Restart unhealthy tasks

CHANGELOG.md

@@ -2,6 +2,7 @@
 
 IMPROVEMENTS:
  * api: Metrics endpoint exposes Prometheus formatted metrics [GH-3171]
+ * discovery: Allow restarting unhealthy tasks with `check_restart` [GH-3105]
  * telemetry: Add support for tagged metrics for Nomad clients [GH-3147]
 
 BUG FIXES:

api/tasks.go

@@ -79,6 +79,71 @@ func (r *RestartPolicy) Merge(rp *RestartPolicy) {
 	}
 }
 
+// CheckRestart describes if and when a task should be restarted based on
+// failing health checks.
+type CheckRestart struct {
+	Limit          int            `mapstructure:"limit"`
+	Grace          *time.Duration `mapstructure:"grace_period"`
+	IgnoreWarnings bool           `mapstructure:"ignore_warnings"`
+}
+
+// Canonicalize CheckRestart fields if not nil.
+func (c *CheckRestart) Canonicalize() {
+	if c == nil {
+		return
+	}
+
+	if c.Grace == nil {
+		c.Grace = helper.TimeToPtr(1 * time.Second)
+	}
+}
+
+// Copy returns a copy of CheckRestart or nil if unset.
+func (c *CheckRestart) Copy() *CheckRestart {
+	if c == nil {
+		return nil
+	}
+
+	nc := new(CheckRestart)
+	nc.Limit = c.Limit
+	if c.Grace != nil {
+		g := *c.Grace
+		nc.Grace = &g
+	}
+	nc.IgnoreWarnings = c.IgnoreWarnings
+	return nc
+}
+
+// Merge values from other CheckRestart over default values on this
+// CheckRestart and return merged copy.
+func (c *CheckRestart) Merge(o *CheckRestart) *CheckRestart {
+	if c == nil {
+		// Just return other
+		return o
+	}
+
+	nc := c.Copy()
+
+	if o == nil {
+		// Nothing to merge
+		return nc
+	}
+
+	if nc.Limit == 0 {
+		nc.Limit = o.Limit
+	}
+
+	if nc.Grace == nil {
+		nc.Grace = o.Grace
+	}
+
+	if nc.IgnoreWarnings {
+		nc.IgnoreWarnings = o.IgnoreWarnings
+	}
+
+	return nc
+}
+
 // The ServiceCheck data model represents the consul health check that
 // Nomad registers for a Task
 type ServiceCheck struct {
@@ -96,16 +161,18 @@ type ServiceCheck struct {
 	TLSSkipVerify bool   `mapstructure:"tls_skip_verify"`
 	Header        map[string][]string
 	Method        string
+	CheckRestart  *CheckRestart `mapstructure:"check_restart"`
 }
 
 // The Service model represents a Consul service definition
 type Service struct {
-	Id          string
-	Name        string
-	Tags        []string
-	PortLabel   string `mapstructure:"port"`
-	AddressMode string `mapstructure:"address_mode"`
-	Checks      []ServiceCheck
+	Id           string
+	Name         string
+	Tags         []string
+	PortLabel    string `mapstructure:"port"`
+	AddressMode  string `mapstructure:"address_mode"`
+	Checks       []ServiceCheck
+	CheckRestart *CheckRestart `mapstructure:"check_restart"`
 }
 
 func (s *Service) Canonicalize(t *Task, tg *TaskGroup, job *Job) {
@@ -117,6 +184,13 @@ func (s *Service) Canonicalize(t *Task, tg *TaskGroup, job *Job) {
 	if s.AddressMode == "" {
 		s.AddressMode = "auto"
 	}
+
+	s.CheckRestart.Canonicalize()
+
+	for _, c := range s.Checks {
+		c.CheckRestart.Canonicalize()
+		c.CheckRestart = c.CheckRestart.Merge(s.CheckRestart)
+	}
 }
 
 // EphemeralDisk is an ephemeral disk object

client/alloc_runner.go

@@ -336,7 +336,8 @@ func (r *AllocRunner) RestoreState() error {
 			// Restart task runner if RestoreState gave a reason
 			if restartReason != "" {
 				r.logger.Printf("[INFO] client: restarting alloc %s task %s: %v", r.allocID, name, restartReason)
-				tr.Restart("upgrade", restartReason)
+				const failure = false
+				tr.Restart("upgrade", restartReason, failure)
 			}
 		} else {
 			tr.Destroy(taskDestroyEvent)

