poimenandrock
4b5ecfcd49
CI runner needs kubectl access to create Tekton PipelineRuns for integration testing. ## Root Cause The runner egress NetworkPolicy blocks `192.168.1.0/24` (LAN). The K8s API server runs on control-plane nodes in that subnet (`192.168.1.166:6443`). kubectl from inside the DinD container times out. ## Fix Allow TCP port 6443 to `192.168.1.0/24` — scoped to control-plane API server only. ## Required By homelab-frontend PR #25 (Tekton integration testing) — CI creates PipelineRuns via kubectl. --------- Co-authored-by: rock <[email protected]> Reviewed-on: #48 Co-authored-by: poimen <[email protected]>
62 lines
2.2 KiB
YAML
62 lines
2.2 KiB
YAML
apiVersion: networking.k8s.io/v1
|
|
kind: NetworkPolicy
|
|
metadata:
|
|
name: {{ .Release.Name }}-egress
|
|
namespace: {{ .Release.Namespace }}
|
|
spec:
|
|
podSelector:
|
|
matchLabels:
|
|
app: {{ .Release.Name }}
|
|
policyTypes: [Egress]
|
|
egress:
|
|
# Forgejo — same cicd namespace (git push, registry push/pull)
|
|
- to:
|
|
- podSelector: {}
|
|
# CoreDNS
|
|
- to:
|
|
- namespaceSelector:
|
|
matchLabels:
|
|
kubernetes.io/metadata.name: kube-system
|
|
ports:
|
|
- protocol: UDP
|
|
port: 53
|
|
- protocol: TCP
|
|
port: 53
|
|
# Internet (action deps, base images) — never LAN or pod network
|
|
- to:
|
|
- ipBlock:
|
|
cidr: 0.0.0.0/0
|
|
except:
|
|
- 192.168.1.0/24
|
|
- 10.244.0.0/16
|
|
# Kubernetes API server — CI needs kubectl access to create
|
|
# Tekton PipelineRuns for integration testing.
|
|
# The API server runs on control-plane nodes (192.168.1.0/24);
|
|
# we allow port 6443 only to that subnet.
|
|
- to:
|
|
- ipBlock:
|
|
cidr: 192.168.1.0/24
|
|
ports:
|
|
- protocol: TCP
|
|
port: 6443
|
|
# ingress-nginx, which is how forgejo.riotpiao.com resolves (CoreDNS
|
|
# rewrites that name to ingress-nginx-controller.ingress-nginx.svc).
|
|
# Image pushes go to that name so the tag matches what containerd pulls
|
|
# on the nodes; without this the whole /24 and pod CIDR are denied above
|
|
# and `docker push`/`docker login` hang until they time out.
|
|
#
|
|
# Was an ipBlock pinned to the ingress-nginx LoadBalancer's LAN IP. That
|
|
# stopped matching once DNS started resolving the name to the Service's
|
|
# ClusterIP instead of the LB IP: Cilium enforces egress against the
|
|
# post-DNAT pod IP, which falls inside the 10.244.0.0/16 exclusion above,
|
|
# so every request silently hung rather than erroring. A namespaceSelector
|
|
# follows the Service wherever it resolves and needs no IP to stay in
|
|
# sync with -- same pattern as the kube-system DNS rule above.
|
|
- to:
|
|
- namespaceSelector:
|
|
matchLabels:
|
|
kubernetes.io/metadata.name: ingress-nginx
|
|
ports:
|
|
- protocol: TCP
|
|
port: 443
|