Cloud computing and service computing can be integrated into a distributed computing platform which provides a set of services to their users. Different services could cooperate with each other to accomplish more intricate tasks. However, third-party cloud services may cause failures or quality violation of workflow applications. Especially in some critical business scenarios, service performance violations need to be proactively prevented. It is important to identify the violation of service quality for service management. In this work, we propose a new SLA (Service Level Agreement) monitoring model to manage quality of workflow. Multivariate control chart is employed to detect the violations. Violation identifying approach based multivariate control chart can be used to avoid failures of workflow applications in an effective way.