fix(host_agent): a stalled NFS mount no longer freezes the agent (#5760)

#5760 "One stalled NFS mount freezes the whole agent: statvfs has no
deadline, so the host goes silent while it still answers ping". The agent
read every mount inline in its only loop; a hard-mounted share whose server
stops blocks statvfs in the kernel with no timeout. Each mount is now read
in a daemon thread under one 5 s deadline. One that misses it is sent as
stalled, and isn't probed again while its last probe is still blocked. A
mount that errors (a stale handle) is left out instead of emptying the list.

Steward counts them as mounts_stalled (alertable: "an NFS share stopped"),
and skips the fullest-disk figure when every mount stalled.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

tests/plugins/host_agent/test_agent_collectors.py

@@ -109,3 +109,45 @@ def test_collect_metadata(tmp_path):
     assert meta["kernel"] == "6.8.0-45-generic"
     assert meta["arch"] == "x86_64"
     assert "Ubuntu 24.04" in meta["distro"]
+
+
+def test_a_mount_that_does_not_answer_is_reported_stalled_and_not_probed_again():
+    """#5760: statvfs on a stalled NFS share blocks with no timeout, and read
+    inline that froze the whole agent. The sample goes on without it, and a
+    probe still blocked isn't stacked on by the next sample."""
+    import threading
+    release = threading.Event()
+    calls: list[str] = []
+
+    def _usage(path):
+        calls.append(path)
+        if path == "/nfs/media":
+            release.wait(5)
+        return type("R", (), {"total": 1000, "used": 400, "free": 600})()
+
+    a._stuck_mounts.clear()
+    try:
+        with patch.object(a.shutil, "disk_usage", side_effect=_usage), \
+                patch.object(a.os, "statvfs", side_effect=OSError):
+            first = a.collect_storage(["/", "/nfs/media"], deadline=0.2)
+            second = a.collect_storage(["/", "/nfs/media"], deadline=0.2)
+        assert first == second == [
+            {"mount": "/", "total_bytes": 1000, "used_bytes": 400},
+            {"mount": "/nfs/media", "stalled": True},
+        ]
+        assert calls.count("/nfs/media") == 1
+    finally:
+        release.set()
+        a._stuck_mounts.clear()
+
+
+def test_a_mount_that_errors_is_left_out_not_the_whole_list():
+    def _usage(path):
+        if path == "/nfs/gone":
+            raise OSError(116, "Stale file handle")
+        return type("R", (), {"total": 10, "used": 1, "free": 9})()
+
+    with patch.object(a.shutil, "disk_usage", side_effect=_usage), \
+            patch.object(a.os, "statvfs", side_effect=OSError):
+        out = a.collect_storage(["/", "/nfs/gone"])
+    assert out == [{"mount": "/", "total_bytes": 10, "used_bytes": 1}]

tests/plugins/host_agent/test_expand_metrics.py

@@ -93,3 +93,14 @@ def test_a_sensor_with_nothing_wired_to_it_does_not_set_the_host_s_temperature()
     assert idx[("temp_c_max", "util")] == 60.0
     assert idx[("temp_c", "util:temp:AUXTIN0")] == 110.0
     assert idx[("temp_c", "util:temp:PCH_CHIP_CPU_MAX_TEMP")] == 0.0
+
+
+def test_stalled_mounts_are_counted_and_have_no_figures():
+    """#5760: a mount the agent couldn't read in time."""
+    sample = {"ts": "2026-10-14T21:00:00+00:00",
+              "storage": [{"mount": "/", "total_bytes": 100, "used_bytes": 50},
+                          {"mount": "/nfs/media", "stalled": True}]}
+    idx = _index(_expand_sample_to_metrics(sample, "d3", datetime.now(timezone.utc)))
+    assert idx[("mounts_stalled", "d3")] == 1
+    assert idx[("disk_used_pct_worst", "d3")] == 50.0
+    assert ("disk_used_pct", "d3:/nfs/media") not in idx