Savvy IT leaders know that every application and workload cannot or should not be moved to or remain in the public cloud. IDC said last year that as many as 50% of public cloud workloads could be repatriated back to private clouds or on-premises infrastructure.
An Enterprise Strategy Group survey of 1,200 organizations recently found that 77% of them have actually repatriated their mission-critical business workloads from public cloud to their data centers.
Despite the justified hype and hoopla around the cloud, on-premises data centers running a mix of applications and other workloads―most notably, business-critical workloads―are here to stay.
Recently examining this, IDC found that five reasons why organizations prefer to host workloads on a private versus public cloud are:
- Security concerns
- Regulatory or internal policy constraints
- Latency requirements with other apps that cannot be moved to public cloud
- Latency requirements with physical entities on premises
- Total cost of ownership
Last year, as if anticipating the new normal of a remote workforce, Evaluator Group found that hosting high numbers of knowledge- and power-worker VDI desktops on premises on converged infrastructure was more financially attractive than on public cloud.
Reimagining the roles of public cloud and data centers
Data centers are here to stay, but so is public cloud. Gartner projects that public cloud has been growing 17.3% annually since 2018 and is expected to reach $50 billion by the end of this year. The fastest growing cloud segment, infrastructure as a service (IaaS), is growing at 24%.
Because organizations are repatriating critical business workloads from public cloud to their data centers and using more public cloud, IT teams need to reimagine cloud as a hybrid of IaaS on public cloud and on-premises private cloud. In fact, a 2020 IDC study found that hybrid cloud is the new standard:
- 60% of businesses have a hybrid cloud and use the framework for deploying each application on its optimal location.
- 28% of businesses do not have a hybrid cloud in place but plan to implement one in the next year.
- Only 12% do not have a hybrid cloud in place and do not plan to implement one in the next year.
Top-reported uses for hybrid cloud are:
- Migrating workloads or applications between locations based on the optimal location for them
- Running parts of an application in one location and parts in another
- Running test/dev/staging and production in different locations
- Data and storage tiering for older or archive data
- Backup of applications and disaster recovery
- Meeting short-term bursting needs
For workload placement strategy, you need to evaluate each workload’s characteristics and match those requirements to your on-premises infrastructure versus public cloud. Key considerations are:
- Storage and compute heavy or light
- Performance and latency
- Amount of data and required data management services
- Backup, recovery or continuous processing
- Security, regulatory and internal governance requirements
Hybrid cloud operations
Public cloud, though not financially (or technically) attractive for all workloads, is certainly appealing because of its on-demand and pay-as-you-go model. Application developers and IT operations staff can request infrastructure as a service without having to configure and manage the underlying hardware. In other words, they just have to focus on the app.
The good news is that next-generation infrastructure offers the same operational and consumption experience in the data center as the public cloud―while satisfying certain security, compliance, latency and cost requirements that public cloud cannot meet.
A common operations experience and flexible consumption across private and public cloud simplifies IT operations for on- and off-premises workloads, making it a win-win. It enables IT to be more agile and shift from time-consuming infrastructure and complex multicloud management to transformative application projects and IT services.
Evaluating data center modernization solutions for hybrid cloud strategy
IT teams can accelerate hybrid cloud initiatives with the latest generation of CI and HCI systems, built to modernize data centers for supporting digitally transforming applications and for simplified hybrid cloud operations.
When evaluating solutions, look for these attributes:
- Full-stack, fully supported convergence: A cloud-like turnkey system with all the data center technologies integrated, supported and lifecycle sustained by the vendor: servers, storage, data protection, network fabric and VMware virtualization.
- Cloud-enabled infrastructure: Storage and data protection with built-in hybrid cloud tiering for backup, replication and recovery, archiving and disaster recovery in the public cloud.
- Automation for cloud-like operations experience: Automation for infrastructure-as-a-service that can deliver, expand and upgrade resources to support new and growing workloads as quickly as the business need them.
- An on-demand cloud-consumption financial model: Automated on-demand access to additional compute and storage capacity in the system, so you scale up and down and pay for resources only when you use them.
Despite the benefits of public cloud, on-premises infrastructure is not going away. In fact, the majority of organizations are looking to modernize their data centers for a more rational hybrid cloud strategy that offers workload placement options.
When it comes to building the right foundation for hybrid cloud, converged and hyperconverged infrastructure offers the strengths of a public cloud experience with the technical, governance and business advantages of an on-premises data center.
IT modernization can’t wait. Learn more about what modern converged solutions can do for you by reviewing the other articles and resources in this special TechTarget information hub and visiting Dell Technologies converged and hyperconverged infrastructure.