fix(mqtt): back off the Sparkplug primary-host claim loop after failures (#602 STAB-27)

Claude-Session: https://claude.ai/code/session_01Ef2ozxS7QPanGT8uoATbTe
2026-10-08 07:29:08 -04:00
parent 1f4f4fca8f
commit 275b240ebd
2 changed files with 99 additions and 5 deletions
@@ -238,6 +238,53 @@ public sealed class SparkplugPrimaryHostTests
        await host.DisposeAsync();
    }

+    /// <summary>
+    ///     STAB-27 — a claim the broker keeps refusing used to be redialled (and Warning-logged) on every
+    ///     one-second role sample, forever. It now backs off. The clock is frozen, so the real one-second
+    ///     loop ticks several times without the backoff window ever opening; advancing the clock opens it.
+    /// </summary>
+    [Fact]
+    public async Task A_failing_claim_backs_off_instead_of_redialling_every_tick()
+    {
+        var clock = new FrozenClock();
+        var host = new SparkplugPrimaryHost(
+            Unreachable(), "drv-1", new FixedRole(shouldActAsPrimary: true), timeProvider: clock);
+
+        await host.StartAsync(TestContext.Current.CancellationToken);
+        host.ClaimAttemptsForTests.ShouldBe(1);
+
+        // Several real loop ticks elapse while the (frozen) backoff window stays shut.
+        await Task.Delay(TimeSpan.FromMilliseconds(3200), TestContext.Current.CancellationToken);
+        host.ClaimAttemptsForTests.ShouldBe(1, "no further claim while inside the backoff window");
+
+        // Positive control: once the window passes, the very next tick does try again.
+        clock.Advance(TimeSpan.FromSeconds(5));
+        (await WaitUntilAsync(() => host.ClaimAttemptsForTests >= 2, TimeSpan.FromSeconds(10)))
+            .ShouldBeTrue("the claim must resume after the backoff window");
+
+        await host.DisposeAsync();
+    }
+
+    /// <summary>The ladder is exponential from 2 s and capped, so a long outage settles at one attempt per cap.</summary>
+    [Theory]
+    [InlineData(1, 2)]
+    [InlineData(2, 4)]
+    [InlineData(3, 8)]
+    [InlineData(4, 16)]
+    [InlineData(5, 30)]
+    [InlineData(40, 30)]
+    public void Claim_backoff_is_exponential_and_capped(int failures, int expectedSeconds)
+        => SparkplugPrimaryHost.ClaimBackoffFor(failures).ShouldBe(TimeSpan.FromSeconds(expectedSeconds));
+
+    private sealed class FrozenClock : TimeProvider
+    {
+        private DateTimeOffset _now = new(2026, 10, 8, 0, 0, 0, TimeSpan.Zero);
+
+        public override DateTimeOffset GetUtcNow() => _now;
+
+        public void Advance(TimeSpan by) => _now += by;
+    }
+
    /// <summary>A broker that refuses instantly, so a connect attempt is observable but not slow.</summary>
    private static MqttDriverOptions Unreachable() =>
        Options() with { Host = "127.0.0.1", Port = 1, ConnectTimeoutSeconds = 1 };