云环境下实现容器部署的加速粒子群优化算法

基于容器的微服务部署是一个具有挑战性的问题，为获得更好的用户体验并给云供应商带来更多的利润，需要在降低微服务的故障率和减少响应时间的同时提高资源利用率。提出了一种改进的加速粒子群优化算法，用于解决集群中微服务容器部署的多目标优化问题。该算法通过考虑微服务之间的调用关系，使得容器聚集在一起，从而降低服务的数据传输成本、减少故障率，并提高集群资源利用率。与现有部署算法相比，实验结果表明，所提出的优化算法在服务间的数据传输开销、故障率和资源利用率等性能指标上有明显改善。具体表现在：容器聚集度的提升达到40%以上，数据传输消耗平均有提升4%以上，故障率减少10%～20%,利用率提升3%左右。

Accelerated particle swarm optimization algorithm for container deployment in cloud environments

Container-based microservice deployment is a challenging problem that aims to improve user experience and increase cloud providers'' profitability by reducing microservice failure rates and response times while maximizing resource utilization. This paper presented an enhanced accelerated particle swarm optimization algorithm to tackle the multi-objective optimization problem of microservice container deployment in a cluster. By considering the invocation relationships between microservices, the algorithm facilitated the aggregation of containers, thereby reducing data transmission costs, lowering failure rates, and enhancing cluster resource utilization. Experimental results demonstrate that the proposed optimization algorithm yields significant improvements in performance measures including data transmission overhead, failure rate, and resource utilization when compared to existing deployment algorithms. Specifically, the algorithm achieves a container aggregation improvement exceeding 40%, an average increase in data transmission consumption of over 4%, a decrease in failure rate by 10% to 20%, and an increase in utilization rate by roughly 3%. The aforementioned findings attest to the efficacy of the proposed algorithm.