Hashing Heterogeneity Aware Dynamic Capacity Provisioning in the Cloud

Data centers consume more amounts of energy in terms of power supply and cooling required processing the production machines. HHADC is a hopeful approach for reducing energy consumption by dynamically adjusting the number of working machines to match resource demands. However, existing solutions have not fully satisfied the heterogeneity of workload and machine hardware found in data center environment. The data centers often comprise capability of the heterogeneous machines with different configuration capacities and power consumption characteristics. The Memory details of both data center production machines and workloads will lead to sub-optimal power-savings and long delays, due to inconvenience between work load requirements and the resources offered by the production machines. To address this limitation, we create a HHADC, scheme for cloud data centers. First use the K-means clustering algorithm to divide processing load into distinct classes with same characteristics in terms of resource and performance needed. Then we have a technique HHADC technique that dynamically adjusting the number of machines to reduce the total power energy consumption and responding delay. HHADC can reduce energy by 25 percent compared to Harmony solutions.