Skip to content


The diagram below represents the WebSight CMS high-level logical architecture.

WebSight - logical architecture

Content authors use CMS to create an experience by managing content, assets and pages (layout). Once the experience is ready to go public, authors push (in other words, publish) it to the backend experience storage.

On the frontend, the experience is delivered to the site visitors. They call the web server, which pulls static HTML files/assets/JavaScript/ CSS scripts from the backened experience storage and serves them immediately.


The experience storage has no access to CMS. That means that if, for example, there is no page in the storage, then site visitors see a 404 message.

The above description is called a push model and assumes that the pages are prepared for the user in advance. Note that within this architecture, it is possible to switch off the CMS part, which has no impact on delivering pages.

Our CMS is a Java-based OSGi application (WebSight CMS CE) with a NoSQL database (MongoDB). WebSight CMS supports content management and the generation of static HTML pages. MongoDB database stores content and assets.

WebSight - logical architecture with MongoDB


Multiple instances of WebSight CMS can connect to MongoDB, which supports online deployments and enables CMS scalability.


We use containers to ship WebSight CMS for deployment on multiple environments, from developers' local computers to public clouds. Read our "Why we decided to ship and develop the OSGi application in containers" article to understand the benefits of containers in our stack.

Once developers create a WebSight project using the Maven archetype, they produce the following Docker images:

  • An NGINX image (web server) with addtional project-specific configurations
  • A customized CMS image with core WebSight CMS and project-specific modules and configurations

Docker Compose

With more than one container in the platform, we need a tool for defining and running multi-container Docker applications. With Compose, we can use the Compose YAML file model to:

  • Configure all platform components (services)
  • Enable communication between them
  • Handle data persistance with volumes

The following diagram presents all WebSight CMS containers and volumes together.

WebSight - logical architecture

The diagram above reflects the containers' logical architecture. Docker Compose configuration specifies the following services:

  • The nginx service (experience delivery)
  • The cms service (content management and pages generation)
  • The mongo service (content database)

Services use volumes (external storage) to save durable data outside the container to separate the lifecycle of containers that use it from the data itself. When Docker Compose destroys a container (e.g., on crash or re-deploy), all data from its internal storage is lost. We specify the following volumes:

  • The experience storage volume is shared by NGINX (read) and WebSight CMS (write) containers
  • The content volume keeps content and assets
  • The logs volume keeps all WebSight CMS application logs


Working with containers enables consistency and standardization. With Docker, we can use the same tooling to work locally as well as to deploy production environments to the cloud.

For example:

When using Docker Compose, the same command is responsible for creating local and cloud instances (docker compose up).

Read more about Docker Compose Amazon Elastic Container Service integration.