What to do when there’s no silver lining to your public cloud experience

Australia, Jul 30, 2024

Like many Australian companies recently, your IT team may have shifted from on-premises infrastructure or hosted data centres across to public cloud to access the many benefits of hyperscale cloud computing. While this shift has helped to bring processes online and empower a hybrid workforce with computing power at scale, in some cases it has led to unexpected bill shock each month.

This is often the situation when the shift from on-premises to public cloud occurs rapidly and without an adequate roadmap or application modernisation strategy. In these cases, using traditional IaaS methodologies like virtual machines can greatly contribute to the increase in monthly cloud costs. In these cases, using traditional IaaS methodologies like virtual machines can greatly contribute to the increase in monthly cloud costs.

Hyperscale cloud computing is built different.

Hyperscale cloud computing allows organisations to access large-scale data processing and storage on demand, with the option to pay for what you use. It’s a scalable, secure, and flexible cloud computing solution, particularly for organisations that want to develop, trial and test AI, IoT, machine learning, and big data projects. The major hyperscale providers globally include Amazon Web Services, Google Cloud Platform, Microsoft Azure, IBM Cloud and Alibaba Cloud.

Most virtual machines designed for on-premises infrastructure are typically not optimal for public cloud environments. They require more support and can limit scalability, which is one of the critical benefits of hyperscale computing. By modernising using cloud-native applications over IaaS, teams can instead take advantage of leaner processes, cloud native functionality like AL or ML and lower their ongoing costs.

Progress sometimes takes two steps forward, one step back.

The pandemic forced many organisations in Australia to take hasty steps to move their applications into a hyperscaler cloud. The aim at the time, was to enable remote access to their workforce, with little consideration around cost, security or future state planning.

At Logicalis Australia, we’re now seeing customers come to us for help with repatriation to shift their virtual machines back to on-premises to achieve short-term financial gain. Reasons can include cost control, performance issues, security, or compliance issues such as data sovereignty, or to gain more customisation and control for applications.

However, they are finding this path is not as simple as anticipated. It’s important to weigh up the available options before making the move to ensure this is the best action to take.

Reflecting broader changes in modern business, many technology budgets have shifted from focusing on longer-term capital expenditure (CapEx) towards the shorter-term benefits of operating expenses (OpEx). Previously, CapEx budgeting was important for on-premises, and in switching back, some teams are finding their OpEx budgets are no longer aligned with the CapEx needs of maintaining on-premises infrastructure.

Being caught between budgeting styles and the choice of public cloud, on-premises infrastructure, or a mix of the two is a reality that many Australian organisations now face. Let’s look at some options to consider when facing a similar situation.

Modernise your environment.
As outlined above, modernising your systems to use cloud-native apps and features can help significantly lower your ongoing costs. Cutting out the additional processing required by virtual machines built for on-premises, this option leads to lower monthly costs over the long term. The downside is that this is an upfront investment and will take time to see any return.
2. Bring VMs back on-premises if your requirements are seasonal or predictable.
This option is not suitable for every type of organisation. If your computing requirements are stable and not likely to change, this shift can deliver immediate financial gains in the short term. In most cases, a hybrid approach can be the sweet spot. This involves assessing workloads and determining which “cloud” makes the most sense based on cost, functionality, capacity and application lifecycle. Based on the workload, some applications can even move from private cloud infrastructure to public cloud infrastructure dynamically based if you need a burst in capacity or storage. Having an updated operating plan and application roadmap is imperative.
3. Repatriation with an OpEx mindset.
While the early appeal of cloud computing offered many advantages, the reality of the experience can lead some organisations to consider repatriation back to on-premises or data centres. Many major vendors have incorporated an OpEx billing model for on-premises infrastructure to help organisations adjust. These include Dell Apex, and HPE GreenLake.
As always, getting the right advice about which strategy will deliver the most benefits at the right time for your organisation is essential. Our team at Logicalis Australia has the expertise and experience to advise you of the best options to suit your specific circumstances.

Topic

Cloud