The volume of datasets to be handled by scientific applications is increasing abruptly. Data-intensive sciences challenged by the big data problems need more elastic and scalable computing infrastructure than traditional infrastructure adhesive to compute-intensive computing applications. Cloud computing is rising alternative to existing compute-intensive high performance computing infrastructures. In this work we present an astronomical time series data analysis on cloud computing as a typical data-intensive scientific application. We implemented a private IaaS cloud which is virtual resource provision service to data analysis applications. We utilize OpenNebula as a virtual machine man- ager and implemented virtual cluster service which gives virtual private cluster instances based on user demand. Detecting variable bright stars from SuperWASP time series data is successfully done in our virtual clusters, which shows the viability of cloud computing for data-intensive sciences.