Are businesses moving away from the cloud?
The cloud computing hype-train of the 2000s has come and gone and cloud computing is now ubiquitous within any application driven business in one way or another.
Cloud computing has revolutionised the way we deploy applications, provision resources, and improve availability and accessibility for the end-users. As with other significant inventions there are unanticipated consequences as the product settles into the mainstream. This is no different with the Cloud. Recent industry discussions reveal that the Cloud is not always the answer and that the use of on-premise computing is starting to find its way into more and more use cases.
Most businesses, large and small, are using one or more cloud services. These services offer feature-rich, relatively inexpensive productivity applications allowing their customers to focus on core business objectives. The exponential growth in the volume of cloud storage required by businesses has created challenges for hyperscalers, and as a result, running costs have increased, seeing rising costs passed down to their customers. This includes the likes of SaaS providers. This process has seen cloud become more expensive than previously promised, resulting in a bloated cost model whereby costly inputs are forcing costly outputs from the SaaS provider to the end customer.
The reality is that going to the cloud is not always cheaper. You may think that “if you’re only paying for what you use, if you use less, you’ll pay less”. However, realising these cost savings requires active management.
Most companies believe that if a server or router dies, a new one takes over and it is seamless to their applications and clients. This is not necessarily the case. If your environment’s configuration doesn’t explicitly address redundancy, you’ll run the risk that an event at a single cloud provider zone or region could affect your applications and data. Different cloud providers address this in different ways. For example, some provide regions and availability zones where you can deploy compute and storage instances in multiple zones for applications that require high availability. The underlying hardware will be located in multiple fault domains, providing diversity of location and systems. This redundancy, obviously, comes at a cost.
You may also believe that you don’t really need much of a technology team, since everything is taken care of by the cloud provider. The reality however is that without extensive investments in scripting and automation, cloud SaaS deployment and management frequently require similar amounts of effort to an on-premises environment. While some tasks like hardware support are now outsourced to the cloud provider, a combination of existing and new administrative tasks are still your company’s responsibility.
There is also often a lack of transparency and visibility into cloud spending. Many companies find it difficult to track their spending in the cloud due to cloud sprawl, an unmanageable number of cloud resources and services, or even confusing and complicated bills from providers. This leads to hidden or unnecessary costs and uninformed conversations when it comes to budgeting or other strategic decisions.
Forecasting is also often inaccurate which leads to over-provisioning. Organisations may fail to right size cloud resources such as compute instances and storage to the amount actually needed, leading to deploying and paying for cloud resources that sit unused.
Companies do not always benefit from the strengths of the cloud as their architecting for the cloud is done poorly. A lack of foresight can lead to designing applications for cloud platforms or re-architecting legacy applications for the cloud without a clear idea of the costs involved.
By moving applications to the cloud with little or no changes, organisations that choose the “lift and shift” approach migrate workloads from on premise to the cloud with as few changes as possible. Since these applications are not optimised for the cloud, spending goes into fixing performance issues and migration failures.
Cost management in the cloud requires thoughtful planning and continuous monitoring. Hiring talent with experience, optimising costs and using consumption-based cloud cost models can help companies better forecast and manage spend for their businesses. Additionally, organisations can invest in tools that bring transparency and predictability to cloud usage and costs. With the right talent and tools in place, companies can gain the visibility and knowledge into their cloud spending to make informed business decisions. Some important factors to consider when architecting for the cloud include design complexity, cost to implement and operational processes and skill sets to support the system.
Hiring talent that has experience tackling these challenges in the cloud, plus implementing the right technology in place to support them can help companies address cost optimization challenges proactively. Organisations need to plan carefully and consider their provider when moving to the cloud. This involves assessing the specific needs of the business and aligning them with those of the chosen service provider.
Digiata and our partners offer both cloud and on-premise installations for all the reasons above. We understand that there is always a discussion to be had before deciding which way to go.