Technical Product Information - ADONIS 19

Overview​

Starting with ADONIS 19, the product is delivered to on-premise customers in a fully containerised form. This marks an important step toward improved scalability, portability and operational flexibility. By adopting a modern container-based delivery model, ADONIS can be deployed more efficiently in current and future IT environments while preserving the familiar capabilities of earlier versions.

Architecture​

With ADONIS 19, the application is distributed as a set of OCI-compliant container images. This approach maintains the functional architecture of previous ADONIS versions while enabling a flexible and modern deployment model.

The main application components are executed as two separate containers:

• web server container

• application server container

The corresponding images are made available through an OCI-compliant registry.

ADONIS supports two deployment variants:

• Kubernetes deployment, using the supplied Helm chart

• Docker Compose deployment, using the supplied Compose configuration

Both deployment options orchestrate the same two containerised application components. Kubernetes offers advanced orchestration capabilities with specialized resource types and integration features and in general enterprise-grade cluster architecture, while Docker Compose provides a simpler setup for environments that do not require full Kubernetes capabilities.

The ADONIS database is typically operated outside the Kubernetes cluster or Docker environment. Customers continue to manage their existing database server independently.

Kubernetes Deployment​

The following diagram illustrates how ADONIS 19 is deployed in a Kubernetes environment. The application components run in a dedicated namespace, while supporting infrastructure components are typically provided in other namespaces.

Prerequisites on Kubernetes​

The Kubernetes cluster in which ADONIS will run must provide:

• An Ingress controller or Gateway API controller for routing end-user traffic

• A StorageClass for provisioning persistent volumes

  • allowVolumeExpansion=true is highly recommended

It is recommended that the cluster also provides:

• cert-manager for automating TLS certificate handling for ingress or HTTPRoute

• ClamAV, exposing a TCP endpoint for malware scanning during file upload/download

Architectural Details​

ADONIS Namespace​

The ADONIS application components run in their own namespace and consist of:

• HttpRoute / Ingress, which acts as the entry point for external traffic into the ADONIS namespace.

• web server tier, where the web server is exposed through a Kubernetes service and uses a persistent volume for storing local logs.

• application server tier, where the application server runs as a separate deployment with its own service and persistent storage for the full-text search index and local logs.

Other Namespaces​

Cluster-level components that support the ADONIS deployment:

• Ingress controller / gateway API controller for routing external traffic

• StorageClass used to provision persistent volumes

• cert-manager (optional) for TLS certificate automation

• ClamAV (optional) for malware scanning

Deployment Behaviour​

• Both the web server and application server deployments support vertical scaling through memory and CPU requests/limits.

• From Kubernetes 1.35 onwards, requests and limits can be adjusted at pod level without recreating the Pod.

• Each deployment uses a single replica; ADONIS 19 does not support horizontal scaling with multiple replicas.

• Local logs are required to enable creation of a Support Information Package (SIP) within the application, which is necessary for technical support.

External Database​

ADONIS uses an external PostgreSQL or Microsoft SQL Server database, which runs outside the Kubernetes cluster and is managed independently.

Docker Compose Deployment​

The following diagram illustrates how ADONIS 19 can be deployed on a single host using Docker Compose-compatible tooling such as Docker or Podman.

In this setup, the ADONIS web server and application server containers run on the same machine. Persistent data is stored using Docker volumes, and external access is typically provided through a reverse proxy.

Architectural Details​

ADONIS Docker Compose Stack​

The Compose stack consists of:

• reverse proxy (e.g. nginx), which handles incoming HTTP(S) requests and forwards them to the ADONIS web server

• web server container

• application server container

• volumes, used to store local logs and the full-text search index

Deployment Behaviour​

• Both the web server and application server run on a single host.

• Container volumes ensure persistence for logs and search index data.

• Scaling is limited to a single-node setup; multi-replica or distributed deployment is not supported.

• Local logs stored in volumes enable the creation of a Support Information Package (SIP) for hotline support.

External Database​

ADONIS uses an external PostgreSQL or Microsoft SQL Server database, which runs outside the Kubernetes cluster and is managed independently.

Registry and Updates​

ADONIS 19 is distributed in a container-based format. The following release artefacts are provided.

Container Images​

The ADONIS application is distributed as a set of container images, including:

• web server image – provides the web interface and handles HTTP(S) requests

• application server image – executes the ADONIS application logic and backend operations.

