Shifting Priorities for Workload Placement Require a Hybrid IT Approach

In the digital economy, addressing the rapidly changing priorities of placing workloads and data close to users and applications is a top focus for IT leaders worldwide. Managing the diverse placement of workloads introduces the need for a hybrid approach to IT infrastructure. According to the 2022 Global Tech Trends Survey by Equinix, 38% of global IT decision-makers are deploying cloud services via a hybrid cloud model, making it the most common deployment approach.

Hybrid IT provides the flexibility enterprises need to place various workloads across on-premises data centers, colocation facilities, the cloud and the digital edge. Workloads include distributed data stores, containerized microservices, deep learning applications, analytics on big data, 3D modelling, real-time streaming data, caches, and distributed databases, in-memory databases, highly scalable NoSQL databases, and data warehouses. There are multiple factors and criteria that influence how enterprises develop strategies to optimize workload placement, which we’ll explore in this article.

Priorities drive workload placement strategy development

Some of the challenges associated with workload placement are driven by the speed at which businesses must respond to new market conditions. The rapid pace of change can lead to making ad hoc decisions on workload deployments without sufficient input from IT during the planning stages, limiting visibility into what drove decisions and how IT can help meet business expectations.

Taking time to identify key workload attributes and understand business needs and pain points will inform IT on the best way forward. Rather than deploying workloads in traditional locations, decision-making is linked to any combination of speed, flexibility, scalability, agility, security, compliance, customer experience, cost savings, data protection, and disaster recovery.

For example, workloads that require speed and flexibility can benefit from placement in a public cloud, while hybrid cloud may make more sense for workloads that are dependent on data protection and disaster recovery. SaaS and cloud-native businesses are starting to distribute mature workloads into cloud-adjacent markets or move functions closer to the edge to reduce processing costs and latency.

Data management is another factor when it comes to selecting locations for workload placement. Strict data regulations may require businesses to store data on-premises or in a colocation facility with on-ramps to applications running in the cloud. Doing so also eliminates data egress fees, one of the hidden costs of cloud.

When heavy transaction processing takes place at the edge, businesses must store their data in edge colocation facilities or clouds to ensure low latency. If multiple workloads share the same data source, it’s important to consider individual compliance, performance or visibility requirements; locating the workloads separately and setting up connectivity between the data and end users may be the best choice.

Ensuring access to partners, customers or service providers in a dynamic interconnected ecosystem may also be a factor in determining workload placement. It’s been said that your ecosystem is your infrastructure, given the prevalence of ecosystem partners that provide digital infrastructure solutions and digital services such as software-defined interconnection and virtual network function devices.

Flexible workload placement requires a hybrid approach

Developing a strategy that guides the selection of one or more deployment options will reflect the hybrid nature of optimal workload placement. It’s important to build in the flexibility to design the best-fit deployment options for various combinations of workload placement priorities.

For example, companies that plan on accessing data in the cloud can use Bare Metal as a Service for a near cloud backup solution or as a disaster recovery solution for infrastructure residing in a colocation facility. Another deployment option is to repatriate workloads from the cloud to dedicated infrastructure for cost control and performance. Setting up your network first will simplify repatriation and introduce an option for when your business needs to scale up.

Ongoing assessment of placement decisions and business outcomes leads to continuous improvement in decision-making and opportunities to pivot when factors that drove previous workload placement decisions change over time. IT leaders must consider both security and performance when moving data and workloads between cloud services and the end users and applications that require access.

What to expect from strategic workload placement

According to the Leaders’ Guide to Digital Infrastructure, when customers deploy workload placement to the right locations, they report:

• 30–40% in operational savings from eliminating cloud connectivity overhead and shifting fixed CAPEX to flexible OPEX models
• 80% performance improvement with purpose-built, workload-optimized HCI that provides significant performance advantages over generic virtualized infrastructure—at a reduced cost
• 5–10x reduced deployment time achieved through automated digital infrastructure provisioning
• 3-4x lower data egress costs, 20x data processing speeds and 10x faster backup and data transfer

Defining a workload placement strategy and putting distributed digital infrastructure and digital services into place can help companies get started achieving similar results.

Putting a workload placement strategy into motion

The on-demand digital infrastructure and digital services available on Platform Equinix® enable enterprises to use a hybrid approach to workload placement on physical and virtual deployment environments across networking, compute and storage. Customers can choose from hybrid multicloud platforms such as VMware and Nutanix, storage partners such as Pure Storage and Dell Technologies, virtual network function devices (such as load balancers, firewalls, routers, and SD-WAN devices) from leading brands, and Kubernetes tools for managing containerized applications. Delivery of these services is virtual rather than physical; this means infrastructure can be set up in minutes, not months.

As the world’s digital infrastructure company™, Equinix partners with more than 3,000 cloud and IT service providers and 2,000 network service providers. This makes it likely that the providers you plan to use for workload placement are already at Equinix. Software-defined interconnection capabilities from Equinix Fabric™ simplify hybrid multicloud networking for enterprises with on-demand virtual connections to cloud providers and other technology partners. Finally, Equinix Metal® dedicated Bare Metal as a Service capabilities allow companies to deploy single-tenant compute and storage capabilities in proximity to end users and data sources at the edge.

To learn more about how customers use on-demand digital services and infrastructure to unlock digital growth, read the Leaders’ Guide to Digital Infrastructure.

You may also want to read:

GARTNER® – Workload placement in hybrid IT — Making great decisions about what, where, when and why (equinix.com)[1]

[1] Henrique Cecci and David Cappuccio, “Workload Placement in Hybrid IT — Making Great Decisions About What, Where, When and Why,” Gartner, May 22, 2022.