The performance and operational cost of a cloud data center may worsen due to degenerative resource allocation. A novel scheme ’C-DHLB’ Classified Dynamic Hierarchical Load Balancer is introduced to address this issue by the means of resource requirement analysis and allocating an optimized number of computing resources to every application. We formulate the traffic scalability issue as a virtual machine placement optimization with multi-dimensional resource constraints. The model allocates the exact amount of resources and pick out the most optimal host to achieve immediate load balancing effect. Performance evaluation demonstrates that proposed model achieves high throughput, avoid traffic overflow and reduce upto 72.58% of the traffic in the network.