Elasticity is the ability of a cloud system to grow and shrink the amount of computing resources it uses automatically, in step with demand. When traffic rises, more capacity is provisioned; when it falls, capacity is released. The goal is that the available resources always closely match what is actually needed at that moment.
The NIST definition of cloud computing (NIST SP 800-145, September 2011) names “rapid elasticity” as one of the five essential characteristics that distinguish cloud computing. NIST describes capabilities that can be “elastically provisioned and released, in some cases automatically, to scale rapidly outward and inward commensurate with demand.” It adds that, to the consumer, the capabilities available for provisioning often appear unlimited and can be obtained in any quantity at any time.
Elasticity is what underlies the cloud’s pay-as-you-go economics, which NIST captures separately as “measured service”: because resources are added and removed on demand, customers can be billed only for what they consume rather than for a fixed, always-on fleet sized for peak load. This is a sharp break from traditional data centers, where capacity had to be bought in advance for the worst-case scenario and often sat idle.
Elasticity differs from simple scalability. Scalability is the capacity to handle more load by adding resources; elasticity is doing that automatically and reversibly, in both directions, as load changes in real time. Auto-scaling groups, serverless functions, and managed container platforms are all mechanisms that deliver elasticity in practice.