Skip to content
For the complete documentation index, see llms.txt. This page is also available as Markdown.

Troubleshooting

Establish what is wrong, then find the page that fixes it — a triage route for an unknown fault, and an index of where every kind of problem is documented.

Start here when you do not know what is wrong

Diagnostics is the shortest path from "something is broken" to knowing what. It gives you a read-only health snapshot that takes about twenty seconds, a symptom index covering the dozen faults that account for most cases — nothing responding, a sealed secret store, a firewall, an expired certificate, a stopped connector — and a script that collects everything a support case needs.

Run the snapshot first. It tells you which tier is unhappy, and the rest of this section assumes you already know that.

Work from the bottom of the stack upwards. The application services depend on the infrastructure tier, so one problem down there shows up as several services misbehaving at once. Confirm the infrastructure before investigating anything above it.

When you know the symptom

Much of OpenIAM's troubleshooting lives beside the feature it belongs to rather than in this section. This is where each kind of problem is documented.

What is wrong
Where it is covered

An end user sees an error — sign-in, MFA, a request, SSO

One user cannot sign in

SSO or federation is broken

The matching page under Configure federation and application SSO — the SAML, OIDC and JIT pages each carry their own symptom table. To back out a federation you have just set up, or recover from being locked out by one, Federation troubleshooting and recovery

Requests never reach OpenIAM at all

Troubleshoot the gateway — locating the failing hop between the browser and the service

A provisioning operation failed, or succeeded and changed nothing

A synchronization or reconciliation is failing — nothing imported, a run that will not start, the same file imported every time

You do not know which log to read

Effective access disagrees with membership

An object is correct on its own screen but missing from search

Batch or scheduled tasks stopped running

A certification campaign stalled or produced nothing

SoD detection is not reporting violations

Audit events are missing, or timestamps look wrong

One connector is misbehaving

The installation itself failed

You need a deployment back in a clean state

Starting again cleanly — the uninstall script, then reinstall and re-initialize

The areas covered here

Last updated

Was this helpful?