Choosing an approach

Overview

Delphix offers three approaches to Continuous Data delivery, each optimized for a different infrastructure model and operational context. All three approaches are managed through Data Control Tower (DCT), which serves as the central control plane for dataset operations, access control, and governance regardless of the underlying infrastructure. This page describes each approach and the tradeoffs between them to help you determine which best fits your environment and team.

These three approaches are:

  • Continuous Data (CD): The foundational Delphix virtualization solution. Delphix provides virtualized data; the administrator is responsible for managing target infrastructure and installing database binaries. All dataset operations are managed through DCT.

  • Continuous Data for PaaS (CD4P): An extension of Continuous Data that orchestrates cloud-native PaaS databases using native cloud provider APIs. The cloud provider manages infrastructure and database binaries; Delphix manages the data lifecycle through DCT.

  • Continuous Data for Kubernetes (CD4K8s): A complete database delivery solution combining Delphix virtualization with Kubernetes infrastructure and containerized database images. Delphix provides the data, Kubernetes provides the infrastructure, and Docker images supply the database binaries — delivering a fully managed, on-premises PaaS-like experience. All dataset operations flow through DCT, with expanded UI and API coverage planned for a future release.

CD4P and CD4K8s are extensions of Continuous Data, not replacements. Organizations running traditional CD workloads can adopt either extension alongside their existing deployment.
  Continuous Data (CD) Continuous Data for PaaS (CD4P) Continuous Data for Kubernetes (CD4K8s)
What Delphix provides Virtualized data only. Orchestrated PaaS databases (data + cloud orchestration). Data (Delphix), binaries (Docker image), infrastructure (Kubernetes).
Infrastructure model Dedicated on-premises or cloud infrastructure (engines, environments). Cloud-managed PaaS instances (AWS RDS, Azure SQL MI, and others). Kubernetes cluster (nodes, storage, networking).
Data refresh speed Fast (virtualization snapshots). Slower (cloud backup/restore APIs); ~1 TB/4 hr for snapshots; provisioning from snapshot typically takes minutes. Fast (virtualization snapshots).
Storage efficiency High (deduplication, compression, thin clones). Low (full copies; cloud storage costs apply). High (deduplication, compression, thin clones).
Cost profile Infrastructure investment upfront; data storage efficient. Pay-as-you-go cloud; higher per-GB costs. Infrastructure investment (K8s cluster); storage efficient.
Toolking and skill requirements Database and systems administration expertise. Cloud platform knowledge (AWS, Azure, GCP). Kubernetes, Helm, DevOps, and container knowledge.
DCT UI/API coverage Full — all operations available via DCT. Full — all operations available via DCT. Partial today (Helm/operators); full DCT coverage planned.
Self-service capability High — application teams operate via DCT. High — application teams operate via DCT. Medium — requires Kubernetes access and knowledge.
Best for... Mixed on-premises and legacy workloads; high-volume data; teams with existing DB infrastructure. Cloud-first organizations; teams seeking minimal operations overhead; managed PaaS environments. DevOps-mature teams; containerized and microservices architectures; teams wanting a complete, managed database experience.

When to use each approach

Continuous Data (CD)

CD is best suited for organizations with existing on-premises or cloud infrastructure that host dedicated Delphix engines and target environments. It provides the highest degree of control over the virtualization stack and supports the broadest range of database types and configurations. Teams with strong database and systems administration skills will be most effective with this model.

Continuous Data for PaaS (CD4P)

CD4P is best suited for organizations that have standardized on cloud-managed PaaS database services and want to reduce the operational burden of managing infrastructure. Because the cloud provider manages the underlying database engine and compute, teams with cloud platform expertise can manage the full data lifecycle through DCT without requiring deep database administration knowledge. CD4P is the right choice when storage efficiency is less of a priority than operational simplicity.

Continuous Data for Kubernetes (CD4K8s)

CD4K8s is best suited for DevOps-mature teams running containerized applications on Kubernetes who want a complete, self-contained database environment — data, engine, and infrastructure — delivered as a single package. Because Kubernetes and Docker images supply the infrastructure and database binaries, teams do not need to provision dedicated target environments or manage database installations separately. CD4K8s is the right choice when teams are already operating within a Kubernetes-native toolchain and want to extend Delphix data virtualization into that workflow.

CD4K8s operations are managed through DCT. However, full DCT UI and API coverage is planned for a future release — some operations today require Helm charts or the Kubernetes operator model. For current integration details, refer to the Delphix Ecosystem Hub documentation.

Operational model summary

The three approaches differ primarily in where responsibility sits for each layer of the database stack:

Layer CD CD4P CD4K8s
Data Delphix Delphix Delphix
Database binaries Administrator Cloud provider Docker image
Infrastructure Administrator Cloud provider Kubernetes
Lifecycle management DCT (full) DCT (full) DCT + Helm (partial today)