It’s been a couple years since my blog post about the hidden costs of cloud computing. Since then, major cloud providers have signaled their intentions to do something about the problem of high data egress fees, inspired by the new European Data Act and the possibility of similar regulations in other jurisdictions. It’s worth revisiting the topic now to see what’s changed and what hasn’t.
It’s true that some major cloud providers have announced they’re removing data egress fees, but as a recent report from Data Center Dynamics[1] points out, this may only apply in certain situations, such as when a customer is leaving a cloud provider altogether. Therefore, businesses running a multicloud environment may not benefit at all from the recent announcements.
In addition to cost structures that align with their business needs, enterprises are looking for greater diversity of providers, more deployment locations and options, and ultimately, greater flexibility in placing workloads closer to their customers. All these factors may help explain why many enterprises are investing in alternatives to the public cloud. IDC predicts that spending on shared (public) cloud infrastructure and dedicated (private) cloud infrastructure will grow at similar rates between 2023 and 2028: 14.8% CAGR and 12.6% CAGR, respectively.[2]
Let’s look at a few of the challenges businesses run into around cloud costs and how Equinix can help overcome them.
Placing the right workloads in the right locations
Enterprises want to deploy workloads where it makes the most sense for their business. For instance, they may want to:
- Store data adjacent to the cloud but not within the cloud
- Move workloads to their customers quickly and easily
- Spread their data across multiple providers to not have “all their eggs in one basket”
Data egress fees could make effective workload placement more difficult. Greater cost efficiency is one of the multicloud benefits that businesses could miss out on as a result.
There are workloads where it makes sense to pay a premium to work with a particular cloud provider, and others where it’s better to diversify the workload. Businesses facing high data egress fees may not get to make this choice for themselves. Instead, they may be stuck with a difficult decision: either pay the egress fees, or leave the data in place and potentially overpay for cloud services. No matter which option they choose, it could hurt their bottom line.
Deploying infrastructure in a cloud adjacent location can help solve this problem. When you put your data at the edge of the cloud, but not in cloud native storage, you can move copies of your data into cloud services on demand without losing control over the primary dataset. This means that when a particular cloud is no longer the most cost-effective option, you don’t have to keep using it.
Equinix is the ideal place to deploy cloud adjacent infrastructure, due to our global data center platform and market-leading number of cloud on-ramps, providing low-latency connectivity to multiple cloud providers in many locations. In addition, our portfolio of digital infrastructure services includes Equinix Metal®, our dedicated bare metal offering. This gives Equinix customers another valuable option for cost-efficient workload placement.
Bare metal servers are available via a fixed-cost model, making them a good choice for more predictable workloads. In contrast, pay-as-you-go public cloud services may be more cost-effective for workloads where utilization varies from month to month. At Equinix, you can run both bare metal and public cloud workloads as part of the same hybrid environment, and easily change placement as your needs change. Integration with Equinix Fabric®, our virtual networking solution, provides simplified, high-performance multicloud networking to enable this.
Planning and sizing your infrastructure
Many businesses were first attracted to the public cloud by the promise of infrastructure flexibility. They wanted the freedom to deploy the exact capacity they need—and pay only for the capacity they use.
Compared to traditional on-premises data centers, today’s digital infrastructure alternatives provide a much faster way to deploy applications, storage, networking and compute as needed, where needed. Gone are the days when you’d need to wait weeks or months to deploy and operate your services and applications in a data center. Now, you can deploy quickly and in proximity to an ecosystem of partners and suppliers.
However, you still need to be thoughtful about your infrastructure costs, ensure you have access and control over your data, and maintain the flexibility to expand or contract your infrastructure as your business needs change.
Equinix solutions help you get the best of both worlds: cloud-like flexibility with control that’s similar to an on-premises environment. For instance, you can take advantage of private cloud solutions from trusted vendors running on dedicated bare metal, including VMware Cloud Foundation on Equinix Metal and Nutanix Cloud Platform on Equinix Metal. These solutions work just as they would on-premises, while giving you the flexibility and control you need to manage your costs.
Detecting anomalies quickly with multicloud observability
You can’t optimize cloud costs if you don’t know what’s happening inside your cloud environments at all times. Cloud anomalies—behavioral outliers caused by malicious attacks or other unexpected events—often come with a hefty price tag.
They can interrupt your regular business processes, thus preventing you from capitalizing on revenue opportunities. The longer it takes to detect the anomaly, the more expensive it will be to remediate it, since it could spread into different areas of your cloud environment. This is why having an effective observability platform is so important for cost optimization.
However, running an observability platform in a multicloud environment isn’t easy. If you run cloud native observability tools, you’d end up with different tools for different clouds, each with their own unique capabilities. This would make it difficult to get observability that cuts across all your cloud silos.
A third-party observability platform can help solve this issue, but that still leaves the question of where to host that platform. To know what’s happening across all your clouds, you need to ensure data can flow back and forth to your observability platform with no interruptions. This means you need to host the platform on vendor-agnostic infrastructure that treats traffic from each cloud provider equally.
At Equinix, neutrality is in our DNA. It’s part of what makes Equinix digital infrastructure well suited for supporting multicloud observability. Other factors include our aforementioned cloud on-ramps and our Equinix Fabric customer portal. The self-service web portal provides visibility into your multicloud networking, showing where costs are coming from in your cloud environment. When you identify cloud services that are no longer worth the price, the Equinix Fabric portal makes it easy to remove them.
For all three of the use cases named above—workload placement, infrastructure sizing, and observability—multicloud networking is essential. It’s the glue that holds your hybrid environment together, allowing data and workloads to flow wherever they need to go across your cloud and on-premises environments.
To learn more about how to design your multicloud networking for cost efficiency and other benefits, read our guide: 7 Key Questions to Ask when Architecting a Multicloud Network.
[1] Georgia Butler, The slow burn of egress fees, Data Center Dynamics, May 3, 2024.
[2] IDC Press Release, Shared Cloud Infrastructure Spending Continues to Accelerate, Fueled by AI-Related Spending in the First Quarter of 2024, According to IDC, June 28, 2024.
Subscribe to the Equinix Blog