Conditional Random Field-Based Incremental Auto Scaling Algorithm to Enhance Workflow Scheduling in Cloud Computing

Current internet environment has demanded the need for cloud computing technology in handling large information. The paper’s central theme is to evaluate the effectiveness of resources in managing scientific processes in a cloud environment by scheduling, capacity planning, and auto-scaling. Current systems have been linked with high costs because of disturbance of the flow, unsuccessful jobs, inability to meet certain cutoff times and long make span times. Therefore, to overcome these challenges, this research presents the Conditional Random Field-based Incremental Auto-scaling algorithm (CRF-IASA) to improve the cloud efficiency. The CRF-IASA uniquely leverages CRF’s ability to model complex dependencies and interrelationships in resource management. Unlike traditional methods, which often rely on static or heuristic-based decision-making, CRF-IASA dynamically adapts to workload changes by effectively balancing data, memory, and CPU demands. The evaluation of the CRF-IASA is done with a help of the following sample workflows: A Cybershake 1000, Montage 1000, LIGO 1000 and Epigenome 997. Thus, according to the values obtained in the experiment, it is proved that the proposed auto-scaling algorithm is effective with the identified evaluation criteria: This is the total time needed to complete all activities, the energy used in the computing process, the costs incurred, and the rate at which the activities are completed in a schedule. Therefore, the study demonstrates the potential of utilizing CRF-IASA to obtain increased results concerning cloud performance and resource’s consumption with less costs in a more adaptable way.