client/consul.go

@@ -10,8 +10,8 @@ import (
 // ConsulServiceAPI is the interface the Nomad Client uses to register and
 // remove services and checks from Consul.
 type ConsulServiceAPI interface {
-	RegisterTask(allocID string, task *structs.Task, exec driver.ScriptExecutor, net *cstructs.DriverNetwork) error
+	RegisterTask(allocID string, task *structs.Task, restarter consul.TaskRestarter, exec driver.ScriptExecutor, net *cstructs.DriverNetwork) error
 	RemoveTask(allocID string, task *structs.Task)
-	UpdateTask(allocID string, existing, newTask *structs.Task, exec driver.ScriptExecutor, net *cstructs.DriverNetwork) error
+	UpdateTask(allocID string, existing, newTask *structs.Task, restart consul.TaskRestarter, exec driver.ScriptExecutor, net *cstructs.DriverNetwork) error
 	AllocRegistrations(allocID string) (*consul.AllocRegistration, error)
 }

client/consul_template.go

@@ -49,7 +49,7 @@ var (
 // TaskHooks is an interface which provides hooks into the tasks life-cycle
 type TaskHooks interface {
 	// Restart is used to restart the task
-	Restart(source, reason string)
+	Restart(source, reason string, failure bool)
 
 	// Signal is used to signal the task
 	Signal(source, reason string, s os.Signal) error
@@ -439,7 +439,8 @@ func (tm *TaskTemplateManager) handleTemplateRerenders(allRenderedTime time.Time
 				}
 
 				if restart {
-					tm.config.Hooks.Restart(consulTemplateSourceName, "template with change_mode restart re-rendered")
+					const failure = false
+					tm.config.Hooks.Restart(consulTemplateSourceName, "template with change_mode restart re-rendered", failure)
 				} else if len(signals) != 0 {
 					var mErr multierror.Error
 					for signal := range signals {

client/consul_template_test.go

@@ -57,7 +57,7 @@ func NewMockTaskHooks() *MockTaskHooks {
 		EmitEventCh: make(chan struct{}, 1),
 	}
 }
-func (m *MockTaskHooks) Restart(source, reason string) {
+func (m *MockTaskHooks) Restart(source, reason string, failure bool) {
 	m.Restarts++
 	select {
 	case m.RestartCh <- struct{}{}:

client/consul_test.go

@@ -60,15 +60,15 @@ func newMockConsulServiceClient() *mockConsulServiceClient {
 	return &m
 }
 
-func (m *mockConsulServiceClient) UpdateTask(allocID string, old, new *structs.Task, exec driver.ScriptExecutor, net *cstructs.DriverNetwork) error {
+func (m *mockConsulServiceClient) UpdateTask(allocID string, old, new *structs.Task, restarter consul.TaskRestarter, exec driver.ScriptExecutor, net *cstructs.DriverNetwork) error {
 	m.mu.Lock()
 	defer m.mu.Unlock()
 	m.logger.Printf("[TEST] mock_consul: UpdateTask(%q, %v, %v, %T, %x)", allocID, old, new, exec, net.Hash())
 	m.ops = append(m.ops, newMockConsulOp("update", allocID, new, exec, net))
 	return nil
 }
 
-func (m *mockConsulServiceClient) RegisterTask(allocID string, task *structs.Task, exec driver.ScriptExecutor, net *cstructs.DriverNetwork) error {
+func (m *mockConsulServiceClient) RegisterTask(allocID string, task *structs.Task, restarter consul.TaskRestarter, exec driver.ScriptExecutor, net *cstructs.DriverNetwork) error {
 	m.mu.Lock()
 	defer m.mu.Unlock()
 	m.logger.Printf("[TEST] mock_consul: RegisterTask(%q, %q, %T, %x)", allocID, task.Name, exec, net.Hash())

client/restarts.go

@@ -37,6 +37,7 @@ type RestartTracker struct {
 	waitRes          *dstructs.WaitResult
 	startErr         error
 	restartTriggered bool      // Whether the task has been signalled to be restarted
+	failure          bool      // Whether a failure triggered the restart
 	count            int       // Current number of attempts.
 	onSuccess        bool      // Whether to restart on successful exit code.
 	startTime        time.Time // When the interval began
@@ -59,6 +60,7 @@ func (r *RestartTracker) SetStartError(err error) *RestartTracker {
 	r.lock.Lock()
 	defer r.lock.Unlock()
 	r.startErr = err
+	r.failure = true
 	return r
 }
 
@@ -67,15 +69,20 @@ func (r *RestartTracker) SetWaitResult(res *dstructs.WaitResult) *RestartTracker
 	r.lock.Lock()
 	defer r.lock.Unlock()
 	r.waitRes = res
+	r.failure = true
 	return r
 }
 
 // SetRestartTriggered is used to mark that the task has been signalled to be
 // restarted
-func (r *RestartTracker) SetRestartTriggered() *RestartTracker {
+func (r *RestartTracker) SetRestartTriggered(failure bool) *RestartTracker {
 	r.lock.Lock()
 	defer r.lock.Unlock()
-	r.restartTriggered = true
+	if failure {
+		r.failure = true
+	} else {
+		r.restartTriggered = true
+	}
 	return r
 }
 
