Design and Implementation of HCI Platform for Remote Experiments

This paper designs and implements a remote experiment platform based on hyper-converged infrastructure and virtual private network technology. The platform addresses limitations of traditional platforms in resource allocation, data storage, and network stability. By adopting key technologies such as distributed computing, software-defined storage, high-availability clusters, and intelligent load balancing, the platform significantly improves reliability and resource utilization. A Markov chain is used to model fault recovery capability, and the M/G/1 queuing model is applied to optimize the load capacity of node servers. Experimental results show that the platform achieves significant improvements in data access latency, fault recovery time, and concurrent processing capability compared with conventional infrastructure. Teaching practice feedback is positive, indicating considerable application potential.