Currently, the scalability of computer systems on the cloud platform is critical to the successful processing and presentation of constantly changing information on the Internet,in consequence of the fact that maintaining adequate resources to handle peak loads on infrastructure significantly increases the high cost of its maintenance. In such circumstances,the scalability of cloud computing technology provides its users with real-time access to computing resources, depending on the required load. This article studies the architecture of the dynamic scaling Web applications based on thresholds in a virtualized cloud environment. The use of technology scaling load demonstrates compelling advantages, including the ability to handle sudden load on the system, provide resources based on user demand and maintaining more efficient use of resources, thus reducing the material costs of infrastructure maintenance and its management.