Helm Charts​

For Kubernetes-based deployments, ADONIS provides a Helm chart for each released ADONIS version. This chart defines the Kubernetes resources and configuration parameters required to run ADONIS components in a containerised environment.

The delivery includes an umbrella chart that defines the full ADONIS deployment.

The umbrella chart contains a central values.yaml file, which specifies:

• container image references

• resource settings

• service configuration

• environment-specific parameters

These values can be customised to suit the target environment.

The Helm chart contains only deployment configuration, such as:

• container image definitions

• Kubernetes resource specifications

• configuration parameters

The Helm chart does not contain or preload application data.

Application data is stored persistently in the configured database and is therefore not affected by deploying, upgrading, or redeploying the Helm chart. This ensures that deployments remain stateless at the application layer, while persistent data is managed independently via the database and configured persistent storage.

Artifact Distribution​

Container Registry​

All container images and Helm charts are distributed through an official, BOC-managed OCI-compliant registry.

• Customers receive an email invitation to onboard to the registry.

• Registry access is secured using authenticated, role-based access control.

• Once onboarded, customers can manage their own access credentials (e.g. access keys or tokens).

These credentials can be used to securely authenticate and pull the required ADONIS artifacts from the registry via:

• container runtimes (e.g. Docker, containerd, Podman)

• Kubernetes environments

• CI/CD pipelines or deployment automation tools

Air-Gapped Environments​

Container images may be mirrored into internal registries to support restricted or air-gapped environments.

Examples of internal registry solutions include:

• JFrog Artifactory

• Harbor

• Nexus Repository

Updates​

Updates for container images and Helm charts follow the standard ADONIS release and maintenance process. New versions of the images are published to the OCI container registry as part of official product releases, maintenance updates, or security fixes.

Customers are notified about new releases through the usual communication channels (such as release notes and customer information). Updated images can then be pulled from the registry and deployed according to the customer’s internal deployment and update procedures.

Customers retain full control over when updates are applied in their environments.

System Requirements​

Client Requirements​

The software requirements for the ADONIS web client remain unchanged compared to previous ADONIS versions.

Supported Browsers​

Using the latest browser versions is recommended to ensure full functionality and security.

Server Requirements​

Container Runtime​

ADONIS 19 supports OCI-compliant container runtimes on AMD64 (x86_64) architecture. The following platforms are supported or expected to work:

The BOC Group operates ADONIS in its own SaaS environment using Kubernetes, and uses Docker and Docker Compose for internal development, testing, and validation. These environments are therefore fully validated.

Platforms classified as "expected to be compatible" are not individually tested or validated but are expected to function correctly if they conform to the respective Kubernetes and OCI standards.

Operating System​

ADONIS 19 can run on modern Linux distributions that support Kubernetes or Docker/containerd, such as:

• Ubuntu

• Debian

• Red Hat Enterprise Linux (RHEL)

• SUSE Linux Enterprise Server (SLES)

Any modern Linux distribution capable of running the chosen container runtime is expected to be compatible.

Kubernetes and Deployment Tooling​

ADONIS 19 deployments are typically performed using Kubernetes and Helm.

Database​

ADONIS requires an external database to store all persistent repository data. Database requirements remain unchanged compared to the previous ADONIS version. Database requirements are documented in the ADONIS 19 Hardware/Software Requirements.

Persistent Storage​

Persistent storage is required for the full-text search index and for application log files. Storage is typically provided using Kubernetes Persistent Volumes (PV) and Persistent Volume Claims (PVC), configured via the delivered Helm chart.

Networking​

Container images are distributed via an OCI-compliant registry and accessed over HTTPS.

Typical networking requirements include:

ADONIS is typically exposed via a Kubernetes Ingress controller, Gateway API controller or an external reverse proxy.

The ADONIS web server container exposes port 8080 and does not terminate HTTPS connections itself.

For encrypted access (HTTPS), TLS termination must therefore be performed by the Ingress/gateway component in Kubernetes or by a reverse proxy in Docker Compose deployments.

Hardware Requirements​

General Principles​

Hardware requirements for ADONIS 19 running in containers are broadly comparable to previous Windows-based deployments. The ADONIS 19 Hardware/Software Requirements provide general hardware sizing recommendations.

Typical Component Resources​

The following values provide typical CPU and memory requirements for each ADONIS container.

