| CPC G06F 9/5077 (2013.01) | 14 Claims |
|
1. A method of optimizing storing efficiency of a cloud computing platform used in a server; the server presents desktop as a server (DaaS) layer and infrastructure as a server (IaaS) layer; the DaaS layer provides a virtual desktop, which is directly accessed by an authorized user through the cloud computing platform; the server comprises a storage medium with computer programs and a processor; the processor executes computer programs in the storage medium to implement the following steps:
obtaining a first input/output operations per second (IOPS) value of a virtual disk while a user accesses the virtual desktop, being one of a plurality of virtual disks comprised in an instance, by the processor;
determining whether the first IOPS value is larger than a predefined threshold value in the IaaS layer, by the processor;
obtaining a warning message, which is generated in the IaaS layer, at the DaaS layer using a Restful Application Programming Interface (API) in response to that the first IOPS value is determined to be larger than the predefined threshold value;
training models in the DaaS layer of the server, each of which corresponds to different predefined time durations in one day, by the processor;
matching one of the models based on one of the predefined time durations in the DaaS layer of the server, by the processor;
obtaining a predictive IOPS value based on an output of a matched model of one of the models, by the processor; and
setting the predictive IOPS value as a maximum IOPS value of the virtual disk, by the processor;
wherein the step of training the models, each of which corresponds to the different predefined time durations in one day, by the processor comprises:
recording IOPS values of the virtual disk at time intervals in a same predefined time duration in the storage medium, by the processor;
and
setting an average IOPS value as the predictive IOPS value, by the processor.
|