Fix zombie worker tasks after missing transition (#5316)

distributed/tests/test_worker.py

@@ -2984,6 +2984,57 @@ async def test_who_has_consistent_remove_replica(c, s, *workers):
     assert s.tasks[f1.key].suspicious == 0
 
 
+@gen_cluster(client=True)
+async def test_missing_released_zombie_tasks(c, s, a, b):
+    """
+    Ensure that no fetch/flight tasks are left in the task dict of a
+    worker after everything was released
+    """
+    a.total_in_connections = 0
+    f1 = c.submit(inc, 1, key="f1", workers=[a.address])
+    f2 = c.submit(inc, f1, key="f2", workers=[b.address])
+    key = f1.key
+
+    while key not in b.tasks or b.tasks[key].state != "fetch":
+        await asyncio.sleep(0.01)
+
+    await a.close(report=False)
+
+    del f1, f2
+
+    while b.tasks:
+        await asyncio.sleep(0.01)
+
+
+@gen_cluster(client=True)
+async def test_missing_released_zombie_tasks_2(c, s, a, b):
+    a.total_in_connections = 0
+    f1 = c.submit(inc, 1, key="f1", workers=[a.address])
+    f2 = c.submit(inc, f1, key="f2", workers=[b.address])
+
+    while f1.key not in b.tasks:
+        await asyncio.sleep(0)
+
+    ts = b.tasks[f1.key]
+    assert ts.state == "fetch"
+
+    # A few things can happen to clear who_has. The dominant process is upon
+    # connection failure to a worker. Regardless of how the set was cleared, the
+    # task will be transitioned to missing where the worker is trying to
+    # reaquire this information from the scheduler. While this is happening on
+    # worker side, the tasks are released and we want to ensure that no dangling
+    # zombie tasks are left on the worker
+    ts.who_has.clear()
+
+    del f1, f2
+
+    while b.tasks:
+        await asyncio.sleep(0.01)
+
+    story = b.story(ts)
+    assert any("missing" in msg for msg in story)
+
+
 @pytest.mark.slow
 @gen_cluster(
     client=True,

distributed/worker.py

@@ -191,6 +191,7 @@ def __init__(self, key, runspec=None):
         self.who_has = set()
         self.coming_from = None
         self.waiting_for_data = set()
+        self.waiters = set()
         self.resource_restrictions = {}
         self.exception = None
         self.exception_text = ""
@@ -1854,6 +1855,7 @@ def transition_released_waiting(self, ts, *, stimulus_id):
         for dep_ts in ts.dependencies:
             if not dep_ts.state == "memory":
                 ts.waiting_for_data.add(dep_ts)
+                dep_ts.waiters.add(ts)
 
         if ts.waiting_for_data:
             self.waiting_for_data_count += 1
@@ -2669,19 +2671,6 @@ async def find_missing(self):
                 who_has = {k: v for k, v in who_has.items() if v}
                 self.update_who_has(who_has, stimulus_id=stimulus_id)
 
-                if self._missing_dep_flight:
-                    logger.debug(
-                        "No new workers found for %s", self._missing_dep_flight
-                    )
-                    recommendations = {
-                        dep: "released"
-                        for dep in self._missing_dep_flight
-                        if dep.state == "missing"
-                    }
-                    self.transitions(
-                        recommendations=recommendations, stimulus_id=stimulus_id
-                    )
-
             finally:
                 # This is quite arbitrary but the heartbeat has scaling implemented
                 self.periodic_callbacks[
@@ -2792,9 +2781,11 @@ def release_key(
                         self.available_resources[resource] += quantity
 
             for d in ts.dependencies:
-                ts.waiting_for_data.discard(ts)
-                if not d.dependents and d.state in {"flight", "fetch", "missing"}:
-                    recommendations[d] = "released"
+                ts.waiting_for_data.discard(d)
+                d.waiters.discard(ts)
+
+                if not d.waiters and d.state in {"flight", "fetch", "missing"}:
+                    recommendations[d] = "forgotten"
 
             ts.waiting_for_data.clear()
             ts.nbytes = None