When resources need to be scaled up in traditional hosting, Service Providers might need to make new investments in hardware in the on-premise data centre, which immediately includes additional costs for space, electricity, cooling, staff, management and monitoring, etc. Equally, when they need to scale down, they still have to pay a cost to maintain the hardware, even when it is sitting idle. This always entails a business risk if these costs cannot be shared between different end customers or if the extra hardware remains unused for most of the time.
The second scenario is scalability in the cloud, such as in Microsoft Azure. Scalability in Azure is a breeze compared to traditional hosting, as scalability is really ingrained in the environment, without the need for new investments. Therefore, it is a very useful functionality for service providers, both from a provisioning perspective – allowing scaling through programmatic controls rather than hardware configuration – and from a cost perspective. It allows them to minimise business risks and meet the flexibility and efficiency requirements of their end customers.
From a cost perspective, there are two main reasons why scalability in Azure has major advantages over the on-premise environment. First, scaling in Azure does not require any upfront investments in hardware unlike the on-premise environment. Second, because the cloud has a pay-per-use pricing model, users only pay for the resources they consume. There is no burden of ownership of unused resources, which significantly reduces the business risk of upfront investments. From a provisioning perspective, it is important to realise that scalability is a core functionality of the Azure cloud; in most cases only requiring programmatic controls or pre-configured VM scale sets to enable scaling.
Azure cloud offers two ways of scaling: scaling out or in, also known as horizontal scaling, and scaling up or down, also known as vertical scaling.
With Windows Virtual Desktop – the new Microsoft Azure service designed specifically for virtual desktop infrastructure and Remote Desktop Session Host management – Microsoft also introduced load balancing as a turn-key offering and a new way of scaling. First, there was breadth-first load balancing, where users were distributed horizontally from host 1 to host 2, 3 and then back to 1: distributing the load between virtual machines. Now, there is also depth-first load balancing, which means that host 1 will be filled first, before users are redirected to host 2.
Scalability and load balancing in Azure is a key enabler for Service Providers to respond to changing circumstances of their end customers, both foreseen and unforeseen, for remote work, growth, seasonal influences, new or discontinued projects or services and more. In addition to scalability, automation is a core functionality of Azure, without manual configuration, downtime, security risks as in the on-premise data centre. Simple automation rules and scripts make it easy to enable or disable scalability procedures, as required.
Scalability in Azure enables Service Providers to provide high availability and maximum performance to their end customers with the ability to control costs. If you want to know more about scalability in your data centre, contact me.
If you’re interested in joining one of our upcoming events or want to know more about the Azure cloud, check out the Insight portal for the latest updates on cloud and data centre transformation or contact your local Insight representative.