Today, a wide variety of applications are available as SaaS applications. These are by definition the fairly vanilla ones for which there is a large demand, like Office 365. But many businesses believe that they still require a more traditional network with servers and storage to run the remainder of their systems that aren’t easily replaced with SaaS apps. The good news though is that you can also migrate some or all of these to the cloud as well.
The global hyperscale cloud providers like Microsoft Azure, Amazon Web Services and Google Cloud Platform allow you to run both Microsoft and open source based application stacks. This allows you to design and deploy exactly the same set up you have on premise in the cloud.
Why would you move a workload that is currently hosted on premise into the cloud?
- Opex rather than capex
- Pay for only what you use
- No contractual commitment
- Scale up / down or out / in to meet demand dynamically
- Deploy solutions anywhere on the globe to be near your customers or your business units
- Access to a huge range of capabilities that you can utilise and deploy instantly
- Geo-replication and resilience
So here are three reasons why you should consider deploying new workloads or moving existing workloads to the cloud:
- You can choose which data centre you deploy your workloads and data to (including UK data centres), and you can encrypt them with your own key so that the cloud provider can’t access any of your data. The hyperscale providers have invested tens of billions of dollars in building and securing these facilities and services which are now used by governments and large corporations around the globe. So the data sovereignty, privacy and security requirements have been very well addressed, and you would have to have a very specific and stringent requirement to prevent you moving to the cloud for one of these reasons.
- The Capex costs of owning and operating your own servers are: purchase of the server hardware, purchase of the server software (e.g. Windows Server 2012, SQL Server 2008), purchase of the backup device. The Opex costs of owning and operating your own servers are: electricity to power and cool, hardware maintenance contract, software maintenance contract, floor space, ongoing server management (either in-house time or a cost with a support company). If you move the same server into the cloud, then all the above costs are replaced with one Opex cost, without any termination fees and you only pay for what you use.
- Another consideration is the sizing and usage of resources in the cloud – you can size (and thus pay) for only what you need at that time as you can easily (and automatically if you wish) scale up and down. You can also turn servers off when you’re not using them (and automate this if you wish) and thus stop paying for them. Most businesses adopt a hybrid approach where they move some workloads into the cloud, and keep some on premise and connect the two networks together to provide a seamless experience to users.
If you have in-house expertise in deploying and managing infrastructure and applications on premise, then you already have almost all the skills you need to do the same in the cloud. The only real difference is that you deploy solutions through a web portal rather than ordering servers and software, waiting for them to arrive and then building them.
So, in summary, you can deploy pretty much any existing or new workload into the cloud with one of the hyperscale providers. You have nothing to lose by at least investigating this option! In the next few blog posts we will address some specific workloads in more detail.
If you want to know more about how to move your workloads to the cloud, you can attend our “Running your Production Systems in Azure” workshop that we are running together with Microsoft at their Campus in Reading on Tuesday morning 25 April at 10am. To register and to find out more details of the event click here.
f Back to Blog posts