test(harness): widen the deploy-seal timeout to match a real two-node deploy

Comparing full-suite runs on this branch against master in a clean worktree: both fail 11 tests, 9 identical. Master's two extras are load-flaky unit tests (AbCip Probe_loops, Galaxy EventPumpBoundedChannel); this branch's two extras were both Host.IntegrationTests deploy-path tests — the area this change re-times — so they are attributable here rather than to background flakiness. Cause is margin, not correctness. These waits used to observe a coordinator that sealed instantly on an empty expected-ack set; they now observe a real ApplyAck round-trip from every configured node. 15s was enormous margin against "instant" and thin against the real thing, so they failed only under full-suite CPU contention. Hoisted to TwoNodeClusterHarness.DeploySealTimeout (45s) so the reason is recorded once rather than as four unexplained numbers. DriverReconnectE2eTests is deliberately left alone: it seeds both ClusterNode rows itself and unconditionally, so its expected-ack set is identical before and after this change. Widening its timeout would be papering over a flake this change did not cause. Host.IntegrationTests 195/201; sole failure AbCip_Green_AgainstSim, verified failing on master. Claude-Session: https://claude.ai/code/session_01GASWkNEi68FSCtvr6rLoEW
2026-07-22 09:28:32 -04:00
parent 14746f2995
commit 0b7c53f64f
5 changed files with 18 additions and 8 deletions
@@ -40,7 +40,7 @@ public sealed class DeployHappyPathTests
            await using var db = await CreateDbAsync(harness);
            var d = await db.Deployments.AsNoTracking().FirstOrDefaultAsync(d => d.DeploymentId == deploymentId, Ct);
            return d?.Status == DeploymentStatus.Sealed;
-        }, TimeSpan.FromSeconds(15));
+        }, TwoNodeClusterHarness.DeploySealTimeout);

        await using var db = await CreateDbAsync(harness);
        var deployment = await db.Deployments.AsNoTracking()
@@ -74,7 +74,7 @@ public sealed class DeployHappyPathTests
            var d = await db.Deployments.AsNoTracking()
                .FirstOrDefaultAsync(d => d.DeploymentId == first.DeploymentId!.Value.Value, Ct);
            return d?.Status == DeploymentStatus.Sealed;
-        }, TimeSpan.FromSeconds(15));
+        }, TwoNodeClusterHarness.DeploySealTimeout);

        // Same DB state → same revision hash → AdminOperations short-circuits with NoChanges,
        // OR (if it accepts) the second dispatch is idempotent (DriverHostActor recognises
@@ -90,7 +90,7 @@ public sealed class DeployHappyPathTests
                var d = await db.Deployments.AsNoTracking()
                    .FirstOrDefaultAsync(d => d.DeploymentId == second.DeploymentId!.Value.Value, Ct);
                return d?.Status == DeploymentStatus.Sealed;
-            }, TimeSpan.FromSeconds(15));
+            }, TwoNodeClusterHarness.DeploySealTimeout);

            second.RevisionHash!.Value.Value.ShouldBe(first.RevisionHash!.Value.Value);
        }
@@ -85,7 +85,7 @@ public sealed class EquipmentNamespaceMaterializationTests
                return true;
            }
            return false;
-        }, TimeSpan.FromSeconds(15));
+        }, TwoNodeClusterHarness.DeploySealTimeout);

        var composition = DeploymentArtifact.ParseComposition(artifact);

@@ -182,7 +182,7 @@ public sealed class EquipmentNamespaceMaterializationTests
                return true;
            }
            return false;
-        }, TimeSpan.FromSeconds(20));
+        }, TwoNodeClusterHarness.DeploySealTimeout);

        await using (var db = await CreateDbAsync(harness))
        {
@@ -87,7 +87,7 @@ public sealed class FailoverDuringDeployTests
            await using var pollDb = await CreateDbAsync(harness);
            return await pollDb.NodeDeploymentStates.AsNoTracking()
                .CountAsync(s => s.DeploymentId == deploymentId, Ct) == 2;
-        }, TimeSpan.FromSeconds(15));
+        }, TwoNodeClusterHarness.DeploySealTimeout);

        await using var db = await CreateDbAsync(harness);
        var nodeStates = await db.NodeDeploymentStates.AsNoTracking()
@@ -143,7 +143,7 @@ public sealed class FailoverDuringDeployTests
            var d = await pollDb.Deployments.AsNoTracking()
                .FirstOrDefaultAsync(d => d.DeploymentId == deploymentId, Ct);
            return d?.Status == DeploymentStatus.Sealed;
-        }, TimeSpan.FromSeconds(15));
+        }, TwoNodeClusterHarness.DeploySealTimeout);

        await using var db = await CreateDbAsync(harness);
        var nodeStates = await db.NodeDeploymentStates.AsNoTracking()
@@ -67,7 +67,7 @@ public sealed class FleetDiagnosticsRoundTripTests
            {
                var snap = await diagnostics.GetDiagnosticsAsync(nodeId, Ct);
                return snap.CurrentRevision == expectedRev;
-            }, TimeSpan.FromSeconds(15));
+            }, TwoNodeClusterHarness.DeploySealTimeout);
        }
    }

@@ -96,6 +96,16 @@ public sealed class TwoNodeClusterHarness : IAsyncDisposable
    // host:port into NodeId so the cluster membership stays distinct on different ports.
    public const string LoopbackHost = "127.0.0.1";

+    /// <summary>
+    ///     How long a deploy-seal assertion waits. Generous on purpose: since per-cluster mesh
+    ///     Phase 1 the coordinator waits for a real ApplyAck from every configured node before
+    ///     sealing, so this measures an end-to-end two-node deploy rather than — as it did when the
+    ///     expected-ack set came up empty — an immediate seal. The old 15s had enormous margin
+    ///     against "instant" and thin margin against the real thing, which showed up as deploy
+    ///     tests failing only under full-suite CPU contention.
+    /// </summary>
+    public static readonly TimeSpan DeploySealTimeout = TimeSpan.FromSeconds(45);
+
    /// <summary>Gets the Akka ActorSystem for node A.</summary>
    public ActorSystem NodeASystem => NodeA.Services.GetRequiredService<ActorSystem>();
    /// <summary>Gets the Akka ActorSystem for node B.</summary>