Deploy and manage Azure compute resources
Pick the compute service, then the size
VMs, VM scale sets, App Service, Container Instances, AKS, and Functions each map to a scenario. For VMs, know the size families (B burstable, D general purpose, E memory optimized, F compute optimized, N GPU) and the disk types: Standard HDD, Standard SSD, Premium SSD, Premium SSD v2, and Ultra Disk. The temporary disk is not persisted, and only the deallocated state stops compute billing.
Availability is a design decision
An availability set uses fault domains and update domains inside one datacenter and targets 99.95%. Availability zones spread instances across datacenters in a region and target 99.99%. Neither can be changed after a VM is created, so the choice is permanent. Scale sets add autoscale rules and, when created with zones, zone balancing.
Deployment and configuration
ARM templates and Bicep deploy declaratively. Incremental mode adds and updates; complete mode deletes resources in the resource group that are absent from the template, a classic exam trap. Run what-if before deploying. VM extensions such as Custom Script and Desired State Configuration handle post-provisioning configuration.
App Service specifics
The App Service plan is the billing and scale unit, and several apps can share it. Scale up changes the plan tier for more CPU and memory; scale out adds instances and requires a tier that supports it. Deployment slots let you stage a version and swap it into production.
Worked rule of thumb
To resize a VM, check whether the target size exists on the current hardware cluster. If not, deallocate the VM, resize, then start it. Expect questions where a resize fails only because the VM is still running.