As a key type of applications running in cloud environments, big data analytics are being generously hosted in virtualized platforms. However, in many cloud data centers, such data analytics clusters are not used efficiently and thereby cost extra overhead. The cost-effectiveness of such data analytics clusters in cloud data centers is the key concern of this paper. In this paper, we design, implement and evaluate the InSTechAH, an autoscale schemascheme for a Hadoop system in a private cloud, which attempts to improve the resource utilization in cloud data centers as well as to maintain required quality of services. In this system, we design the multilayer node model to reduce interference from other services by automatically scaling the clusters according to the autoscale algorithm we introduced. We then build the resource scheduling model which use prediction based scheduling method to reduce the cost brought by scaling. We evaluate our schemascheme partly on a real data trace and partly on simulation, with Hadoop as the parallel data analytics frameworks and OpenStack as the cloud management architecture, to show the efficiency of InSTechAH system.