Exam AZ-300 topic 5 question 18 discussion

The D is correct, because Scale In is also needed.

Since the company no longer have the installer or the source code, they can only lift and shift the onprem vm or physical server into Azure. The question also mentions that the infra was right sized which also means that there are at least 2 or more VMs/Physical Machines involved with the application. We also do not know if there was 2 or 5 or 10 or 20 machines handling the load. When creating the VMSS, we can specify our own VM images. (https://azure.microsoft.com/en-us/services/virtual-machine-scale-sets/#overview) When creating a VMSS, we can also specify a custom scale policy to scale up and down when necessary. Answer C mentions scaling up by 1. While this answer is suitable to a certain extent and there is no mention about saving costs, it doesn’t mention scaling down which means that resources are not optimised when load goes down. From an architect perspective, it would in the interest of the company to save on cost instead of leaving unnecessary VMs running. My answer preference is therefore ‘D’.

I think you need the source code to add AppInsight for D.... should be C

It can't be C because the scale out will be done on the average CPU usage of one instance not all instances. It means that as long the first VM is busy after the cooldown period it will increase one more and so on.

In order to use Azure monitor to autoscale you need to install Application Insights into the web app but you don't have de code then should be C?

Azure Monitor autoscale applies only to Virtual Machine Scale Sets, Cloud Services, App Service - Web Apps, Azure Data Explorer Cluster , Integration Service Environment and API Management services? Since we cannot create a VM scale set to an exisitng VM. Are we assuming the VM/Website is hosted to a WebAPP and going to use the Azure Monitor to enable Autoscaling? as per the link. https://docs.microsoft.com/en-us/azure/azure-monitor/platform/autoscale-custom-metric

People stating scale out is not needed, are too black or white. You don' t know which Size they chose for the VMSS, maybe it's a very weak VM which means scaling in and out is relevant... After all they moved from onprem so it's not like they are forced to use the same settings in Azure. C and D both work for me

For further knowledge: https://docs.microsoft.com/en-us/azure/azure-monitor/platform/autoscale-custom-metric

The ques is simple but demands well thoughts. Only scale out is not required. Since the usage is unknown Scale in also needs to be considered. Hence Option D as based on Metrics VMs will be set.

it could be C as well