provide -no-shutdown-delay flag for job/alloc stop (#11596)

Some operators use very long group/task `shutdown_delay` settings to
safely drain network connections to their workloads after service
deregistration. But during incident response, they may want to cause
that drain to be skipped so they can quickly shed load.

Provide a `-no-shutdown-delay` flag on the `nomad alloc stop` and
`nomad job stop` commands that bypasses the delay. This sets a new
desired transition state on the affected allocations that the
allocation/task runner will identify during pre-kill on the client.

Note (as documented here) that using this flag will almost always
result in failed inbound network connections for workloads as the
tasks will exit before clients receive updated service discovery
information and won't be gracefully drained.

client/allocrunner/taskrunner/task_runner.go

@@ -112,6 +112,11 @@ type TaskRunner struct {
 	killErr     error
 	killErrLock sync.Mutex
 
+	// shutdownDelayCtx is a context from the alloc runner which will
+	// tell us to exit early from shutdown_delay
+	shutdownDelayCtx      context.Context
+	shutdownDelayCancelFn context.CancelFunc
+
 	// Logger is the logger for the task runner.
 	logger log.Logger
 
@@ -287,6 +292,13 @@ type Config struct {
 
 	// startConditionMetCtx is done when TR should start the task
 	StartConditionMetCtx <-chan struct{}
+
+	// ShutdownDelayCtx is a context from the alloc runner which will
+	// tell us to exit early from shutdown_delay
+	ShutdownDelayCtx context.Context
+
+	// ShutdownDelayCancelFn should only be used in testing.
+	ShutdownDelayCancelFn context.CancelFunc
 }
 
 func NewTaskRunner(config *Config) (*TaskRunner, error) {
@@ -342,6 +354,8 @@ func NewTaskRunner(config *Config) (*TaskRunner, error) {
 		maxEvents:              defaultMaxEvents,
 		serversContactedCh:     config.ServersContactedCh,
 		startConditionMetCtx:   config.StartConditionMetCtx,
+		shutdownDelayCtx:       config.ShutdownDelayCtx,
+		shutdownDelayCancelFn:  config.ShutdownDelayCancelFn,
 	}
 
 	// Create the logger based on the allocation ID
@@ -895,6 +909,8 @@ func (tr *TaskRunner) handleKill(resultCh <-chan *drivers.ExitResult) *drivers.E
 		select {
 		case result := <-resultCh:
 			return result
+		case <-tr.shutdownDelayCtx.Done():
+			break
 		case <-time.After(delay):
 		}
 	}
@@ -1478,3 +1494,9 @@ func (tr *TaskRunner) DriverCapabilities() (*drivers.Capabilities, error) {
 func (tr *TaskRunner) SetAllocHookResources(res *cstructs.AllocHookResources) {
 	tr.allocHookResources = res
 }
+
+// shutdownDelayCancel is used for testing only and cancels the
+// shutdownDelayCtx
+func (tr *TaskRunner) shutdownDelayCancel() {
+	tr.shutdownDelayCancelFn()
+}