fix(0006d): enforce homogeneous cluster posture + publish posture on /healthz (audit 0008 N1)

A cluster is only as secure as its weakest node: the data plane forwards every
subject between nodes, so one node running without enforced auth lets an
unauthenticated peer Subscribe(">") on it and harvest the traffic forwarded from
the ACL'd nodes.

- validateClusterConfig now takes the auth mode and REFUSES to join a cluster
  unless --bus-auth enforce, regardless of bind (a clustered node is a production
  node; there is no safe dev cluster without auth). This binary therefore cannot
  BE the weak node.
- Server.Posture {enforce,acl,tls,cluster,store} is published on /healthz (non
  secret operational metadata, probe stays unauthenticated) so a monitor or peer
  can detect a cluster member not running enforce+ACL+TLS — covering a peer that
  runs a tampered/old binary outside this node's control.

Tests:
- TestAttack0008_N1: a clustered node with --bus-auth off is refused; the same
  node with enforce + full route security is allowed.
- TestClusterConfigPolicy: extended with off/soft clustered cases (refused) and
  the mode parameter throughout.
- TestHealthExposesPosture: /healthz returns the posture booleans + store backend.

CGO_ENABLED=0 go build/vet/test green; govulncheck 0 reachable.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
2026-06-07 17:17:37 +02:00
parent 18ee7c469b
commit 9b96537aa6
5 changed files with 151 additions and 22 deletions
+18 -1
View File
@@ -83,7 +83,17 @@ func validateBootConfig(bind string, mode membership.AuthMode, tlsCert, tlsKey s
// The three route-TLS paths are all-or-nothing (mutual TLS needs the node cert,
// its key, and the CA together), independent of the bind, so a partial TLS
// config never silently degrades to plaintext routes.
func validateClusterConfig(clusterName, bind, user, pass, rtCert, rtKey, rtCA string) error {
//
// Homogeneous posture (issue 0006d, audit 0008 N1): a cluster is only as secure
// as its weakest node — the data plane forwards every subject between nodes, so a
// single node running without enforced auth lets an unauthenticated peer
// Subscribe(">") on it and harvest the traffic forwarded from the ACL'd nodes.
// This node therefore REFUSES to join a cluster unless it runs --bus-auth enforce,
// regardless of bind: a clustered node is a production node, and there is no safe
// "dev cluster without auth". (A peer running a tampered binary is out of this
// node's control; /healthz exposes each node's posture so a monitor can detect
// one that is not enforce+ACL — see Server.Posture.)
func validateClusterConfig(clusterName, bind, user, pass, rtCert, rtKey, rtCA string, mode membership.AuthMode) error {
rtAny := rtCert != "" || rtKey != "" || rtCA != ""
rtAll := rtCert != "" && rtKey != "" && rtCA != ""
if rtAny && !rtAll {
@@ -93,6 +103,13 @@ func validateClusterConfig(clusterName, bind, user, pass, rtCert, rtKey, rtCA st
if clusterName == "" {
return nil // standalone: no route layer to secure
}
// A clustered node MUST enforce auth (homogeneous posture). Checked before the
// loopback shortcut so even a loopback cluster cannot form without enforce.
if mode != membership.AuthEnforce {
return fmt.Errorf(
"refusing to start: cluster %q requires --bus-auth enforce; a cluster node without enforced auth+ACL lets an unauthenticated peer harvest the traffic forwarded from the other nodes (audit 0008 N1) — every node must run the same enforce+ACL+TLS posture",
clusterName)
}
if isLoopbackBind(bind) {
return nil // loopback cluster is dev-only and unreachable from outside
}