Health summary: a node is failing after two failures in a row; one is a warning, a good run ends both

The summary read degraded for 15 minutes after any single node_error, even
once the node had run well again and including one bad input from a visitor.
The controller now counts a node's failures since it last ran well and notes
when it recovered; both carry across rebuilds and an acknowledgement resets
the count. Two in a row inside the 15 minute window is "failing" (a problem),
one is listed under warnings, and nodes.failing keeps meaning "failing now".

backend/tests/flow/test_node_failures.py

@@ -0,0 +1,58 @@
+"""A node's failures in a row, which the health summary reads."""
+
+import asyncio
+from pathlib import Path
+
+from fluksio.flow.controller import FlowController, LoadedNode
+from fluksio.flow.events import EventBus
+from fluksio.flow.store import FlowStore
+
+NODE = "house.calc"
+
+
+def _failure(ts: float) -> dict:
+    return {"type": "node_error", "node": NODE, "error": "boom", "ts": ts}
+
+
+def test_failures_count_until_the_node_runs_well(tmp_path: Path) -> None:
+    async def scenario() -> LoadedNode:
+        events = EventBus()
+        events.bind(asyncio.get_running_loop())
+        controller = FlowController(FlowStore(tmp_path / "flows"), events=events)
+        controller.loaded[NODE] = LoadedNode(id=NODE, flow="house")
+        watcher = asyncio.create_task(controller._watch_failures())
+        await asyncio.sleep(0)
+
+        events.publish(_failure(1.0))
+        events.publish(_failure(2.0))
+        await asyncio.sleep(0.05)
+        assert controller.loaded[NODE].failures_in_a_row == 2
+
+        events.publish({"type": "node_executed", "node": NODE, "ts": 3.0})
+        await asyncio.sleep(0.05)
+        watcher.cancel()
+        return controller.loaded[NODE]
+
+    entry = asyncio.run(scenario())
+
+    assert entry.failures_in_a_row == 0
+    assert entry.last_ok_ts == 3.0
+    # What it failed with stays for the canvas until someone acknowledges it.
+    assert entry.last_error == "boom"
+
+
+def test_a_rebuilt_node_keeps_its_failures() -> None:
+    old = LoadedNode(
+        id=NODE,
+        flow="house",
+        last_error="boom",
+        last_error_ts=5.0,
+        failures_in_a_row=3,
+        last_ok_ts=1.0,
+    )
+    new = LoadedNode(id=NODE, flow="house")
+
+    new.inherit_failure(old)
+
+    assert (new.last_error, new.last_error_ts) == ("boom", 5.0)
+    assert (new.failures_in_a_row, new.last_ok_ts) == (3, 1.0)