hive-c0re/hivectl/hive-agent: pause as a job-queue DAG node (closes #3056)

hive-c0re/src/coordinator.rs

@@ -162,6 +162,14 @@ pub struct Coordinator {
     /// agent is in this set — the inbound fence. Cleared when the agent
     /// reports `GracefulStopComplete` or the container is stopped.
     graceful_stop_pending: Mutex<HashSet<String>>,
+    /// Agents with a pause acknowledgement in progress — set by the pause
+    /// DAG's signal node, cleared when the agent reports
+    /// `PauseAcknowledged` (or the drain node's timeout fallback fires).
+    /// Same shape as `graceful_stop_pending`, one set per concern rather
+    /// than reusing it: a pause and a graceful stop are independent
+    /// orchestrations that can be in flight for different agents
+    /// simultaneously.
+    pause_pending: Mutex<HashSet<String>>,
     /// Logical agent names that were running at the last broad-scope
     /// `hivectl stop`. A subsequent broad-scope `hivectl start` restores
     /// only this set (intersected with the requested scope) rather than
@@ -582,6 +590,7 @@ impl Coordinator {
             recent_transient: Mutex::new(HashMap::new()),
             recent_crashes: Mutex::new(HashMap::new()),
             graceful_stop_pending: Mutex::new(HashSet::new()),
+            pause_pending: Mutex::new(HashSet::new()),
             last_stopped_running: Mutex::new(None),
             dashboard_events,
             event_seq: AtomicU64::new(0),
@@ -1188,6 +1197,26 @@ impl Coordinator {
         self.graceful_stop_pending.lock().unwrap().remove(name);
     }
 
+    /// Mark `name` as having a pause acknowledgement in progress. Set by
+    /// the pause DAG's signal node; the drain node polls
+    /// `is_pause_pending` until the agent reports back or the timeout
+    /// fallback fires.
+    pub fn mark_pause_pending(&self, name: &str) {
+        self.pause_pending.lock().unwrap().insert(name.to_owned());
+    }
+
+    /// Whether a pause acknowledgement is pending for `name`.
+    #[must_use]
+    pub fn is_pause_pending(&self, name: &str) -> bool {
+        self.pause_pending.lock().unwrap().contains(name)
+    }
+
+    /// Clear the pause-pending flag for `name` (agent reported
+    /// `PauseAcknowledged`, or the drain node's wait finished / timed out).
+    pub fn clear_pause_pending(&self, name: &str) {
+        self.pause_pending.lock().unwrap().remove(name);
+    }
+
     /// Record the set of agents that were running at a broad-scope
     /// `hivectl stop`, so the next broad-scope `start` restores exactly
     /// this set. See the `last_stopped_running` field doc. Persists a