@@ -106,6 +113,7 @@ func (r *RestartTracker) GetState() (string, time.Duration) {
 		r.startErr = nil
 		r.waitRes = nil
 		r.restartTriggered = false
+		r.failure = false
 	}()
 
 	// Hot path if a restart was triggered
@@ -134,66 +142,40 @@ func (r *RestartTracker) GetState() (string, time.Duration) {
 		r.startTime = now
 	}
 
-	if r.startErr != nil {
-		return r.handleStartError()
-	} else if r.waitRes != nil {
-		return r.handleWaitResult()
-	}
-
-	return "", 0
-}
-
-// handleStartError returns the new state and potential wait duration for
-// restarting the task after it was not successfully started. On start errors,
-// the restart policy is always treated as fail mode to ensure we don't
-// infinitely try to start a task.
-func (r *RestartTracker) handleStartError() (string, time.Duration) {
-	// If the error is not recoverable, do not restart.
-	if !structs.IsRecoverable(r.startErr) {
-		r.reason = ReasonUnrecoverableErrror
-		return structs.TaskNotRestarting, 0
-	}
-
-	if r.count > r.policy.Attempts {
-		if r.policy.Mode == structs.RestartPolicyModeFail {
-			r.reason = fmt.Sprintf(
-				`Exceeded allowed attempts %d in interval %v and mode is "fail"`,
-				r.policy.Attempts, r.policy.Interval)
-			return structs.TaskNotRestarting, 0
-		} else {
-			r.reason = ReasonDelay
-			return structs.TaskRestarting, r.getDelay()
+	// Handle restarts due to failures
+	if r.failure {
+		if r.startErr != nil {
+			// If the error is not recoverable, do not restart.
+			if !structs.IsRecoverable(r.startErr) {
+				r.reason = ReasonUnrecoverableErrror
+				return structs.TaskNotRestarting, 0
+			}
+		} else if r.waitRes != nil {
+			// If the task started successfully and restart on success isn't specified,
+			// don't restart but don't mark as failed.
+			if r.waitRes.Successful() && !r.onSuccess {
+				r.reason = "Restart unnecessary as task terminated successfully"
+				return structs.TaskTerminated, 0
+			}
 		}
-	}
-
-	r.reason = ReasonWithinPolicy
-	return structs.TaskRestarting, r.jitter()
-}
 
-// handleWaitResult returns the new state and potential wait duration for
-// restarting the task after it has exited.
-func (r *RestartTracker) handleWaitResult() (string, time.Duration) {
-	// If the task started successfully and restart on success isn't specified,
-	// don't restart but don't mark as failed.
-	if r.waitRes.Successful() && !r.onSuccess {
-		r.reason = "Restart unnecessary as task terminated successfully"
-		return structs.TaskTerminated, 0
-	}
-
-	if r.count > r.policy.Attempts {
-		if r.policy.Mode == structs.RestartPolicyModeFail {
-			r.reason = fmt.Sprintf(
-				`Exceeded allowed attempts %d in interval %v and mode is "fail"`,
-				r.policy.Attempts, r.policy.Interval)
-			return structs.TaskNotRestarting, 0
-		} else {
-			r.reason = ReasonDelay
-			return structs.TaskRestarting, r.getDelay()
+		if r.count > r.policy.Attempts {
+			if r.policy.Mode == structs.RestartPolicyModeFail {
+				r.reason = fmt.Sprintf(
+					`Exceeded allowed attempts %d in interval %v and mode is "fail"`,
+					r.policy.Attempts, r.policy.Interval)
+				return structs.TaskNotRestarting, 0
+			} else {
+				r.reason = ReasonDelay
+				return structs.TaskRestarting, r.getDelay()
+			}
 		}
+
+		r.reason = ReasonWithinPolicy
+		return structs.TaskRestarting, r.jitter()
 	}
 
-	r.reason = ReasonWithinPolicy
-	return structs.TaskRestarting, r.jitter()
+	return "", 0
 }
 
 // getDelay returns the delay time to enter the next interval.

client/restarts_test.go

@@ -99,7 +99,7 @@ func TestClient_RestartTracker_RestartTriggered(t *testing.T) {
 	p := testPolicy(true, structs.RestartPolicyModeFail)
 	p.Attempts = 0
 	rt := newRestartTracker(p, structs.JobTypeService)
-	if state, when := rt.SetRestartTriggered().GetState(); state != structs.TaskRestarting && when != 0 {
+	if state, when := rt.SetRestartTriggered(false).GetState(); state != structs.TaskRestarting && when != 0 {
 		t.Fatalf("expect restart immediately, got %v %v", state, when)
 	}
 }