Introduction to Image/Container anatomy

Container data flows and running state

The diagram below shows the anatomy of a container with flows of data into the container and how it transitions to the eventual running state.

Data Class	Default Location	Use	Description
VAULT		ext	Secret information from external Vault (i.e. HashiCorp Vault). Items like passwords, certificates, keys, etc...
ORCH		ext	Environment variables from secrets, configmaps and/or env/envfile resources from orchestration (i.e. docker, k8s.
SERVER PROFILE		ext	Product server profile from either an external repository (i.e. git) or external volume (i.e. aws s3).
SERVER BITS	/opt/server	ro	Uncompressed copy of the product software. Provided by image.
SECRETS	/run/secrets	ro	Read Only secrets residing on non-persistent storage (i.e. /run/secrets).
IN	/opt/in	ro	Volume intended to receive all incoming server-profile information.
ENV	/opt/staging/.env	mem	Environment variable settings used by hooks and product to configure container.
STAGING	/opt/staging	tmp	Temporary space used to prepare configuration and store variable settings before being moved to OUT
OUT	/opt/out	rw	Combo of product bits/configuration resulting in running container configuration.
PERSISTENT VOLUME		rw	Persistent location of product bits/configuration in external storage (i.e. AWS EBS)

Because of many factors of how an image is deployed, the options available and recommended for use of the elements in the previous table can vary greatly:

• Deployment Environment - Kubernetes, Cloud Vendor, Local Docker
• CI/CD Tools - Kubectl, Helm, Kustomize, Terraform
• Source Maintenance - Git, Cloud Vendor Volumes
• Customer Environment - Development, Test, QA, Stage, Prod
• Security - Test/QA/Production Data, Secrets, Certificates, Secret Management Tools

Examples might look like:

File flowchart example

The following diagram shows how files can enter and flow through the container:

Production Example

The following diagram shows an example in a high-level production scenario in an Amazon Web Services (AWS) EKS environment, where:

• HashiCorp Vault is used to provide secrets to the container.
• Helm is used to create k8s resources and deploy them.
• AWS EBS volumes is used to persist the state of the container.

Development Example

The following diagram shows an example in a high-level development scenario in an Azure AKS environment, where:

• No secrets management is used.
• Simple kubectl is used to deploy k8s resources.
• AWS EBS volumes is used to persist the state of the container.

Customizing the Containers

You can customize our product containers by:

• Customizing server profiles

  The server profiles supply configuration, data, and environment information to the product containers at startup. You can use our server profiles or use them as a baseline for creating your own.

  You can find these in Baseline server profiles in our pingidentity-server-profiles repository.

• Environment substitution

  You can deploy configurations in multiple environments with minimal changes by removing literal values and replacing them with environment variables.

• Customizing YAML files

  In the stack-related directories for the deployment examples, you can find the YAML files used to configure the Docker stack deployment. The YAML files can contain startup configuration settings or references to startup configuration settings, such as environment variables, for the stack.

  You can try different configuration settings using these YAML files or use them as a baseline for creating your own.

• Using DevOps hooks

  Hooks are DevOps shell scripts, generally specific to a product, that you can use to automate certain operations.

  You can find the hooks for our builds in the Docker builds product directories.

• Using release tags

  We use sets of tags for each released build image. These tags identify whether the image is a specific stable release, the latest stable release, or current (potentially unstable) builds. You can find the release tag information in Docker images.

  You can try different tags in either the standalone startup scripts for the deployment examples or the YAML files for the orchestrated deployment examples.

• Adding a message of the day (MOTD)

  You can use a motd.json file to add message of the day information that will be used by the DevOps images.