typhoon/google-cloud/container-linux/kubernetes/network.tf
Dalton Hubble e9c8520359 Add experimental Cilium CNI provider
* Accept experimental CNI `networking` mode "cilium"
* Run Cilium v1.8.0-rc4 with overlay vxlan tunnels and a
minimal set of features. We're interested in:
  * IPAM: Divide pod_cidr into /24 subnets per node
  * CNI networking pod-to-pod, pod-to-external
  * BPF masquerade
  * NetworkPolicy as defined by Kubernetes (no L7 Policy)
* Continue using kube-proxy with Cilium probe mode
* Firewall changes:
  * Require UDP 8472 for vxlan (Linux kernel default) between nodes
  * Optional ICMP echo(8) between nodes for host reachability
    (health)
  * Optional TCP 4240 between nodes for endpoint reachability (health)

Known Issues:

* Containers with `hostPort` don't listen on all host addresses,
these workloads must use `hostNetwork` for now
https://github.com/cilium/cilium/issues/12116
* Erroneous warning on Fedora CoreOS
https://github.com/cilium/cilium/issues/10256

Note: This is experimental. It is not listed in docs and may be
changed or removed without a deprecation notice

Related:

* https://github.com/poseidon/terraform-render-bootstrap/pull/192
* https://github.com/cilium/cilium/issues/12217
2020-06-21 20:41:53 -07:00

220 lines
5.4 KiB
HCL

resource "google_compute_network" "network" {
name = var.cluster_name
description = "Network for the ${var.cluster_name} cluster"
auto_create_subnetworks = true
timeouts {
delete = "6m"
}
}
resource "google_compute_firewall" "allow-ssh" {
name = "${var.cluster_name}-allow-ssh"
network = google_compute_network.network.name
allow {
protocol = "tcp"
ports = [22]
}
source_ranges = ["0.0.0.0/0"]
target_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
}
resource "google_compute_firewall" "internal-etcd" {
name = "${var.cluster_name}-internal-etcd"
network = google_compute_network.network.name
allow {
protocol = "tcp"
ports = [2379, 2380]
}
source_tags = ["${var.cluster_name}-controller"]
target_tags = ["${var.cluster_name}-controller"]
}
# Allow Prometheus to scrape etcd metrics
resource "google_compute_firewall" "internal-etcd-metrics" {
name = "${var.cluster_name}-internal-etcd-metrics"
network = google_compute_network.network.name
allow {
protocol = "tcp"
ports = [2381]
}
source_tags = ["${var.cluster_name}-worker"]
target_tags = ["${var.cluster_name}-controller"]
}
# Allow Prometheus to scrape kube-scheduler and kube-controller-manager metrics
resource "google_compute_firewall" "internal-kube-metrics" {
name = "${var.cluster_name}-internal-kube-metrics"
network = google_compute_network.network.name
allow {
protocol = "tcp"
ports = [10251, 10252]
}
source_tags = ["${var.cluster_name}-worker"]
target_tags = ["${var.cluster_name}-controller"]
}
resource "google_compute_firewall" "allow-apiserver" {
name = "${var.cluster_name}-allow-apiserver"
network = google_compute_network.network.name
allow {
protocol = "tcp"
ports = [6443]
}
source_ranges = ["0.0.0.0/0"]
target_tags = ["${var.cluster_name}-controller"]
}
# BGP and IPIP
# https://docs.projectcalico.org/latest/reference/public-cloud/gce
resource "google_compute_firewall" "internal-bgp" {
count = var.networking != "flannel" ? 1 : 0
name = "${var.cluster_name}-internal-bgp"
network = google_compute_network.network.name
allow {
protocol = "tcp"
ports = ["179"]
}
allow {
protocol = "ipip"
}
source_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
target_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
}
# flannel VXLAN
resource "google_compute_firewall" "internal-vxlan" {
count = var.networking == "flannel" ? 1 : 0
name = "${var.cluster_name}-internal-vxlan"
network = google_compute_network.network.name
allow {
protocol = "udp"
ports = [4789]
}
source_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
target_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
}
# Cilium VXLAN
resource "google_compute_firewall" "internal-linux-vxlan" {
count = var.networking == "cilium" ? 1 : 0
name = "${var.cluster_name}-linux-vxlan"
network = google_compute_network.network.name
allow {
protocol = "udp"
ports = [8472]
}
# Cilium health
allow {
protocol = "icmp"
}
allow {
protocol = "tcp"
ports = [4240]
}
source_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
target_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
}
# Allow Prometheus to scrape node-exporter daemonset
resource "google_compute_firewall" "internal-node-exporter" {
name = "${var.cluster_name}-internal-node-exporter"
network = google_compute_network.network.name
allow {
protocol = "tcp"
ports = [9100]
}
source_tags = ["${var.cluster_name}-worker"]
target_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
}
# Allow Prometheus to scrape kube-proxy metrics
resource "google_compute_firewall" "internal-kube-proxy" {
name = "${var.cluster_name}-internal-kube-proxy"
network = google_compute_network.network.name
allow {
protocol = "tcp"
ports = [10249]
}
source_tags = ["${var.cluster_name}-worker"]
target_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
}
# Allow apiserver to access kubelets for exec, log, port-forward
resource "google_compute_firewall" "internal-kubelet" {
name = "${var.cluster_name}-internal-kubelet"
network = google_compute_network.network.name
allow {
protocol = "tcp"
ports = [10250]
}
# allow Prometheus to scrape kubelet metrics too
source_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
target_tags = ["${var.cluster_name}-controller", "${var.cluster_name}-worker"]
}
# Workers
resource "google_compute_firewall" "allow-ingress" {
name = "${var.cluster_name}-allow-ingress"
network = google_compute_network.network.name
allow {
protocol = "tcp"
ports = [80, 443]
}
source_ranges = ["0.0.0.0/0"]
target_tags = ["${var.cluster_name}-worker"]
}
resource "google_compute_firewall" "google-ingress-health-checks" {
name = "${var.cluster_name}-ingress-health"
network = google_compute_network.network.name
allow {
protocol = "tcp"
ports = [10254]
}
# https://cloud.google.com/load-balancing/docs/health-check-concepts#method
source_ranges = [
"35.191.0.0/16",
"130.211.0.0/22",
"35.191.0.0/16",
"209.85.152.0/22",
"209.85.204.0/22",
]
target_tags = ["${var.cluster_name}-worker"]
}