Azure Functions vs. Azure Virtual Machines

They're great to embed logic and code in a medium-small, cloud-native application, but they can become quite limiting for complex, enterprise applications.

Incentivized

It's well suited to delivering information about our sports events as during the events a lot of processing power is needed and instantly becomes available by scaling out when the event is over the service can be scaled right back making massive savings. We use it for football, horse racing, Olympics games etc, it is also used when things happen in the world like right now there is a lot of concern over the Russia and Ukraine conflict, since the demand for this information is high we instantly scale to meet the demand of our news feed services. I believe up to 90% of the UK's News, sports and media information actually passes through our computer systems, we are a market leading news and information service and Azure Virtual Machines provide us with the reliability that we need so that we can provide a rock solid reliable news and information service to the world.

Incentivized

• They natively integrate with many triggers from other Azure services, like Blob Storage or Event Grid, which is super handy when creating cloud-native applications on Azure (data wrangling pipelines, business process automation, data ingestion for IoT, ...)
• They natively support many common languages and frameworks, which makes them easily approachable by teams with a diverse background
• They are cheap solutions for low-usage or "seasonal" applications that exhibits a recurring usage/non-usage pattern (batch processing, montly reports, ...)

Incentivized

• You can login to Azure VMs using SSO with your Azure Ad account
• Azure VMs are securely accessible from anywhere in the world, with Azure Bastion
• You can execute scripts on the VM from the Azure portal without logging in to it

Incentivized

• My biggest complaint is that they promote a development model that tightly couples the infrastructure with the app logic. This can be fine in many scenarios, but it can take some time to build the right abstractions if you want to decouple you application from this deployment model. This is true at least using .NET functions.
• In some points, they "leak" their abstraction and - from what I understood - they're actually based on the App Service/Web App "WebJob SDK" infrastructure. This makes sense, since they also share some legacy behavior from their ancestor.
• For larger projects, their mixing of logic, code and infrastructure can become difficult to manage. In these situations, good App Services or brand new Container Apps could be a better fit.

Incentivized

• Pricing can be a concern if you are truly agnostic to which cloud you are building your particular solution in.
• The UI, as is the case with any cloud provider, is crowded.
• As with any cloud provider, it can be difficult to tune in exactly the right amount of servers for your needs...you might find yourself under/overprovisioning.

Incentivized

They are very easy and intuitive, very easy to deploy

Incentivized

I give the overall support for Azure Virtual Machines a 7 because I think while the overall support do a great job there are still areas that it could improve on such as efficiency and speed. So while I only give it a 7 and it has some issues it is still better than the overall support at Amazon EC2 Auto Scaling.

Incentivized

This is the most straightforward and easy-to-implement server less solution. App Service is great, but it's designed for websites, and it cannot scale automatically as easily as Azure Functions. Container Apps is a robust and scalable choice, but they need much more planning, development and general work to implement. Container Instances are the same as Container Apps, but they are extremely more limited in termos of capacity. Kubernetes Service si the classic pod container on Azure, but it requires highly skilled professional, and there are not many scenario where it should be used, especially in smaller teams.

Incentivized

Oracle Cloud Infrastructure's Compute supports Windows, but is really geared towards Linux. You can install Windows on some Linode servers, but again, Linode is geared towards Linux. AWS does everything but is complex and can have high costs. If you want to host Windows servers in the cloud, nothing beats Azure. From licensing to management, Microsoft Azure provides the easiest way to deploy and manage Windows Servers in the cloud, especially if you utilize other Microsoft services like Microsoft 365 an Visual Studio subscriptions.

Incentivized

• They allowed me to create solutions with low TCO for the customer, which loves the result and the low price, that helped me create solutions for more clients in less time.
• You can save up to 100% of your compute bill, if you stay under a certain tenant conditions.

Incentivized

• It's so easy to spin up new instances, that it becomes also to easy to have to many of them to manage. Many teams end up with a couple of hundreds of VMs after a short while, making the whole thing very hard to maneuver
• Azure VMs are the next step for us to rely on Onprem servers, and leaving the management of the infrastructure to the professionals
• The ease of use, is also important when our main focus is to deliver new applications and integrations fast, and not having to worry about infrastructure. We sell bottles, not CPUs

Incentivized