Usage of ActiveInCluster should be justified (#10485)

## What changed?

1. Add linter rule to guard`ActiveInCluster` usage.
2. Fixed the callsites for `ActiveInCluster` where should use
`ActiveClusterName` instead.
3. Add explicit `//nolint:forbidigo justifications` for genuine
namespace-level checks.

## Why?

ActiveInCluster is namespace-level and not businessID-aware. For
workflow-scoped decisions, the active cluster must be resolved with
`ActiveClusterName(routingKey)` so future businessID-aware routing does
not incorrectly treat an entire namespace as active or standby.

## How did you test it?
- [x] built
- [ ] run locally and tested manually
- [x] covered by existing tests
- [ ] added new unit test(s)
- [ ] added new functional test(s)

Author: jiechenz

.github/.golangci.yml

@@ -52,6 +52,8 @@ linters:
           msg: "Use await.Require / s.Await for assertion conditions, or await.RequireTrue / s.AwaitTrue for bool predicates, instead of testify Eventually helpers"
         - pattern: 'assert\.\w+'
           msg: "Use require.X / protorequire.X instead of assert.X / protoassert.X — assert doesn't stop the test on failure."
+        - pattern: '\bActiveInCluster\b'
+          msg: "ActiveInCluster is not businessID-aware. For any decision tied to a workflow / businessID, use ns.ActiveClusterName(routingKey) instead. For genuine namespace-level checks (no workflow context), suppress with //nolint:forbidigo // <justification>."
     depguard:
       rules:
         main:
@@ -200,6 +202,10 @@ linters:
         text: "FunctionalTestBase"
         linters:
           - forbidigo
+      - path: common/namespace/(namespace|replication_resolver)(_test)?\.go$ # ActiveInCluster definition and direct unit tests
+        text: "ActiveInCluster"
+        linters:
+          - forbidigo
       - path-except: tests/.+_test\.go # only enforce in test files
         text: "context.Background"
         linters:

common/namespace/nsregistry/registry.go

@@ -185,6 +185,7 @@ func DefaultNamespaceStateChanged(currentClusterName string, oldNS *namespace.Na
 		oldNS.State() != newNS.State() ||
 		oldNS.Name() != newNS.Name() ||
 		oldNS.IsGlobalNamespace() != newNS.IsGlobalNamespace() ||
+		//nolint:forbidigo // ns-wide state diff for cache invalidation.
 		oldNS.ActiveInCluster(currentClusterName) != newNS.ActiveInCluster(currentClusterName) ||
 		oldNS.ReplicationState("") != newNS.ReplicationState("")
 }

service/frontend/nexus_handler.go

@@ -187,6 +187,7 @@ func (c *operationContext) interceptRequest(
 		return commonnexus.ConvertGRPCError(err, false)
 	}
 
+	//nolint:forbidigo // Nexus requests are not tied to a business ID by design (see line 184)
 	if !c.namespace.ActiveInCluster(c.clusterMetadata.GetCurrentClusterName()) {
 		if c.shouldForwardRequest(ctx, header) {
 			// Handler methods should have special logic to forward requests if this method returns

service/frontend/workflow_handler.go

@@ -414,6 +414,7 @@ func (wh *WorkflowHandler) Start() {
 
 			if ns.IsGlobalNamespace() &&
 				ns.ReplicationPolicy() == namespace.ReplicationPolicyMultiCluster &&
+				//nolint:forbidigo // namespace state-change callback; cancels all pollers on ns deactivation
 				!ns.ActiveInCluster(wh.clusterMetadata.GetCurrentClusterName()) {
 				pollers, ok := wh.outstandingPollers.Get(ns.ID().String())
 				if ok {

service/history/api/deleteworkflow/api.go

@@ -55,7 +55,7 @@ func Invoke(
 		if err != nil {
 			return nil, err
 		}
-		if ns.ActiveInCluster(shardContext.GetClusterMetadata().GetCurrentClusterName()) {
+		if ns.ActiveClusterName(namespace.RoutingKey{ID: request.WorkflowExecution.GetWorkflowId()}) == shardContext.GetClusterMetadata().GetCurrentClusterName() {
 			// If workflow execution is running and in active cluster.
 			if err := api.UpdateWorkflowWithNew(
 				shardContext,

