Three challenges for your cost management
When we talk to customers about their use of the public cloud, cost always comes up as a major pain point. As companies look to optimize their cloud costs, they often encounter the following challenges:
1. The change from a CapEx to an OpEx model
In a CapEx model, you make an initial investment for a fixed term. This investment is usually relatively large, but once the investment has been made, you know that you will be able to reap the benefits for the specified period without additional investment.
In the OpEx model of cloud platforms such as Microsoft Azure, you work with a pay-per-use model where you pay monthly, depending on your usage, instead of making any upfront investments. As your use can vary greatly, this makes costs much harder to predict, especially in the beginning, when you will have to rely mainly on guesswork. Even as time goes on, you will constantly have to keep an eye on the costs if you do not want to lose control.
2. Budgeting for projects
In addition, using cloud services make it more difficult to keep track of budgets for individual projects. Not only are these costs more unpredictable, but it is not always clear which resources in the cloud contribute to the cost of which project. This poses challenges in terms of visibility and control over project budgets.
3. Shadow IT
Before the emergence of the cloud, virtually all use of IT in an organization was via service requests to the IT department. As a result, the IT department retained an overview of costs. The low barrier to entry of cloud platforms and SaaS solutions means that anyone can now easily start using resources in the cloud without the IT department knowing about it. This so-called ‘shadow IT’ will of course also have an impact on costs.
The broader picture of shadow IT is that the cloud makes it more difficult to correctly pass on the charges for the use of resources by different internal departments. For accurate internal invoicing, you need insight into the usage of each department because of the pay-per-use model.
Regain control of your costs
Fortunately, the cloud not only presents you with these challenges, but also offers all kinds of tools to help you take back control of your costs. These include:
1. Reserved instances and policies
The fact that costs are more unpredictable due to the pay-per-use model can be partially addressed by analysing which uses you can predict very well. For example, if you know that you will need a specific virtual machine 24/7 for the next three years, reserve it. Azure offers reserved instances for this purpose: by reserving use, you receive a discount on the cost. However, for a test environment that you only need to set up every now and then, you are better off with the pay-per-use model. So take a critical look at all your workloads to see which might work better as reserved instances.
Another tool that Azure offers are policies. These can be used to impose restrictions on the use of all resources in the cloud. This prevents users from starting up virtual machines that cost too much, for example in a region where network traffic to your other workloads is too expensive. You can define who can do what, and with which resources in the cloud, with a high degree of granularity.
To keep the costs of specific projects in check, it is important that you practise, matching the capacity of the cloud resources (amount of RAM, CPU, storage, etc.) to the workloads you run on it. That way you won’t waste any capacity.
Capacity requirements are dynamic, but the flexibility of the cloud is the perfect way to deal with this. For example, if you know that your website gets more visitors between 6 and 8 p.m. every day, let that server automatically upscale at those times and then downscale again.
Microsoft Azure allows you to assign tags to servers and other resources. A tag consists of a key and a value, for example the key ‘environment’ with the value ‘production.’ You can assign tags to indicate the department responsible for the resource, the project for which the resource is used, or an internal cost centre.
If you consistently apply tagging to all resources you use in Azure, you can easily break down costs by project or department for internal billing. Any resources that have not been tagged are easy to identify and then fix. It is also advisable to set up a warning as soon as a project is approaching the limit on its budget.
Our approach: from assessment to targeted actions
Feeling out of control with your cloud costs? Thanks to more than 25 years of experience in IT infrastructure solutions and an in-depth knowledge of applications, Cegeka can help put you back in the driving seat. Our approach consists of the following three phases:
In the first step, we take a look at your cloud environment in Azure: how are you using it, what exactly do you need and what impact does this have on costs? In this way, we discover problems, look at areas for improvement and list options for saving costs.
Once we have determined all possible cost optimizations, we will discuss them with you and set priorities. We take a pragmatic approach and pay special attention to quick wins such as the removal of unused resources. And if the outcome of the assessment is that your cloud application would be cheaper if redesigned, we can help you with that too.
3. Targeted actions
On the basis of our discussions, we will draw up a detailed action plan setting out the changes needed to reduce your costs. In addition, we will also determine the actions needed to keep your cloud environment cost-efficient afterwards.
If you do not yet use the cloud and you’d like us to guide you through a cloud migration and handle your cloud environment thereafter, we will keep a constant eye on the costs, from beginning to end.
The ultimate goal is to reduce your cloud costs, freeing up a budget that you can spend on, for example, modernizing your applications. This will give your digital transformation a boost.