Abstract: The multi-tenant data stored in cloud using replica technology must be reasonably placed, meet the requirements for data access by tenants and management by service providers. According to the characteristics of multi-tenant data and the load of the nodes, for the outweight nodes and the ultralight nodes, through adjusting the number and position of the replicas to maintenance and optimization the placement strategy so that meet the SLA requirements at the same time minimize the overall cost, realize the balance between the quality of service and the cost of management. Experimental results through comparison with random placement strategy and greedy placement policy demonstrate the feasibility and effectiveness of the proposed strategy.