Processing at the edge or when disconnected; empower Dev Ops in your own datacentre; unlock Azure without compliance roadblocks.
In my previous posts I have provided an overview of Azure Stack and discussed the two commercial models available for using Azure Stack. Azure Stack use cases within Australia can be grouped into two broad categories, those where reliable connectivity to public Azure is both available and permitted and those where it is not. As discussed in my last post this also has a bearing on the commercial model you will use.
In Australia, we have Azure data centres in country – and a new one on the way – so data sovereignty is less of a concern. But that does not mean that Azure is suitable for all organisations. Azure Stack may be the answer for you if you are looking at building modern applications, need to overcome the tyrannies of distance or if you simply need to control where the data lives. In this post, I will explore some of these connected use cases and I’ll cover off some of the disconnected ones in the next post.
Modern applications on premises
For organisations with in-house software development teams there are a couple of ways they could use Azure Stack. For example an organisation developing consumer facing applications may need the kind of hyperscale that public cloud can offer for production, but running dev and test on Azure Stack could significantly reduce the operational cost of these environments.
Azure Stack may also be of use if internal development teams have a strategy to move line of business applications to more modern architectures comprised of microservices and containers, but for the time being they need to integrate with legacy systems that just are not able to go to the cloud. Many of these legacy systems simply need low latency connectivity so running the modern application physically near the legacy components is an interim step. There is no reason such a hybrid application could not still leverage public cloud for functionality that is not available in Azure Stack, batch processing or hyperscale. This flexibility makes hybrid applications exceptionally powerful in bridging legacy and cloud workloads.
Distance and Isolation
Let’s face it – Australia is big. Really big. Both the Azure data centres are located on the east coast. The latency across the country is going to be around 45ms if you are going to a main centre like Perth. There are whole industries in Australia where a significant portion of activity occurs well away from main centres. Mining, oil & gas, agriculture, forestry and tourism to name just a few. Many sites in these industries are off the beaten path and leverage satellite connectivity. While this can offer high bandwidth it also includes high latency.
Deploying modern applications on Azure Stack in these scenarios keeps the execution of the applications closer to the users. This will offer a better user experience and can also mitigate the risk of intermittent connectivity challenges across country.
Controlling where your data lives
I’ve talked to a number of our customers that are looking at Azure Stack because they want to build and run modern applications, but they can’t use public cloud. In one case, this was because the customer operates in a highly regulated industry and customer data must stay internal to the organisation. Another customer I spoke to was similar, except they are an IP based company and this requirement was dictated by their own security policy. A third customer simply had a mandate from their parent company – thou shalt not use public cloud. For all three customers, Azure Stack will allow them to meet their business objectives and maintain compliance.
In all three of these scenarios allowing Azure Stack to connect to the internet for maintenance and to report usage metrics is allowed. In some cases organisations won’t be able to have any connectivity to the internet – I’ll explore some of those use cases in my next post.