Cloud Elasticity
Elasticity is the ability of the cloud to “stretch and squeeze” IT resources in order to meet the needs of changing workloads. An elastic cloud system would autonomously commission and decommission computational resources proportional to how much work is there and the nature of said work
The goal is to avoid over-provisioning (way too many resources for the job) and under-provisioning (not enough resources for the job), I compare this concept with the idea of control theory — dynamically regulate the system output and correct for error. Sometimes, positive feedback correction needs to be applied, sometimes, negative feedback correction
A highly elastic cloud system supports faster autonomic provisioning of resources to quickly meet changing demands. Faster responses
- Eliminates upfront costs and investments. You don’t have to spend hefty sums of money into infrastructure and worry about whether all of that is going to remain unused or none of that is still enough
- Saves time. Elastic cloud networks handle the provisioning of resources quickly and on-demand. Without that, manually provisioning whatever’s required can take a lot of time
- Elastic cloud networks are highly scalable because of their adaptability.
- Pay-as-you-go models are usually followed. This enables proportional resource allocation and utilization
Cloud resource balancing
Resource utilization should match resource allocation