Actual resource usage depends on factors such as:

• number of concurrent users

• size and complexity of the repository

• modelling and reporting activity

• application workload patterns

Example Resource Configuration​

The values below show how typical ADONIS sizing can be mapped to Kubernetes resource settings. They are based on the recommended values above and should be adjusted according to real workload.

Web Server Container​

Application Server Container​

Deployment Configuration and Setup​

Configuration Approach​

Configuration of ADONIS 19 follows two complementary mechanisms:

Settings and Configuration Stored Within the Application​

These settings are managed either individually by users or centrally through the ADONIS Administration. They are stored in the ADONIS database.

Examples include:

• authentication configuration

• rights administration

• REST API settings

• application-level preferences

Settings and Configuration Read From the Environment​

These settings are read by the application from the runtime environment.

In earlier ADONIS versions, such settings were stored in .properties or .conf files inside the application server (e.g. server.conf) or web application (e.g. adoxx_web.properties) directories.

With ADONIS 19, the application is distributed as container images for the application server and the web server. Because containers are ephemeral, file-based configuration is no longer supported.

Instead, all deployment-specific configuration is now injected via environment variables:

• Kubernetes: variables are defined in the values.yaml file and passed to containers when the Helm chart is deployed.

• Docker: variables are defined in docker-compose.yml and applied when the containers start.

Configuration of the Application Server​

Configuration properties for the application server are provided as environment variables via either:

• the Helm values.yaml file (Kubernetes deployments), or

• the docker-compose.yml file (Docker deployments)

All application server variables are prefixed with ADOXX_.

The following examples show how to configure the database name and the application worker (aworker) ports.

Kubernetes:

aserver:
  env:
    - name: ADOXX_SERVER_DBNAME
      value: adodb
    - name: ADOXX_SERVER_PORTS
      value: 5432,5432

Docker:

aserver:
  image: adonis.image/aserver
  container_name: aserver
  ...
  environment:
    - ADOXX_SERVER_DBNAME=adodb
    - ADOXX_SERVER_PORTS=54321,54322

Configuration of the Web Server​

Configuration properties for the web server are also provided as environment variables via the values.yaml (Kubernetes) or docker-compose.yml (Docker) files.

Web server variables are prefixed with AXW_PROPERTIES_.

The following examples show how the used application server and aworker processes are configured and how the maximum size for REST log files is increased to 300MB.

Kubernetes:

webserver:
  env:
    - name: AXW_PROPERTIES_aservers
      value: AS1:adonis.host
    - name: AXW_PROPERTIES_aworkers
      value: AS1:54321,AS1:54322
    - name: AXW_PROPERTIES_axw_logger_appender_rest_maxfilesize
      value: 300MB

Docker:

webserver:
  image: adonis.image/webserver
  container_name: webserver
  ...
  environment:
    - AXW_PROPERTIES_aservers=AS1:adonis.host
    - AXW_PROPERTIES_aworkers=AS1:54321,AS1:54322
    - AXW_PROPERTIES_axw_logger_appender_rest_maxfilesize=300MB

Kubernetes Configuration Example​

The following values.yaml example shows a minimal configuration for deploying ADONIS with:

• one application server instance

• multiple aworkers

• database connection settings

• logging configuration

• runtime parameters

It defines the environment variables for both the application server (aserver) and web server (webserver) components, along with resource settings and optional features such as mail and lifecycle endpoints, serving as the input for Kubernetes-based deployment via Helm.

aserver:
  env:
  - name: ADOXX_SERVER_DBNAME
    value: adodb
  - name: ADOXX_SERVER_DBTYPE
    value: PostgreSQL
  - name: ADOXX_SERVER_DBPORT
    value: 5432
  - name: ADOXX_SERVER_DBHOST
    value: postgres.company.com
  - name: ADOXX_SERVER_WEBTIERAUTH_TRUSTED_HOSTS
    value: 127.0.0.*
  - name: ADOXX_LOG_MODE
    value: MIXED
  resources:
    limits:
      memory: 4096Mi
    requests:
      cpu: 500m
      memory: 2048Mi
  workerConf:
    count: 2
instance:
  name: adonis19
webserver:
  env:
  - name: AXW_PROPERTIES_mail_enabled
    value: 'true'
  - name: AXW_PROPERTIES_mail_server
    value: smtp.company.com 
  - name: AXW_PROPERTIES_mail_server_port
    value: '25'
  - name: AXW_PROPERTIES_mail_sender
    value: ADONIS <no-reply@company.com>
  - name: AXW_PROPERTIES_mail_reply_to
    value: no-reply@company.com
  - name: JAVA_OPTS
    value: -Xms512m -Xmx2048m
  - name: AXW_PROPERTIES_base_url
    value: https://adonis19.company.com
  - name: AXW_PROPERTIES_lifecycleops_endpoint_enable
    value: 'true'
  - name: AXW_PROPERTIES_axw_logger_console_enabled
    value: 'true'
  - name: AXW_PROPERTIES_axw_logger_root_level
    value: INFO
  httpRoute:
    hostnames:
    - adonis19.company.com
  resources:
    limits:
      memory: 4096Mi
    requests:
      cpu: 100m
      memory: 2048Mi

