proceedings of the 32nd international conference on the foundations of software engineering 2024 (pp. 429-39). ACM.
[Crossref] [Google Scholar]
[23]
Hanafy WA, Liang Q, Bashir N, Souza A, Irwin D, Shenoy P. Going green for less green: optimizing the cost of reducing cloud carbon emissions. In proceedings of the 29th international conference on architectural support for programming languages and operating systems. 2024 (pp. 479-96). ACM.
[Crossref] [Google Scholar]
[24]
Bostandoost R, Hanafy WA, Lechowicz A, Bashir N, Shenoy P, Hajiesmaili M. Data-driven algorithm selection for carbon-aware scheduling. ACM SIGENERGY Energy Informatics Review. 2024; 4(5):148-53.
[Crossref] [Google Scholar]
[25]
Lechowicz A, Shenoy R, Bashir N, Hajiesmaili M, Wierman A, Delimitrou C. Carbon-and precedence-aware scheduling for data processing clusters. In proceedings of the carbon and precedence-aware scheduling for data processing clusters conference 2025 (pp. 1241-44). ACM.
[Crossref] [Google Scholar]
[26]
Beena BM, Prashanth CS, Manideep TS, Saragadam S, Karthik G. A green cloud-based framework for energy-efficient task scheduling using carbon intensity data for heterogeneous cloud servers. IEEE Access. 2025; 13:73916-38.
[Crossref] [Google Scholar]
[27]
Bahreini T, Tantawi AN, Tardieu O. Caspian: a carbon-aware workload scheduler in multi-cluster kubernetes environments. In 32nd international conference on modeling, analysis and simulation of computer and telecommunication systems (MASCOTS) 2024 (pp. 1-8). IEEE.
[Crossref] [Google Scholar]
[28]
Piontek T, Haghshenas K, Aiello M. Carbon emission-aware job scheduling for Kubernetes deployments. The Journal of Supercomputing. 2024; 80(1):549-69.
[Crossref] [Google Scholar]
[29]
Souza A, Jasoria S, Chakrabarty B, Bridgwater A, Lundberg A, Skogh F, et al. Casper: carbon-aware scheduling and provisioning for distributed web services. In proceedings of the 14th international green and sustainable computing conference 2023 (pp. 67-73). ACM.
[Crossref] [Google Scholar]
[30]
Hanafy WA, Liang Q, Bashir N, Irwin D, Shenoy P. Carbonscaler: leveraging cloud workload elasticity for optimizing carbon-efficiency. Proceedings of the ACM on Measurement and Analysis of Computing Systems. 2023; 7(3):1-28.
[Crossref] [Google Scholar]
[31]
Furnadzhiev R, Shopov M, Kakanakov N. Efficient orchestration of distributed workloads in multi-region kubernetes cluster. Computers. 2025; 14(4):1-18.
[Crossref] [Google Scholar]
[32]
Ghafouri S, Abdipoor S, Doyle J. Smart-kube: energy-aware and fair kubernetes job scheduler using deep reinforcement learning. In 8th international conference on smart cloud (SmartCloud) 2023 (pp. 154-63). IEEE.
[Crossref] [Google Scholar]
[33]
Lai WK, Wang YC, Wei SC. Delay-aware container scheduling in kubernetes. IEEE Internet of Things Journal. 2023; 10(13):11813-24.
[Crossref] [Google Scholar]
[34]
Carvalho M, Macedo DF. Container scheduling in co-located environments using QoE awareness. IEEE Transactions on Network and Service Management. 2023; 20(3):3247-60.
[Crossref] [Google Scholar]
[35]
Di CN, Poltronieri F, Santos J, Zaccarini M, Tortonesi M, Stefanelli C, et al. Multi-objective scheduling and resource allocation of kubernetes replicas across the compute continuum. In 20th international conference on network and service management (CNSM) 2024 (pp. 1-9). IEEE.
[Crossref] [Google Scholar]