d602063448
cluster-issuer.yaml: name → letsencrypt-prod-{{ tenant.slug }}, hard-pin
apiTokenSecretRef.name to cloudflare-api-token-{{ tenant.slug }} so it
matches the ESO-created Secret. ACME account key also slug-suffixed
for tenant isolation. Pre-0.7.3 the unsuffixed letsencrypt-prod
mismatched what instance.go:504 stamps into per-instance Certificates
(letsencrypt-prod-<slug>), so cert-manager logged 'Referenced
ClusterIssuer not found' and erp2 served Traefik default cert forever.
tenants-wildcard-cert.yaml: issuerRef.name → letsencrypt-prod-{{ $.Values.tenant.slug }}
to match the renamed ClusterIssuer.
values.yaml: secrets.cloudflareTokenSecret block deprecated (the chart
no longer reads it; kept for back-compat with external overrides).
Diagnosed in the qsoft2 migrate test 2026-05-09.
67 lines
3.0 KiB
YAML
67 lines
3.0 KiB
YAML
apiVersion: v2
|
|
name: cluster-platform-v3
|
|
description: |
|
|
Per-cluster platform infrastructure for OdooSky v3. ArgoCD-managed
|
|
on every connected customer K8s cluster. Provides:
|
|
- odoosky-system namespace (where Tower spawns build Jobs and
|
|
stores cluster-private credentials sourced from OpenBao)
|
|
- Local container registry (Distribution v2). In-cluster
|
|
BuildKit Jobs push addon images here; the chart consumes
|
|
them as image volumes. Sovereignty + GFW resistance: no
|
|
cross-cluster image transfer.
|
|
- cert-manager + Traefik (vendored via Helm dependencies)
|
|
so the substrate that used to be installed by bootstrap.sh
|
|
now lives in Git, deployed by Tower's per-cluster Argo
|
|
Application. Customer's "Connect Server" terminal stops
|
|
at "kubeconfig sent" — the slow ACME wait happens here in
|
|
the background.
|
|
- tenants Namespace + tenants-wildcard Certificate. Per-tenant
|
|
via .Values.tenant.{domain,wildcardHost}; cert-manager's
|
|
DNS-01 solver pulls the Cloudflare token from the
|
|
`cloudflare-api-token` Secret Tower kubectl-applies into the
|
|
cert-manager namespace at Connect time (secrets stay out of
|
|
Git).
|
|
|
|
type: application
|
|
version: 0.7.3
|
|
appVersion: "0.7.3"
|
|
|
|
# All 4 subcharts now resolve from registry.odoosky.cloud (mirrored
|
|
# 2026-05-08). Mirror-first discipline + China-region readiness: a
|
|
# Jetstack / Traefik / Longhorn / external-secrets-io outage no longer
|
|
# blocks new tenant cluster bootstrap.
|
|
#
|
|
# Original upstream sources (for re-sync if a chart bumps):
|
|
# cert-manager → https://charts.jetstack.io
|
|
# traefik → https://traefik.github.io/charts
|
|
# longhorn → https://charts.longhorn.io
|
|
# external-secrets → https://charts.external-secrets.io
|
|
#
|
|
# Re-sync recipe: `helm pull <chart> --repo <upstream> --version <v>`
|
|
# then `helm push <tgz> oci://registry.odoosky.cloud/odoosky/docker-mirror/charts`.
|
|
dependencies:
|
|
- name: cert-manager
|
|
version: "v1.16.1"
|
|
repository: "oci://registry.odoosky.cloud/odoosky/docker-mirror/charts"
|
|
condition: certManager.enabled
|
|
- name: traefik
|
|
version: "33.2.1"
|
|
repository: "oci://registry.odoosky.cloud/odoosky/docker-mirror/charts"
|
|
condition: traefik.enabled
|
|
# Longhorn — CSI block storage with snapshot + clone primitives.
|
|
# See ADR 0003 (in odooskyv3 monorepo). Phase 1 declares the
|
|
# dependency but the chart's default is `longhorn.enabled=false`,
|
|
# so `helm dep update` skips it on render unless a per-cluster
|
|
# Argo Application sets the flag.
|
|
- name: longhorn
|
|
version: "1.7.2"
|
|
repository: "oci://registry.odoosky.cloud/odoosky/docker-mirror/charts"
|
|
condition: longhorn.enabled
|
|
# External Secrets Operator — declarative Secret delivery from
|
|
# OpenBao. Replaces Tower's imperative kubectl-stamp pattern for
|
|
# the 5 substrate Secrets (Item #9, all phases shipped 2026-05-07).
|
|
- name: external-secrets
|
|
version: "0.10.7"
|
|
repository: "oci://registry.odoosky.cloud/odoosky/docker-mirror/charts"
|
|
condition: externalSecrets.enabled
|