20th June 2016
According to ITPro, the most common reason why a business will consider migrating their data and infrastructure to a cloud solution is the cost savings that are available.
Traditional data centres can be costly to run. Aside from the capital costs of buying and replacing equipment, there are also costs for the skilled staff required to manage the infrastructure, and the power and communication costs for simply keeping the lights on and data moving from A-to-B.
In theory, using a public cloud service like Microsoft Azure saves on most of these costs. A business no longer needs to buy and maintain their own servers, and the resource costs like technical support, and power are handled by the provider. Using a public cloud service also strips out the cost of backups and disaster recovery. Microsoft Azure operates across multiple data centres, and information can be backed up in near real time, meaning that the risk of outage and data loss is significantly reduced.
Despite this, many organisations find that moving to the cloud doesn’t always provide the cost savings they expected, and in some cases, the way services are charged for means that a monthly Azure subscription can get out of control.
There are a number of reasons why the cost of the cloud solution can increase to higher levels. Some of these are simply due to increasing need over time – a business may need additional VMs as new staff come on board, or require additional storage space to meet their demands as the amount of information they produce and process increases. There are also some fundamental issues that can arise from the initial move to the cloud.
Typically, when a business makes the decision to migrate to the cloud they build a new infrastructure using Virtual Machines that replicates what they had before. Azure makes it really simple to provision VMs at different specifications – the number of cores, processor speed, memory and storage.
A lack of analysis of usage and storage requirements during the environment planning phase can have big implications for the long-term cost. Virtual infrastructure that is under or over the required specification will result in cost inflation.
If you create a Microsoft Azure subscription direct with Microsoft, the specification for an entry level VM is as follows:
This configuration (A0) is intended as a very low spec machine for testing simple applications. The amount of memory and storage is too small to enable serious use, and while it might be plausible to use it for limited processing, the limitations of the machine rule it out for practical purposes.
A more typical specification for a VM would be as follows:
The D3 VM outlined above costs 34x more than the entry-level VM but is more capable. If a business has initially under specified their cloud environment and needs to upgrade to more powerful machines in order to get the functionality their staff need, there will be a significant jump in price.
Proper planning and analysis of the requirements as part of the architecting process mean that there won’t be any nasty surprises around the corner.
No-one wants to pay for something that isn’t being used, however quite often with the cloud, there is a need to do just that. Traditionally, you would create a server farm which is powerful enough to handle peak levels of demand. This ensures that user experience – whether for hosting or application use – remains acceptable at all times. However, it also means that a business can end up with a lot of redundant capacity in their infrastructure.
When we started working with Microsoft Azure in 2010 as a hosting platform for SharePoint, we quickly found that the subscription levels meant that costs for our clients could rise quite quickly. SharePoint requires multiple VMs to handle different elements of the platform including document storage, application and DB hosting, and also Search Indexing. Separate VMs are also needed to handle the Active Directory and user management.
For most businesses, the level of resource required to operate the environment changes significantly over time. At peak times, the VMs require more power and memory to handle user requests and data management, while at times of low usage, there will be too much capacity.
Cloud Control automates the resources in an environment, adding more machines and power when needed, and scaling them back when they are not. This means that our Azure clients get true Pay as You Go cloud infrastructure and are billed on actual consumption.
Cloud Control also handles a lot of additional management tasks such as backups, Active Directory, and OS level updates and patches. On average, our clients save more than 30% on the cost of going direct.
In the longer term, cloud costs will still tend to rise – as you add more storage or additional applications, you will still need to accommodate them – but Cloud Control can help you reduce the cost of your subscription almost immediately.
Why not call us today on 0207 099 0632 to talk to a member of our technical team about how Cloud Control works and how it can help to save your business money on Microsoft Azure.Contact us
Call now on 0203 697 0302 to speak to a member of our team