Components

Data volume

To ensure data durability and architectural flexibility, all application file data in the V4 platform is allocated separately from the compute layer and stored on an independent, persistent cloud data volume.

Because this data volume exists independently of the core server instance, the compute layer can be destroyed, upgraded, or reconfigured without any risk to the underlying data.

In single-server configurations, this volume acts as the local storage for the application’s stateful files. As you scale to multi-server tiers (Standard, Pro, HA Pro), this same volume serves as a centralized file database, exported via an NFS shared filesystem to feed the horizontally autoscaling web nodes.

Snapshots of the data volume are performed at least daily and serve as a point of recovery against unintended software behaviors, human error, or malicious activity.

Core server

At the heart of the Webscale V4 platform is the core server, a highly optimized, multi-purpose environment configured to handle the entire application ecosystem out of the box. Its role adapts dynamically depending on the architectural tier in use, shifting from an all-in-one host to a specialized backend manager as a stack scales.

Essentials: The core server is the entire environment. It functions as a single-tenant node hosting every layer of your stack: serving the web application code, handling cron jobs, and running critical infrastructure components like MySQL/MariaDB, Valkey, RabbitMQ, and OpenSearch on a localized, high-performance filesystem.

Scaling Upward: As workloads grow and you transition to higher tiers, the core server is systematically offloaded to maximize efficiency and resilience:

• Standard: It hands off the web delivery layer to a dedicated, horizontally autoscaling web tier, retaining responsibility for the database, caching, queueing, and file layers.

• Pro and HA Pro: Critical infrastructure services (databases, caches, and search services) are migrated to cloud-native managed services. The core server still provides message queueing, data volume sharing, and central cron execution.

Regardless of tier, the dev command provides consistent access to a shell in an application container environment. Shared directories and files are configured in the shared stack attribute, and they are automatically mounted into the container at execution time.

Web servers

In scaling tiers (Standard and above), the web delivery layer is decoupled from backend services and managed as a dynamic, horizontally autoscaling cluster.

Application code is provided to a scaling cluster by creating a custom container image and configuring the cluster to run it. Shared directories and files are configured in the shared stack attribute, and they are automatically mounted into the container at execution time.

The configuration of the scaling cluster is managed at the Webscale control plane. CodeDirect is recommended to build and auto deploy new images to a stack.

Cloud native services

For the Pro and HA Pro tiers, the most critical services are offloaded to managed, cloud-native services. This provides dedicated resources, separate backups, and enhanced recoverability, while allowing backend components to scale independently and significantly reducing routine maintenance overhead.

Last modified on July 5, 2026