Following on from covering the basics around security models within Microsoft Azure yesterday today we are going to look into the various compute services available to us in Azure.
Most likely the starting point for anyone in the public cloud.
- Provides a VM from a variety of series and sizes with different capabilities (Sometimes an overwhelming) [Sizes for Virtual machines in Azure](https://docs.microsoft.com/en-us/azure/virtual-machines/sizes)
- We also have a burstable VM type which can be found under the B-Series. This is great for workloads where you can have a low CPU requirement for the most part but require that maybe once a month performance spike requirement.
- There are also Azure-tuned kernels when it comes to specific Linux distributions. [Azure Tuned Kernals](https://docs.microsoft.com/en-us/azure/virtual-machines/linux/endorsed-distros#azure-tuned-kernels)
There is a large selection of templates that can export deployed resource definitions. I like to think about this templating feature to something like AWS CloudFormation or could be Terraform for a multi-cloud option. We will cover Terraform more in the Infrastructure as code section.
In Azure, we have something called Virtual Machine Scale Sets (VMSS) for IaaS. This enables the automatic creation and scale from a gold standard image based on schedules and metrics.
We have not covered containers as a use case and what and how they can and should be needed in our DevOps learning journey but we need to mention that Azure has some specific container-focused services to mention.
Azure Kubernetes Service (AKS) - Provides a managed Kubernetes solution, no need to worry about the control plane or management of the underpinning cluster management. More on Kubernetes also later on.
Azure Container Instances - Containers as a service with Per-Second Billing. Run an image and integrate it with your virtual network, no need for Container Orchestration.
Azure also has the Container Registry which provides a private registry for Docker Images, Helm charts, OCI Artifacts and images. More on this again when we reach the containers section.
We should also mention that a lot of the container services may indeed also leverage containers under the hood but this is abstracted away from your requirement to manage.
The goal with serverless is that we only pay for the runtime of the function and do not have to have running virtual machines or PaaS applications running all the time. We simply run our function when we need it and then it goes away.
Azure Functions - Provides serverless code. If we remember back to our first look into the public cloud we will remember the abstraction layer of management, with serverless functions you are only going to be managing the code.