service/history/api/queryworkflow/api.go

@@ -156,7 +156,7 @@ func Invoke(
 		// 1. the namespace is not active, in this case history is immutable so a query dispatched at any time is consistent
 		// 2. the workflow is not running, whenever a workflow is not running dispatching query directly is consistent
 		// 3. if there is no pending or started workflow tasks it means no events came before query arrived, so its safe to dispatch directly
-		safeToDispatchDirectly := !nsEntry.ActiveInCluster(shardContext.GetClusterMetadata().GetCurrentClusterName()) ||
+		safeToDispatchDirectly := nsEntry.ActiveClusterName(namespace.RoutingKey{ID: workflowKey.WorkflowID}) != shardContext.GetClusterMetadata().GetCurrentClusterName() ||
 			!mutableState.IsWorkflowExecutionRunning() ||
 			(!mutableState.HasPendingWorkflowTask() && !mutableState.HasStartedWorkflowTask())
 		if safeToDispatchDirectly {

service/history/chasm_engine.go

@@ -585,7 +585,7 @@ func (e *ChasmEngine) deleteExecution(
 		if err != nil {
 			return err
 		}
-		if ns.ActiveInCluster(shardContext.GetClusterMetadata().GetCurrentClusterName()) {
+		if ns.ActiveClusterName(namespace.RoutingKey{ID: ref.BusinessID}) == shardContext.GetClusterMetadata().GetCurrentClusterName() {
 			chasmTree, ok := mutableState.ChasmTree().(*chasm.Node)
 			if !ok {
 				return serviceerror.NewInternalf(

service/history/history_engine.go

@@ -393,6 +393,7 @@ func (e *historyEngineImpl) registerNamespaceStateChangeCallback() {
 
 		if ns.IsGlobalNamespace() &&
 			ns.ReplicationPolicy() == namespace.ReplicationPolicyMultiCluster &&
+			//nolint:forbidigo // namespace state-change callback; FailoverNamespace operates per-namespace, no workflow context
 			ns.ActiveInCluster(e.currentClusterName) {
 
 			for _, queueProcessor := range e.queueProcessors {

service/history/ndc/workflow_state_replicator.go

@@ -1218,7 +1218,7 @@ func (r *WorkflowStateReplicatorImpl) bringLocalEventsUpToSourceCurrentBranch(
 	}
 	// In standby cluster, use a background low priority which is higher than the standby task processing
 	callerType := headers.CallerTypeBackgroundLow
-	if ns.ActiveInCluster(r.clusterMetadata.GetCurrentClusterName()) {
+	if ns.ActiveClusterName(namespace.RoutingKey{ID: workflowID}) == r.clusterMetadata.GetCurrentClusterName() {
 		// In active cluster, use lowest priority to minimize the impact to live traffic
 		callerType = headers.CallerTypePreemptable
 	}
@@ -1727,7 +1727,7 @@ func (r *WorkflowStateReplicatorImpl) backfillHistory(
 	}
 	// In standby cluster, use a background low priority which is higher than the standby task processing
 	callerType := headers.CallerTypeBackgroundLow
-	if ns.ActiveInCluster(r.clusterMetadata.GetCurrentClusterName()) {
+	if ns.ActiveClusterName(namespace.RoutingKey{ID: workflowID}) == r.clusterMetadata.GetCurrentClusterName() {
 		// In active cluster, use lowest priority to minimize the impact to live traffic
 		callerType = headers.CallerTypePreemptable
 	}

service/history/queues/executable.go

@@ -861,7 +861,7 @@ func taskBaseMetricTagsWithoutArchetype(
 	ns, err := namespaceRegistry.GetNamespaceByID(namespace.ID(task.GetNamespaceID()))
 	if err == nil {
 		namespaceTag = metrics.NamespaceTag(ns.Name().String())
-		isActive = ns.ActiveInCluster(currentClusterName)
+		isActive = ns.ActiveClusterName(namespace.RoutingKey{ID: task.GetWorkflowID()}) == currentClusterName
 	}
 
 	taskType := taskTypeTagProvider(task, isActive, chasmRegistry)