Safe Execution of Parallel Component Applications on an Unstable Heterogeneous Cluster Platform
Yunfeng Peng, Congming Shi, Guowei Gao, Jianan Wang, Hai Liu · Advances in transdisciplinary engineering · 2024
Parallel component technology improves the development efficiency and performance of parallel software. Parallel component applications often need to be deployed on unstable heterogeneous cluster platforms. By analyzing the running mechanism of parallel component applications, this paper realizes the method of optimizing the deployment and running management of parallel components on the basis of modelling and managing the computing resources of heterogeneous platforms. This method can select nodes with light load to deploy parallel component code, and monitor the status of computing resources during the component operation. Our method can find the unavailable computing resources in time, and redeploy the parallel components running on them to other nodes with light load, which reduces the adverse impact of computing resource failure on the operation of component applications, and ensures the safe execution of parallel component applications on unstable heterogeneous platforms and achieves the expected results.