Docker Compose Configuration Example​

This Docker Compose setup deploys ADONIS with:

• one application server instance

• two aworker ports

• external database connection

• local volume-based logging

• a shared bridge network

services:
 aserver:
    image: <boc-registry>/adonis/aserver:19.0.0
    container_name: aserver
    user: ado
    ports:
      - 54321:54321
      - 54322:54322
    environment:
      - ADOXX_SERVER_PORTS=54321,54322
      - ADOXX_SERVER_DBNAME=adodb
      - ADOXX_SERVER_DBTYPE=PostgreSQL
      - ADOXX_SERVER_DBHOST=postgres.company.com
#      - ADOXX_SERVER_DBPORT=5432
      - ADOXX_SERVER_WEBTIERAUTH_TRUSTED_HOSTS=*.*.*.*
      - ADOXX_LOG_MODE=MIXED
      - TZ=Europe/Vienna
    volumes:
      - ./logs/aserver:/aserver/logs
    shm_size: "300mb"
    networks:
      aserver_webserver:

 webserver:
    image: <boc-registry>/adonis/webserver:19.0.0
    container_name: webserver
    user: ado
    ports:
      - 8888:8080
    environment:
      - JAVA_OPTS=-Xms512m -Xmx2048m -agentlib:jdwp=transport=dt_socket,server=y,suspend=n,address=*:8000
      - AXW_PROPERTIES_aworkers=AS1:54321,AS1:54322
      - AXW_PROPERTIES_aservers=AS1:aserver
      - AXW_PROPERTIES_axw_logger_console_enabled=true
      - AXW_PROPERTIES_axw_webserver_debug_enabled=false
      - AXW_PROPERTIES_axw_logger_root_level=INFO
      - AXW_PROPERTIES_base_url=https://adonis19.company.com
      - AXW_PROPERTIES_lifecycleops_endpoint_enable=true
      - AXW_PROPERTIES_mail_enabled=true
      - AXW_PROPERTIES_mail_server=smtp.company.com
      - AXW_PROPERTIES_mail_server_port=25
      - AXW_PROPERTIES_mail_sender=ADONIS <no-reply@company.com>
      - AXW_PROPERTIES_mail_reply_to=no-reply@company.com
      - TZ=Europe/Vienna
    volumes:
      - ./logs/webserver:/tomcat/logs
    networks:
      aserver_webserver

networks:
  aserver_webserver:
    driver: bridge
    enable_ipv6: true

Authentication and Identity Integration​

Authentication in ADONIS is handled at the web application layer and is independent of the underlying operating system. The web application has historically supported operation on multiple platforms, including Linux, Solaris, and Windows.

LDAP and SAML Authentication​

LDAP and SAML authentication are unaffected by the migration to a Linux-based, containerised deployment.

Communication with identity systems uses standard network protocols (LDAP, HTTP, HTTPS), which are operating system-agnostic.

Existing LDAP or SAML configurations can be reused without OS-specific changes.

Configuration Impact​

Because authentication is handled in the web application and relies on OS-independent communication, the move to containerised deployment introduces no functional changes for supported authentication mechanisms.

Only environment-specific adjustments (e.g. hostnames or network settings) may be required.

Upgrade from ADONIS 16.x/17.x to ADONIS 19​

The Installation Manual contains a migration guide that will help you upgrade ADONIS from an older version to version 19.0. The guide contains all the steps that need to be taken, with everything explained in detail:

• Upgrade from ADONIS 16.0 – 16.7 or 17.0 – 17.4 to ADONIS 19.0

Migration from All Other Versions to ADONIS 19.0​

ADONIS 18.0 is available only for SaaS customers hosted on AWS. As all upgrades from 18.0 to 19.0 are performed by the BOC Cloud Operations team, no migration guide is provided for this version. If you are using ADONIS 15.0 or an earlier version, please contact your ADONIS consultant for assistance with the required upgrade steps.

