Scalability on the Cloud

In contrast to traditional IDCs, cloud resources provide enhanced scalability. Based on different requirements, the scalability can be classified into the following types:

• Vertical scaling is applicable to a single application, independent applications, and stateful applications, particularly when rapid hardware upgrades are necessary to accommodate evolving demands. For instance, during promotional campaigns, resource demands frequently surge multiple times than usual. To address this, enterprises can swiftly upgrade their cloud resource configurations (for example, additional CPUs, increased memory, expanded bandwidth, and greater disk capacity) via the console or OpenAPI. After the promotion, they can scale down these resources back to the previous specifications to achieve cost efficiency.
• Horizontal scaling is applicable to distributed, stateless, and rapidly-changing applications when fixed data resources fail to handle the swift shifts in service requirements. In case of an unexpected surge in enterprise demand or promotional campaigns, you can swiftly scale resources through dynamic scaling policies. Moreover, you can fine-tune resource allocation to seamlessly accommodate business growth.

  The following scaling policies are supported:

  • Scheduled mode: Create a scheduled task to scale resources at a specified time.
  • Metric mode: Create a threshold for triggering an alarm based on resource performance metrics (such as CPU usage and average network traffic). After an alarm is triggered, resources are automatically scaled.
  • Fixed quantity mode: Set the minimum and maximum expected resource quantities. If the quantity of instances is less than the minimum or greater than the maximum, the system automatically scales resources.
  • Manual mode: Manually add, remove, or delete resources.