goldmane-trail: polish follow-ups #57/#59/#61/#62/#63 + digest→#alerts
All checks were successful
ci/woodpecker/push/default Pipeline was successful

Completes the Goldmane who-talks-to-whom trail (ADR-0014), implemented by a
subagent workflow (distinct stacks in parallel, docs last):

- #57 Whisker gated ingress: ingress_factory (whisker.viktorbarzin.me,
  auth=required, Authentik-gated) + a NetworkPolicy allowing traefik->whisker:8081
  (the operator's whisker NP default-denies ingress). calico stack.
- #61 pipeline health: AggregatorDown + DigestFailing Prometheus alerts
  (prometheus_chart_values.tpl) + cluster-health check #48.
- #59 service-identity labels on the multi-Service namespaces (monitoring's 5
  TF-managed deployments + dbaas), with the KYVERNO_LIFECYCLE_V1 marker so they
  update in-place.
- #62/#63 docs: docs/runbooks/goldmane-flow-trail.md (new), service-catalog,
  security.md + monitoring.md east-west sections, ADR-0014 as-built, CONTEXT.md.
  #62 = the SQL to derive the Wave-1 per-namespace egress allowlist from the
  edge table (feeds code-8ywc; enforce-flips out of scope).

Also fixes the digest's Slack target: #security override 404s channel_not_found
because the shared alertmanager_slack_api_url webhook's app isn't a member of
#security (this likely also breaks alertmanager's slack-security receiver — flagged
in the runbook). Routed to #alerts (the webhook's working channel) until the app
is invited; verified a real digest run posts cleanly (360 edges).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
Viktor Barzin 2026-06-25 17:49:25 +00:00
parent 306cdd4cb3
commit 6c5288998f
17 changed files with 626 additions and 11 deletions

View file

@ -212,3 +212,65 @@ resource "kubectl_manifest" "whisker" {
spec = { notifications = "Disabled" }
})
}
# ---------------------------------------------------------------------------
# Gated public ingress for the Whisker UI (infra #57 / ADR-0014).
#
# whisker.viktorbarzin.me -> whisker:8081, Authentik-gated (auth="required":
# Whisker ships NO own login it's an admin observability UI, so Authentik
# forward-auth is the only gate between strangers and the flow view). The
# operator replicated `tls-secret` into calico-system already.
#
# TWO coupled pieces are required because the operator's own `whisker`
# NetworkPolicy (owned by the Whisker CR above) sets policyTypes:[Ingress]
# with NO ingress rules => default-deny on ingress to the whisker pod. The
# additive NP below ORs in a Traefik allow (k8s NetworkPolicies are additive
# across policies selecting the same pod), so we never edit the operator NP.
module "ingress_whisker" {
source = "../../modules/kubernetes/ingress_factory"
dns_type = "proxied"
namespace = "calico-system"
name = "whisker"
service_name = "whisker"
port = 8081
auth = "required"
tls_secret_name = "tls-secret"
extra_annotations = {
"gethomepage.dev/enabled" = "true"
"gethomepage.dev/name" = "Whisker"
"gethomepage.dev/description" = "Calico flow observability (who-talks-to-whom)"
"gethomepage.dev/icon" = "calico.png"
"gethomepage.dev/group" = "Infrastructure"
}
}
# Additive NetworkPolicy: permit Traefik -> whisker:8081. ORs with the
# operator's default-deny `whisker` NP (selecting the same pod) so Traefik
# can reach the UI without touching the operator-owned policy.
resource "kubernetes_network_policy_v1" "whisker_allow_traefik" {
metadata {
name = "whisker-allow-traefik"
namespace = "calico-system"
}
spec {
pod_selector {
match_labels = {
"app.kubernetes.io/name" = "whisker"
}
}
policy_types = ["Ingress"]
ingress {
from {
namespace_selector {
match_labels = {
"kubernetes.io/metadata.name" = "traefik"
}
}
}
ports {
port = "8081"
protocol = "TCP"
}
}
}
}