Files
lmxopcua/src/Server/ZB.MOM.WW.OtOpcUa.OpcUaServer/OtOpcUaSdkServer.cs
T
Joseph Doherty f692f6afe4 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
2026-08-01 05:00:28 -04:00

154 lines
8.9 KiB
C#

using Opc.Ua;
using Opc.Ua.Server;
using ZB.MOM.WW.OtOpcUa.Commons.OpcUa;
using ZB.MOM.WW.OtOpcUa.Core.Abstractions;
using ZB.MOM.WW.OtOpcUa.Core.Authorization;
using ZB.MOM.WW.OtOpcUa.OpcUaServer.Security;
namespace ZB.MOM.WW.OtOpcUa.OpcUaServer;
/// <summary>
/// <see cref="StandardServer"/> subclass that wires in the v2 <see cref="OtOpcUaNodeManager"/>.
/// Exposes the live node manager after start so callers (<see cref="OpcUaApplicationHost"/>,
/// the fused Host's DI binding) can wrap it in a <see cref="SdkAddressSpaceSink"/> and hand
/// it to <c>OpcUaPublishActor</c>.
/// </summary>
public sealed class OtOpcUaSdkServer : StandardServer
{
private OtOpcUaNodeManager? _otOpcUaNodeManager;
/// <summary>The custom node manager once <c>StartAsync</c> has called
/// <see cref="CreateMasterNodeManager"/>. Null until the SDK has bootstrapped.</summary>
public OtOpcUaNodeManager? NodeManager => _otOpcUaNodeManager;
/// <summary>
/// Wire the reverse-path sink for inbound Part 9 alarm method calls onto the created
/// <see cref="OtOpcUaNodeManager"/>. The host calls this after start with a non-blocking
/// <c>mediator.Tell</c> that publishes each <see cref="AlarmCommand"/> onto the
/// <c>alarm-commands</c> DistributedPubSub topic. No-op (returns <c>false</c>) when the node
/// manager has not been created yet, so the caller can detect a too-early call.
/// </summary>
/// <param name="router">The router invoked by the condition handlers once the <c>AlarmAck</c>
/// gate passes; may be <c>null</c> to clear it.</param>
/// <returns><c>true</c> when the router was set on a live node manager; <c>false</c> when no node
/// manager exists yet.</returns>
public bool SetAlarmCommandRouter(Action<AlarmCommand>? router)
{
if (_otOpcUaNodeManager is null) return false;
_otOpcUaNodeManager.AlarmCommandRouter = router;
return true;
}
/// <summary>
/// Wire the reverse-path router for inbound Part 9 Acknowledge of a NATIVE (driver-fed, e.g. Galaxy)
/// condition onto the created <see cref="OtOpcUaNodeManager"/>. The host calls this after start with a
/// non-blocking router that Tells a <c>DriverHostActor.RouteNativeAlarmAck</c> into the local
/// <c>DriverHostActor</c>, which (Primary-gated) routes the acknowledge to the owning driver child.
/// Native conditions are not owned by the scripted-alarm engine, so the node manager branches their
/// inbound Acknowledge to this seam instead of the scripted <see cref="SetAlarmCommandRouter"/> path.
/// Passing <c>null</c> clears it. No-op (returns <c>false</c>) when the node manager has not been
/// created yet, so the caller can detect a too-early call (mirrors <see cref="SetAlarmCommandRouter"/>).
/// </summary>
/// <param name="router">The router invoked by the native condition's Acknowledge handler once the
/// <c>AlarmAck</c> gate passes; may be <c>null</c> to clear it.</param>
/// <returns><c>true</c> when the router was set on a live node manager; <c>false</c> when no node
/// manager exists yet.</returns>
public bool SetNativeAlarmAckRouter(Action<NativeAlarmAck>? router)
{
if (_otOpcUaNodeManager is null) return false;
_otOpcUaNodeManager.NativeAlarmAckRouter = router;
return true;
}
/// <summary>
/// Wire the reverse-path gateway for inbound operator writes to writable equipment-tag nodes onto the
/// created <see cref="OtOpcUaNodeManager"/>. The host calls this after start with an
/// <c>ActorNodeWriteGateway</c> whose <c>WriteAsync</c> kicks off a bounded <c>Ask</c> of the local
/// <c>DriverHostActor</c> (<c>RouteNodeWrite</c>); the node manager calls it fire-and-forget and uses
/// the resolved <c>NodeWriteOutcome</c> to self-correct the node on a failed device write (mirrors
/// <see cref="SetAlarmCommandRouter"/>). Passing <c>null</c> restores the
/// <c>NullOpcUaNodeWriteGateway</c> default (writes unavailable). No-op (returns <c>false</c>) when the
/// node manager has not been created yet, so the caller can detect a too-early call.
