scheduler: retain eval metrics on port collision

.changelog/19933.txt

@@ -0,0 +1,3 @@
+```release-note:bug
+scheduler: Fixed a bug that caused blocked evaluations due to port conflict to not have a reason explaining why the evaluation was blocked
+```

nomad/job_endpoint_test.go

@@ -2489,6 +2489,90 @@ func TestJobRegister_ACL_RejectedBySchedulerConfig(t *testing.T) {
 	}
 }
 
+func TestJobEndpoint_Register_PortCollistion(t *testing.T) {
+	ci.Parallel(t)
+
+	testCases := []struct {
+		name     string
+		configFn func(c *Config)
+	}{
+		{
+			name: "no preemption",
+			configFn: func(c *Config) {
+				c.DefaultSchedulerConfig = structs.SchedulerConfiguration{
+					PreemptionConfig: structs.PreemptionConfig{
+						ServiceSchedulerEnabled: false,
+					},
+				}
+			},
+		},
+		{
+			name: "with preemption",
+			configFn: func(c *Config) {
+				c.DefaultSchedulerConfig = structs.SchedulerConfiguration{
+					PreemptionConfig: structs.PreemptionConfig{
+						ServiceSchedulerEnabled: true,
+					},
+				}
+			},
+		},
+	}
+
+	for _, tc := range testCases {
+		t.Run(tc.name, func(t *testing.T) {
+			s1, cleanupS1 := TestServer(t, tc.configFn)
+			defer cleanupS1()
+			state := s1.fsm.State()
+
+			// Create test node.
+			node := mock.Node()
+			must.NoError(t, state.UpsertNode(structs.MsgTypeTestSetup, 1000, node))
+
+			// Create test job with a static port.
+			job := mock.Job()
+			job.TaskGroups[0].Count = 1
+			job.TaskGroups[0].Networks[0].DynamicPorts = nil
+			job.TaskGroups[0].Networks[0].ReservedPorts = []structs.Port{
+				{Label: "http", Value: 80},
+			}
+			job.TaskGroups[0].Tasks[0].Services = nil
+
+			testutil.RegisterJob(t, s1.RPC, job)
+			testutil.WaitForJobAllocStatus(t, s1.RPC, job, map[string]int{
+				structs.AllocClientStatusPending: 1,
+			})
+
+			// Register second job with port conflict.
+			job2 := job.Copy()
+			job2.ID = fmt.Sprintf("conflict-%s", uuid.Generate())
+			job2.Name = job2.ID
+
+			testutil.RegisterJob(t, s1.RPC, job2)
+
+			// Wait for job registration eval to complete.
+			evals := testutil.WaitForJobEvalStatus(t, s1.RPC, job2, map[string]int{
+				structs.EvalStatusComplete: 1,
+				structs.EvalStatusBlocked:  1,
+			})
+
+			var blockedEval *structs.Evaluation
+			for _, e := range evals {
+				if e.Status == structs.EvalStatusBlocked {
+					blockedEval = e
+					break
+				}
+			}
+
+			// Ensure blocked eval is properly annotated.
+			must.MapLen(t, 1, blockedEval.FailedTGAllocs)
+			must.NotNil(t, blockedEval.FailedTGAllocs["web"])
+			must.Eq(t, map[string]int{
+				"network: reserved port collision http=80": 1,
+			}, blockedEval.FailedTGAllocs["web"].DimensionExhausted)
+		})
+	}
+}
+
 func TestJobEndpoint_Revert(t *testing.T) {
 	ci.Parallel(t)
 

scheduler/rank.go

@@ -322,6 +322,8 @@ OUTER:
 				netPreemptions := preemptor.PreemptForNetwork(ask, netIdx)
 				if netPreemptions == nil {
 					iter.ctx.Logger().Named("binpack").Debug("preemption not possible ", "network_resource", ask)
+					iter.ctx.Metrics().ExhaustedNode(option.Node,
+						fmt.Sprintf("network: %s", err))
 					netIdx.Release()
 					continue OUTER
 				}
@@ -339,6 +341,8 @@ OUTER:
 				offer, err = netIdx.AssignPorts(ask)
 				if err != nil {
 					iter.ctx.Logger().Named("binpack").Debug("unexpected error, unable to create network offer after considering preemption", "error", err)
+					iter.ctx.Metrics().ExhaustedNode(option.Node,
+						fmt.Sprintf("network: %s", err))
 					netIdx.Release()
 					continue OUTER
 				}
@@ -392,6 +396,8 @@ OUTER:
 					netPreemptions := preemptor.PreemptForNetwork(ask, netIdx)
 					if netPreemptions == nil {
 						iter.ctx.Logger().Named("binpack").Debug("preemption not possible ", "network_resource", ask)
+						iter.ctx.Metrics().ExhaustedNode(option.Node,
+							fmt.Sprintf("network: %s", err))
 						netIdx.Release()
 						continue OUTER
 					}
@@ -409,6 +415,8 @@ OUTER:
 					offer, err = netIdx.AssignTaskNetwork(ask)
 					if offer == nil {
 						iter.ctx.Logger().Named("binpack").Debug("unexpected error, unable to create network offer after considering preemption", "error", err)
+						iter.ctx.Metrics().ExhaustedNode(option.Node,
+							fmt.Sprintf("network: %s", err))
 						netIdx.Release()
 						continue OUTER
 					}

testutil/wait.go

@@ -377,6 +377,57 @@ func WaitForJobAllocStatusWithToken(t testing.TB, rpc rpcFn, job *structs.Job, a
 	return allocs
 }
 
+// WaitforJobEvalStatus blocks until the job's evals match the status described
+// in the map of <Eval.Status>: <count>.
+func WaitForJobEvalStatus(t testing.TB, rpc rpcFn, job *structs.Job, evalStatus map[string]int) []*structs.Evaluation {
+	return WaitForJobEvalStatusWithToken(t, rpc, job, evalStatus, "")
+}
+
+// WaitForJobEvalStatusWithToken is the same as WaitforJobEvalStatus  with ACL
+// enabled.
+func WaitForJobEvalStatusWithToken(t testing.TB, rpc rpcFn, job *structs.Job, evalStatus map[string]int, token string) []*structs.Evaluation {
+	var evals []*structs.Evaluation
+
+	errorFunc := func() error {
+		req := &structs.JobSpecificRequest{
+			JobID: job.ID,
+			QueryOptions: structs.QueryOptions{
+				AuthToken: token,
+				Namespace: job.Namespace,
+				Region:    job.Region,
+			},
+		}
+		var resp structs.JobEvaluationsResponse
+		err := rpc("Job.Evaluations", req, &resp)
+		if err != nil {
+			return fmt.Errorf("failed to call Job.Evaluations RPC: %w", err)
+		}
+
+		got := make(map[string]int)
+		for _, eval := range resp.Evaluations {
+			got[eval.Status]++
+		}
+
+		if diff := cmp.Diff(evalStatus, got); diff != "" {
+			return fmt.Errorf("eval status mismatch (-want +got):\n%s", diff)
+		}
+
+		evals = resp.Evaluations
+		return nil
+	}
+
+	must.Wait(t,
+		wait.InitialSuccess(
+			wait.ErrorFunc(errorFunc),
+			wait.Timeout(time.Duration(TestMultiplier())*time.Second),
+			wait.Gap(10*time.Millisecond),
+		),
+		must.Sprintf("failed to wait for job %s eval status", job.ID),
+	)
+
+	return evals
+}
+
 // WaitForFiles blocks until all the files in the slice are present
 func WaitForFiles(t testing.TB, files []string) {
 	WaitForResult(func() (bool, error) {