The cloud provider operates data centres with a vast collections of physical servers. These servers come in various sizes, each with different amount of CPU processing power, memory (RAM), storage, and other resources. Different servers are dedicated for different computational power requirements, such as certain tasks like machine learning and scientific computing demand specialised hardware like GPUs or TPU to accelerate complex calculation.
Virtual Machines (VMs)
The traditional way a business operates is by running applications on servers, there are web applications, databases, email services and more, each application is running on one server, this is because the operating systems didn’t have the capability to run multiple applications securely on a single server. To make efficient use of these of these physical servers, the cloud providers employ virtualisation technology. This allows the to create and run VMs on each physical server. VMs are similar to independent computers that share the computational resources of the physical server. A user who wants to use the could servers access a web page or interface provided by the cloud provider. The user can then configure the specific type of VM they need. They can select options such as the amount of CPU, RAM, and other resources they require. After the user has made their selection, the cloud provider’s system proceeds to provision a VM. This involves selecting an appropriate physical server from the collection of servers and setting up a virtual machine according to the user’s configurations.
User remote connection
Once the VM is provisioned, the customer can remotely connect to it. This connection is typically established through secure protocols like SSH (for Unix-based systems) or remote desktop. The user can then complete the configuration of the virtual machine, start the applications (e.g. a web server) and customise the environment according to their needs.
Manage VM remotely via API
A cloud API, or cloud Application Programming Interface, allows programmer configure, start, stop, etc the cloud VMs from a program. This means that VMs can be initiated in response to increased computational load for example.
Dynamic scalability and cost efficiency
One of the significant advantages of using a cloud API for VM management is the ability to dynamically scale resources to changing needs. For example, if an application experiences a sudden increase in traffic or load, the API can be used to automatically initiate additional VMs to handle the increased demand. Conversely, when the load decreases, unnecessary VMs can be stopped to save costs.
Back to parent node: Cloud Computing