Docs

Components

What this page is for

The pieces Fenwave is made of, and which piece to suspect when something misbehaves. This is the page that turns “Fenwave is broken” into a specific question.

Before you start

Nothing. This is reference material.

The pieces

Piece What it does Symptoms when it is the problem
Web interface The sidebar and every screen in these docs Menu items missing (permissions), stale data after a change (session cache)
Fenwave backend Serves the APIs behind every screen, and holds the registries 401 / 403 / 503 responses; empty lists where data should be
Catalog The register of components, systems, APIs and owners Something you registered is not showing; ownership blank
TechDocs Builds and serves documentation from repositories A component’s Docs tab is empty; a language switch 404s
Argo CD Executes deployments Deployments do not progress
Argo Workflows Executes pipelines and test runs Pipelines and test runs submit but never start
Kubernetes clusters Where everything actually runs Workloads crash-looping, OOM kills, pods not scheduling
Cloud accounts Credentials the clusters authenticate through Clusters go invalid together after a credential rotation
Object storage S3 / GCS buckets Syncs failing; objects missing at a prefix
Fenwave Agent / DevApp Local development tooling on your own machine Local setup problems only — unrelated to the hosted platform

Which piece to suspect

What you see Suspect Go to
A menu item is missing Permissions Access Explorer
403 with a permission name Permissions How permissions work
503 An integration is not configured The relevant Settings area
Deployments stuck Argo CD Settings › Development › CI/CD
Pipelines or test runs never start Argo Workflows CI/CD — a separate integration from Argo CD
Many clusters invalid at once The shared cloud credential Cloud
One workload misbehaving The cluster Cluster Console
Docs missing or 404 on language switch TechDocs Docs

Scenario

“Fenwave is down” — narrowing it in five minutes.

Almost never is the whole platform down. The job is to find which piece.

  1. Can you load the sidebar at all? If yes, the web interface and backend are alive, and this is not an outage — go to step 2.
  2. Is the problem confined to one menu area? A single area failing points at that area’s integration, not the platform.
  3. Deployments stuck but pipelines fine → Argo CD. Pipelines stuck but deployments fine → Argo Workflows. Both fine but one cluster misbehaving → the cluster.
  4. Several clusters invalid at once → a shared cloud credential rotated. One cluster invalid → that cluster.
  5. Only one person affected → permissions, not infrastructure. Confirm with Access Explorer before spending time elsewhere.
  6. Report what you narrowed it to, not “Fenwave is down”. The distinction between step 3’s two branches usually decides who needs to be woken up.

When it doesn’t work

Symptom Cause How to check Fix
Everything returns 503 A backing integration is unconfigured The relevant Settings area Configure and test it
One user affected, others fine Permissions Access Explorer Grant the right role
Everyone affected in one area That area’s integration Its Settings page, “Last Tested” Test and repair it
Data stale after a change Session cache, or a periodic refresh — Reload; wait for the refresh

Next