AAP 2.6 Tested Deployment Models: Growth & Enterprise Topologies Guide

Q: Can I use a topology not listed by Red Hat?

Yes, but Red Hat only provides "commercially reasonable support" for non-tested topologies. Stick to tested topologies for production environments to ensure full support coverage.

Q: Can I start with growth and migrate to enterprise later?

Yes. Start with the growth topology and scale to enterprise as your needs grow. The migration involves adding nodes and redistributing components — the AAP installer handles this.

Q: Do I need a load balancer for growth topology?

No. Growth topology runs everything on a single host. A load balancer is only needed for enterprise topology with multiple gateway or controller nodes.

Q: What's the minimum for a production deployment?

Red Hat recommends the enterprise topology for production. At minimum: 2 gateway nodes (HA), 2 controller nodes, 1 hub, 1 EDA controller, 1 dedicated database server. Total: 7 VMs.

By Luca Berton · Published 2024-01-01 · Category: installation

Complete guide to AAP 2.6 tested deployment models and reference architectures. Covers container growth and enterprise topologies, OpenShift operator.

Red Hat tests Ansible Automation Platform 2.6 against specific deployment topologies — use these reference architectures for supported, production-ready deployments. Choose between growth topologies (single-node, resource-efficient) and enterprise topologies (multi-node, redundant).

Deployment Models Overview

| Mode | Infrastructure | Topologies | |------|---------------|------------| | Containers | VMs / Bare metal (Podman) | Growth, Enterprise | | Operator | Red Hat OpenShift | Growth, Enterprise | | RPM (deprecated) | VMs / Bare metal | Growth, Enterprise |

Container Growth Topology

Best for: Organizations getting started, development environments, or resource-constrained deployments.

All AAP components run on a single VM:

┌─────────────────────────────┐
│        Single RHEL Host     │
│                             │
│  ┌──────────────────────┐   │
│  │  Platform Gateway    │   │
│  ├──────────────────────┤   │
│  │  Automation Controller│  │
│  ├──────────────────────┤   │
│  │  Automation Hub      │   │
│  ├──────────────────────┤   │
│  │  Event-Driven Ansible│   │
│  ├──────────────────────┤   │
│  │  PostgreSQL Database │   │
│  ├──────────────────────┤   │
│  │  Redis (standalone)  │   │
│  └──────────────────────┘   │
└─────────────────────────────┘

System Requirements

| Resource | Minimum | |----------|---------| | RAM | 16 GB (32 GB with hub_seed_collections=true) | | CPUs | 4 | | Disk | 60 GB total (15 GB install dir, 1-3 GB /var/tmp, 10 GB temp for bundled) | | Disk IOPS | 3000 | | OS | RHEL 9.4+ or RHEL 10+ |

Inventory Groups

| Purpose | Group Names | |---------|------------| | All components | automationgateway, automationcontroller, automationhub, automationeda, database |

All groups point to the same host.

Container Enterprise Topology

Best for: Production deployments requiring high availability, redundancy, and scalability.

Components distributed across multiple VMs:

┌──────────────┐  ┌──────────────┐
│  Gateway 1   │  │  Gateway 2   │
└──────┬───────┘  └──────┬───────┘
       │                 │
┌──────┴─────────────────┴───────┐
│         Load Balancer          │
└──────┬─────────────────┬───────┘
       │                 │
┌──────┴───────┐  ┌──────┴───────┐
│ Controller 1 │  │ Controller 2 │
└──────────────┘  └──────────────┘
       │                 │
┌──────┴───────┐  ┌──────┴───────┐
│  Hub Node    │  │  EDA Node    │
└──────────────┘  └──────────────┘
       │                 │
┌──────┴─────────────────┴───────┐
│       PostgreSQL Database      │
│    (external or dedicated)     │
└────────────────────────────────┘

Additional Requirements

• Multiple VMs for redundancy • External load balancer • Redis in cluster mode • Dedicated database server recommended • Receptor mesh for execution nodes

OpenShift Operator Topologies

Operator Growth Topology

Single namespace deployment on OpenShift — suitable for development and small teams.

Operator Enterprise Topology

Multi-namespace or multi-cluster deployment with: • Separate namespaces for controller, hub, and EDA • Horizontal pod autoscaling • External PostgreSQL for data durability • Multi-AZ deployment for HA

RPM Topologies (Deprecated)

⚠️ Only for RHEL 9 during AAP 2.6 lifecycle to support migrations. Will be removed in AAP 2.7.

Same growth/enterprise patterns as containerized, but installed directly via RPM packages instead of Podman containers.

Network Port Requirements

All topologies require these ports:

Platform Communication

| Port | Protocol | From → To | Purpose | |------|----------|-----------|---------| | 80/443 | TCP | Gateway → Controller | Platform API | | 80/443 | TCP | Gateway → Hub | Collection/image access | | 80/443 | TCP | Gateway → EDA | Event routing | | 80/443 | TCP | EDA → Controller | Launch jobs | | 80/443 | TCP | EDA → Hub | Pull decision environments | | 80/443 | TCP | Controller → Hub | Pull collections & EEs |

Database & Infrastructure

| Port | Protocol | From → To | Purpose | |------|----------|-----------|---------| | 5432 | TCP | All components → Database | PostgreSQL | | 6379 | TCP | EDA/Gateway → Redis | Data storage | | 27199 | TCP | Controller → Execution nodes | Receptor mesh |

Component NGINX Ports

| Port | Component | Configurable Via | |------|-----------|-----------------| | 8080/8443 | Controller | controller_nginx_http_port, controller_nginx_https_port | | 8081/8444 | Hub | hub_nginx_http_port, hub_nginx_https_port | | 8082/8445 | EDA | eda_nginx_http_port, eda_nginx_https_port | | 8083/8446 | Gateway | gateway_nginx_http_port, gateway_nginx_https_port |

Supported Configurations

| Setting | Supported Options | |---------|------------------| | OS | RHEL 9.4+, RHEL 10+ | | CPU Architecture | x86_64, AArch64, s390x, ppc64le | | ansible-core | 2.14 (RHEL 9), 2.16 (RHEL 10) | | Database | PostgreSQL 15 (managed), 15/16/17 (external) | | IP Version | IPv4, IPv6 (single-stack and dual-stack) | | Browser | Current Firefox or Chrome |

Choosing Your Topology

Use Growth When:

• Evaluating or testing AAP • Small team (< 20 users) • Budget-constrained • Development/staging environment • < 100 managed hosts

Use Enterprise When:

• Production workloads • High availability required • Multiple teams sharing the platform • Compliance requirements (audit trails, separation) • 100+ managed hosts • 24/7 uptime expectations

Database Considerations

Managed (Installer-Controlled)

• PostgreSQL 15 installed and managed by the AAP installer • Backup/restore handled by AAP tools • Simplest option for growth topology

External (Customer-Provided)

• PostgreSQL 15, 16, or 17 • ICU support must be enabled (not default on AWS RDS, Azure SQL, EDB) • PostgreSQL 16/17 must handle backup/restore independently • Required for enterprise topology with dedicated DB servers

FAQ

Can I use a topology not listed by Red Hat?