feat(security): enforce node ACLs on write, alarms and HistoryRead (#520 stage 5)

`Security:Acl:Enabled` (default false) makes scoped `NodeAcl` grants decide
data-plane authorization at the three seams that have one, and gives HistoryRead
a check it never had. Read, Browse and Subscribe stay ungated — stage 6.

The ACL decision REPLACES the server-wide role check rather than chaining with
it. Chained, a grant would have to confer the role as well as pass the scope
check, and only Cluster-scoped grants confer a role — so every narrow grant
would have stayed inert and the flag would have looked inert too. The
load-bearing test is an ALLOW: a session carrying no roles at all writing
successfully on the strength of its grant. A deny-only test passes against a
build where the ACL is never consulted, because the role gate denies too.

Found en route: the principal never reached the session. The authenticator
passed it as OpcUaUserAuthResult.DisplayName, but RoleCarryingUserIdentity
derives DisplayName from the presented token — for a key-bearing client that is
the static username `apikey`, identical fleet-wide. So the trie's matching key
was dropped at the identity boundary, and the audit ClientUserId on a rejected
write recorded `apikey` for every consuming system. Principal and ApiKeyId are
now carried explicitly.

Beyond the stage as sketched:

- Revocation reaches an OPEN session. Every decision re-checks the
  authenticating key against the current snapshot; without it a key revoked by a
  deployment kept its authority until the client chose to reconnect, which for a
  machine-to-machine consumer can be indefinitely.
- UserAuthorizationState is built per CALL, not per session. The per-session
  shape exists so an LDAP session can re-resolve membership that changed out of
  band; nothing can change out of band under API keys, so the generation stamp is
  the whole invalidation signal and a per-call record picks it up for free.
  IsStale/NeedsRefresh therefore cannot fire here and are documented as vestigial.
- The trie is built inside DataPlaneAuthorityState.Apply, under the same
  generation as the snapshot, rather than by a second wiring at the publish
  actor. An uninstalled trie denies everything — indistinguishable from working
  enforcement. Deleting the install turns 13 of the 22 new tests red.
- The deploy log reports what the BUILD dropped instead of re-deriving it. The
  old count agreed with the builder on one of its four drop reasons; a rule with
  a blank ScopeId, a retired scope kind, or a path not starting at a realm
  vanished with the log reading zero.

Both IApiKeyStore implementations widened to the full grant set. The narrowing
was redundant — ApiKeyInterimRoleMapper applies the same filter itself — and
became a trap once scope was enforced. Pinned by asserting the roles are
unchanged, with the two grants carrying different flags so it cannot pass by
accident.

Falsified by mutation: enforcement never engaging, the write seam ignoring the
ACL, the realm dropped from the scope, the revocation check removed, and the
trie never installed each turn the intended tests red and no others.

Full solution suite green.

Claude-Session: https://claude.ai/code/session_015p7wGqy3YpZNCpDzTpGMKo
This commit is contained in:
Joseph Doherty
2026-08-01 05:00:28 -04:00
parent 9499099028
commit f692f6afe4
25 changed files with 1845 additions and 151 deletions
@@ -138,6 +138,9 @@ builder.AddZbSerilog(o => o.ServiceName = "otopcua");
// to MINT them, and getting this wrong on either side is a silent authentication failure rather than a
// startup error.
builder.Services.Configure<ApiKeyOptions>(builder.Configuration.GetSection(ApiKeyOptions.SectionName));
// #520 stage 5 — Security:Acl:Enabled. Default false; see AclOptions for why the flip must be a config
// change rather than a redeploy, and what it swaps rather than adds.
builder.Services.Configure<AclOptions>(builder.Configuration.GetSection(AclOptions.SectionName));
if (hasAdmin)
builder.Services.AddOtOpcUaConfigDb(builder.Configuration);
@@ -360,6 +363,14 @@ if (hasDriver)
builder.Services.AddSingleton<DataPlaneAuthorityState>();
builder.Services.AddSingleton<IDataPlaneAuthoritySource>(sp => sp.GetRequiredService<DataPlaneAuthorityState>());
builder.Services.AddSingleton<IDataPlaneAuthoritySink>(sp => sp.GetRequiredService<DataPlaneAuthorityState>());
// #520 stage 5 — the tries the ACL gate evaluates against are taken FROM the authority state, not
// registered independently. The state builds the trie inside the same Apply() call that installs
// the snapshot, so resolving any other cache here would hand the evaluator a permanently empty
// one — and an empty trie denies everything, which looks exactly like "the ACLs are working" until
// someone checks that an ALLOW is possible at all. Deriving it from the owner makes that mistake
// unrepresentable.
builder.Services.AddSingleton(sp => sp.GetRequiredService<DataPlaneAuthorityState>().Tries);
}
else
{
@@ -369,8 +380,24 @@ if (hasDriver)
// fix, as the NullOpcUaUserAuthenticator registration below. Only the SINK is genuinely optional
// (nothing publishes artifacts on a node with no driver role).
builder.Services.AddSingleton<IDataPlaneAuthoritySource>(EmptyDataPlaneAuthoritySource.Instance);
// Nothing installs a trie on a node with no driver role; an empty cache denies every scoped
// decision, which is correct for a node that serves no data plane.
builder.Services.AddSingleton<PermissionTrieCache>();
}
// #520 stage 5 — the node-scoped ACL gate. Registered on EVERY node, not just driver ones, because
// OtOpcUaServerHostedService takes it as a constructor parameter and the built-in container resolves
// constructors by whether every parameter is registered, not by whether the parameter type is nullable.
// That is the same trap the IOpcUaUserAuthenticator and IDataPlaneAuthoritySource registrations above
// both carry a comment about; an absent registration is a node that will not start.
builder.Services.AddSingleton<IPermissionEvaluator>(sp =>
new TriePermissionEvaluator(sp.GetRequiredService<PermissionTrieCache>()));
builder.Services.AddSingleton(sp => new DataPlaneAuthorizer(
sp.GetRequiredService<IDataPlaneAuthoritySource>(),
sp.GetRequiredService<IPermissionEvaluator>(),
sp.GetRequiredService<IOptions<AclOptions>>().Value.Enabled,
sp.GetRequiredService<ILoggerFactory>().CreateLogger<DataPlaneAuthorizer>()));
if (hasAdmin)
{
// A node WITH a ConfigDb reads keys straight from it, which is what makes revocation take effect