Installation Guide
Build an OpenIAM deployment — choose a route, provision for it, install, and verify.
Applies to: OpenIAM 2026.8.1
This guide builds a working OpenIAM deployment. It assumes you have been asked to stand one up, that you have a terminal and a change window, and that you have not read anything else about the product.
Start with Choose your route. It matches your infrastructure to one of eleven supported routes and lists the pages that route uses, in order.
Building an evaluation and want the short version? Quick start — one host and Quick start — Kubernetes condense the same procedures onto one page each, with the decisions made for you. They are the fast lane, not a second version of the truth — each step names the page that is the authority for it.
Then read Scope and availability. It states what this release covers — all deployments here are single-zone — and what each shape does and does not protect against. Choosing a shape on an incorrect availability assumption is the most expensive mistake available here.
What this guide builds
RPM, single node
Evaluations, development and test, and production where a single host meets the availability requirement. Bundled, external or managed database
RPM, three nodes
Production needing redundancy across the platform's own tiers, tolerating the loss of one application host within a zone
Kubernetes
Organizations already running a cluster, where OpenIAM should share its deployment pipeline, monitoring and upgrade rhythm
Each runs on your own infrastructure, on AWS, or on Azure. Terraform is the recommended route for a new AWS or Azure Kubernetes environment; Helm is the route for a cluster that already exists.
Related guides
Choose a topology and size it
Upgrade an existing deployment
Configure managed systems, roles, policies and campaigns
Get oriented and see the product working
Getting Started
How to use this guide
Choose your route — and read Scope and availability before committing to a shape.
Before you start — what to provision, and how to obtain the software.
Your route's pages, in the order that route lists them.
Verify — every route ends with a verification page confirming the deployment is ready to configure.
Each page opens with what you are about to create, how long it takes, and whether it can be undone, before it tells you what to run. Every step states what you should see, so you always know where you are.
How commands in this guide are written
Every procedure page that needs values from you sets them as shell variables in one block, before the first command. Fill that block in once and the commands below it can be copied and run as they are.
The same names are used throughout:
OPENIAM_HOST
The host you are installing on
OPENIAM_URL
The address users will type, once you have decided it
NODE1, NODE2, NODE3
The three hosts of a three-node deployment. NODE1 is the first node
DB_HOST, DB_PORT, DB_ADMIN
Your database server, its port, and an administrative account on it
OPENIAM_VERSION
The release you are installing
NS, APP
On Kubernetes: the namespace, and the deployment name
RWX_CLASS
On Kubernetes: the storage class that supports ReadWriteMany
DIAG_IMAGE
On Kubernetes: a small pinned utility image, used only by the diagnostic checks
Angle brackets — <like this> — appear only where a value is illustrative rather than something you would set once, and the surrounding text always says what to put there.
If you get stuck
Every path has a troubleshooting page: what you see, what it means, what to do. Diagnostics has a health snapshot, a symptom index and a collection script for a support case.
Last updated
Was this helpful?