/// </summary>
/// <param name="gateway">The gateway invoked by the writable node's <c>OnWriteValue</c> handler once the
/// <c>WriteOperate</c> gate passes; may be <c>null</c> to restore the Null default.</param>
/// <returns><c>true</c> when the gateway was set on a live node manager; <c>false</c> when no node
/// manager exists yet.</returns>
public bool SetNodeWriteGateway(IOpcUaNodeWriteGateway? gateway)
{
if (_otOpcUaNodeManager is null) return false;
// The NodeWriteGateway setter null-coalesces to the Null default, so a null gateway is intentional
// (restores "writes unavailable"); forgive the nullable-in here.
_otOpcUaNodeManager.NodeWriteGateway = gateway!;
return true;
}
/// <summary>
/// Wire the server-side HistoryRead backend (the <see cref="IHistorianDataSource"/> the node
/// manager's HistoryRead overrides block-bridge to) onto the created
/// <see cref="OtOpcUaNodeManager"/>. The host calls this after start with the DI-resolved source —
/// the <c>NullHistorianDataSource</c> default (GoodNoData-empty reads) or the configured HistorianGateway
/// read client. Passing <c>null</c> restores the Null default (the property setter null-coalesces),
/// i.e. "no historian". No-op (returns <c>false</c>) when the node manager has not been created yet,
/// so the caller can detect a too-early call (mirrors <see cref="SetNodeWriteGateway"/>).
/// </summary>
/// <param name="source">The read backend invoked by the node manager's HistoryRead overrides; may be
/// <c>null</c> to restore the Null default (GoodNoData-empty reads).</param>
/// <returns><c>true</c> when the source was set on a live node manager; <c>false</c> when no node
/// manager exists yet.</returns>
public bool SetHistorianDataSource(IHistorianDataSource? source)
{
if (_otOpcUaNodeManager is null) return false;
// The HistorianDataSource setter null-coalesces to the Null default, so a null source is intentional
// (restores GoodNoData-empty reads); forgive the nullable-in here.
_otOpcUaNodeManager.HistorianDataSource = source!;
return true;
}
/// <summary>
/// Wire the node's data-plane authorization inputs (#520 stage 4) onto the created
/// <see cref="OtOpcUaNodeManager"/> — the cluster it serves, the current auth generation, and the
/// ACL grants. The host calls this after start with the DI-resolved
/// <see cref="IDataPlaneAuthoritySource"/>, which the artifact-apply path refreshes on every
/// deploy. Passing <c>null</c> restores the empty source (authorizes nothing). No-op (returns
/// <c>false</c>) when the node manager has not been created yet, so the caller can detect a
/// too-early call (mirrors <see cref="SetHistorianDataSource"/>).
/// </summary>
/// <param name="source">The node's data-plane authorization inputs; may be <c>null</c> to restore the
/// empty default.</param>
/// <returns><c>true</c> when the source was set on a live node manager; <c>false</c> when no node
/// manager exists yet.</returns>
public bool SetDataPlaneAuthority(IDataPlaneAuthoritySource? source)
{
if (_otOpcUaNodeManager is null) return false;
// The DataPlaneAuthority setter null-coalesces to the empty source, so a null is intentional;
// forgive the nullable-in here.
_otOpcUaNodeManager.DataPlaneAuthority = source!;
return true;
}
/// <summary>
/// Wire the node-scoped ACL gate (#520 stage 5) onto the created <see cref="OtOpcUaNodeManager"/>.
/// The host calls this after start; passing <c>null</c> detaches it, which returns the write and
/// alarm seams to their server-wide role checks and leaves HistoryRead ungated. No-op (returns
/// <c>false</c>) when the node manager has not been created yet, mirroring
/// <see cref="SetDataPlaneAuthority"/>.
/// </summary>
/// <param name="authorizer">The gate, or <c>null</c> to detach.</param>
/// <returns><c>true</c> when set on a live node manager; <c>false</c> when none exists yet.</returns>
public bool SetDataPlaneAuthorizer(DataPlaneAuthorizer? authorizer)
{
if (_otOpcUaNodeManager is null) return false;
_otOpcUaNodeManager.Authorizer = authorizer;
return true;
}
/// <inheritdoc />
protected override MasterNodeManager CreateMasterNodeManager(
IServerInternal server, ApplicationConfiguration configuration)
{
_otOpcUaNodeManager = new OtOpcUaNodeManager(server, configuration);
return new MasterNodeManager(server, configuration, dynamicNamespaceUri: null, _otOpcUaNodeManager);
}
}