Distributed Storage Method for Individual Monitoring Information of Cloud Platform Network Node
Aiming at the problems of unbalanced load and high packet loss rate of monitoring information storage,a distributed storage method for individual monitoring information of cloud platform network node is proposed.The monitoring information source is regarded as a set composed of multiple events,the maximum value of information entropy is calculated through the e-vent probability distribution,and the monitoring information compression and redundancy removal are completed according to the value.The remote server load is calculated by using the remote data backup and recovery algorithm,and the backup task al-location is adjusted in real time according to the load situation to ensure load balance and avoid information loss.We establish a three-tier cloud storage architecture including software services,platform services and infrastructure services,and analyze the functions at different levels.The distributed storage process is divided into access layer,connection layer and archive layer.By calculating the information access probability,it can judge whether the monitoring information meets the storage requirements.For the information that meets the requirements,it can be stored directly.For the information that does not meet the condi-tions,it can use the granularity rate to adjust the elastic expectation value,reduce the information storage congestion,and real-ize distributed storage.Experiments show that the proposed method has good compression effect,can ensure the load balance of storage nodes,and reduce the loss of monitoring information.
万方数据
维普