Security​

Rights and Permissions​

ADONIS container images do not require root privileges.

All processes inside the containers run under a predefined non-root user (ADO).

No additional host-level permissions or elevated Kubernetes privileges are required.

Minimal Image Principle​

ADONIS container images follow a distroless approach.

Unnecessary operating system components and utilities are removed, reducing the attack surface and enforcing the principle of minimalism.

Cluster Resources​

The ADONIS deployment does not require privileged containers or global Kubernetes resources. In particular:

• no privileged mode

• no NodePorts

• no Kubernetes Operators

• no cluster-wide access rights

This allows deployment in restrictive enterprise clusters with strict security policies.

Namespaces and Resource Isolation​

All required Kubernetes resources can be deployed and operated entirely within a single namespace.

This supports namespace-level isolation, role-based access control (RBAC), and organisation-specific security frameworks.

Storage Permissions​

Persistent storage is required only for specific components, such as the full-text search index and application logs.

The required Persistent Volumes and Persistent Volume Claims are defined through the Helm chart and do not require elevated cluster permissions.

Access is limited to the ADONIS application server and web server containers.

Operations​

Scaling​

Horizontal Pod Autoscaling (HPA) is not supported in ADONIS 19.

Scaling is therefore performed primarily through vertical scaling, by adjusting the CPU and memory resources assigned to the application containers.

The delivered Helm charts support deploying multiple application server instances. As with previous ADONIS releases, running two or more application servers is supported and recommended to distribute the workload across multiple servers.

Resource allocation should be sized according to:

• expected workload

• number of concurrent users

• overall system usage

Logging and Monitoring​

Logging​

Both the web server and application server containers generate log information for system events and errors.

To enable creation of the Support Information Package (SIP) within ADONIS, log files are written to persistent storage. A dedicated persistent volume is required for these logs.

The SIP is required for support and troubleshooting.

In addition to file-based logging, ADONIS also outputs logs via standard container logging streams (stdout/stderr).

This enables integration with external logging solutions such as central log collectors, SIEM platforms, or data lakes.

Log Storage and Retention​

Log files are stored on the configured persistent volume.

Retention, rotation, and rollover behaviour can be configured through the logging setup.

The effective retention period depends on:

• logging configuration

• customer infrastructure

• organisational operational policies

Log Levels​

ADONIS uses the following log levels:

Log levels behave the same way as in previous Windows-based versions of ADONIS (ADONIS 17 and earlier).

Log Configuration​

Logging behaviour can be configured using environment variables provided through:

• values.yaml (Kubernetes)

• docker-compose.yml (Docker)

Configurable parameters include:

• log level

• maximum file size

• maximum history

• total log size limits

Details on available parameters are described in the configuration section.

Data Protection in Logs​

Sensitive values and personal data written to logs are pseudonymised wherever applicable.

This behaviour is consistent with earlier ADONIS versions.

Log Output Format​

The general structure of log output remains largely consistent with previous releases.

Some adjustments have been made to improve consistency across different log types.

Monitoring​

Logs Monitoring​

ADONIS does not provide a built-in logging stack.

Customers are expected to integrate ADONIS logs into their existing observability or monitoring infrastructure.

ADONIS supports this through:

• log output via standard container logging mechanisms (stdout/stderr)

• compatibility with common log collection agents (e.g. Fluent Bit)

• forwarding logs to systems such as OpenSearch, Elasticsearch, or similar platforms

In Kubernetes environments, container logs are typically stored temporarily on the node and collected by a log processor before being forwarded to the organisation’s central log system.

Metrics and operational monitoring can be integrated into existing tools such as Prometheus, Grafana, or other enterprise monitoring platforms.

Monitoring Mechanisms​

ADONIS provides built-in monitoring mechanisms for platform-level health checks.

The delivered container images include functionality used for Kubernetes liveness and readiness probes.

Licensing​

Existing ADONIS licensing remains valid for all current usage scenarios.

The deployment model, whether Windows-based or containerised, does not affect the applicable licensing terms.

Support & lifecycle​

The introduction of the containerised deployment model does not change the existing ADONIS support or product lifecycle policies.

All support terms, maintenance services, and lifecycle rules continue to apply as they did for previous ADONIS versions.