Merge branch 'job-service' of https://github.com/vmware/harbor into job-service

api/jobs/replication.go

@@ -64,6 +64,33 @@ func (rj *ReplicationJob) Post() {
 	}
 }
 
+type RepActionReq struct {
+	PolicyID int64  `json:"policy_id"`
+	Action   string `json:"action"`
+}
+
+func (rj *ReplicationJob) HandleAction() {
+	var data RepActionReq
+	rj.DecodeJSONReq(&data)
+	//Currently only support stop action
+	if data.Action != "stop" {
+		log.Errorf("Unrecognized action: %s", data.Action)
+		rj.RenderError(http.StatusBadRequest, fmt.Sprintf("Unrecongized action: %s", data.Action))
+		return
+	}
+	jobs, err := dao.GetRepJobToStop(data.PolicyID)
+	if err != nil {
+		log.Errorf("Failed to get jobs to stop, error: %v", err)
+		rj.RenderError(http.StatusInternalServerError, "Faild to get jobs to stop")
+		return
+	}
+	var jobIDList []int64
+	for _, j := range jobs {
+		jobIDList = append(jobIDList, j.ID)
+	}
+	job.WorkerPool.StopJobs(jobIDList)
+}
+
 // calls the api from UI to get repo list
 func getRepoList(projectID int64) ([]string, error) {
 	uiURL := os.Getenv("UI_URL")

dao/base.go

@@ -64,6 +64,7 @@ func GenerateRandomString() (string, error) {
 
 //InitDB initializes the database
 func InitDB() {
+	//	orm.Debug = true
 	orm.RegisterDriver("mysql", orm.DRMySQL)
 	addr := os.Getenv("MYSQL_HOST")
 	port := os.Getenv("MYSQL_PORT")

dao/dao_test.go

@@ -25,13 +25,13 @@ import (
 	"github.com/vmware/harbor/utils/log"
 )
 
-func execUpdate(o orm.Ormer, sql string, params interface{}) error {
+func execUpdate(o orm.Ormer, sql string, params ...interface{}) error {
 	p, err := o.Raw(sql).Prepare()
 	if err != nil {
 		return err
 	}
 	defer p.Close()
-	_, err = p.Exec(params)
+	_, err = p.Exec(params...)
 	if err != nil {
 		return err
 	}
@@ -95,6 +95,19 @@ func clearUp(username string) {
 		o.Rollback()
 		log.Error(err)
 	}
+
+	err = execUpdate(o, `delete from replication_job where id < 99`)
+	if err != nil {
+		log.Error(err)
+	}
+	err = execUpdate(o, `delete from replication_policy where id < 99`)
+	if err != nil {
+		log.Error(err)
+	}
+	err = execUpdate(o, `delete from replication_target where id < 99`)
+	if err != nil {
+		log.Error(err)
+	}
 	o.Commit()
 }
 
@@ -856,19 +869,23 @@ func TestAddRepJob(t *testing.T) {
 	id, err := AddRepJob(job)
 	if err != nil {
 		t.Errorf("Error occurred in AddRepJob: %v", err)
+		return
 	} else {
 		jobID = id
 	}
 	j, err := GetRepJob(id)
 	if err != nil {
 		t.Errorf("Error occurred in GetRepJob: %v, id: %d", err, id)
+		return
 	}
 	if j == nil {
 		t.Errorf("Unable to find a job with id: %d", id)
+		return
 	}
 	if j.Status != models.JobPending || j.Repository != "library/ubuntu" || j.PolicyID != policyID || j.Operation != "transfer" {
 		t.Errorf("Expected data of job, id: %d, Status: %s, Repository: library/ubuntu, PolicyID: %d, Operation: transfer, "+
 			"but in returned data:, Status: %s, Repository: %s, Operation: %s, PolicyID: %d", id, models.JobPending, policyID, j.Status, j.Repository, j.Operation, j.PolicyID)
+		return
 	}
 }
 
@@ -888,6 +905,11 @@ func TestUpdateRepJobStatus(t *testing.T) {
 	if j.Status != models.JobFinished {
 		t.Errorf("Job's status: %s, expected: %s, id: %d", j.Status, models.JobFinished, jobID)
 	}
+	err = UpdateRepJobStatus(jobID, models.JobPending)
+	if err != nil {
+		t.Errorf("Error occured in UpdateRepJobStatus when update it back to status pending, error: %v, id: %d", err, jobID)
+		return
+	}
 }
 
 func TestGetRepPolicyByProject(t *testing.T) {
@@ -941,6 +963,47 @@ func TestGetRepJobByPolicy(t *testing.T) {
 	}
 }
 
+func TestGetRepoJobToStop(t *testing.T) {
+	jobs := [...]models.RepJob{
+		models.RepJob{
+			Repository: "library/ubuntu",
+			PolicyID:   policyID,
+			Operation:  "transfer",
+			Status:     models.JobRunning,
+		},
+		models.RepJob{
+			Repository: "library/ubuntu",
+			PolicyID:   policyID,
+			Operation:  "transfer",
+			Status:     models.JobFinished,
+		},
+		models.RepJob{
+			Repository: "library/ubuntu",
+			PolicyID:   policyID,
+			Operation:  "transfer",
+			Status:     models.JobCanceled,
+		},
+	}
+	var err error
+	for _, j := range jobs {
+		_, err = AddRepJob(j)
+		if err != nil {
+			log.Errorf("Failed to add Job: %+v, error: %v", j, err)
+			return
+		}
+	}
+	res, err := GetRepJobToStop(policyID)
+	if err != nil {
+		log.Errorf("Failed to Get Jobs, error: %v", err)
+		return
+	}
+	//time.Sleep(15 * time.Second)
+	if len(res) != 2 {
+		log.Errorf("Expected length of stoppable jobs, expected:2, in fact: %d", len(res))
+		return
+	}
+}
+
 func TestDeleteRepTarget(t *testing.T) {
 	err := DeleteRepTarget(targetID)
 	if err != nil {

dao/replication_job.go

@@ -102,11 +102,23 @@ func GetRepJob(id int64) (*models.RepJob, error) {
 	return &j, err
 }
 func GetRepJobByPolicy(policyID int64) ([]*models.RepJob, error) {
-	o := orm.NewOrm()
 	var res []*models.RepJob
-	_, err := o.QueryTable("replication_job").Filter("policy_id", policyID).All(&res)
+	_, err := repJobPolicyIDQs(policyID).All(&res)
 	return res, err
 }
+
+// GetRepJobToStop get jobs that are possibly being handled by workers of a certain policy.
+func GetRepJobToStop(policyID int64) ([]*models.RepJob, error) {
+	var res []*models.RepJob
+	_, err := repJobPolicyIDQs(policyID).Filter("status__in", models.JobPending, models.JobRunning).All(&res)
+	return res, err
+}
+
+func repJobPolicyIDQs(policyID int64) orm.QuerySeter {
+	o := orm.NewOrm()
+	return o.QueryTable("replication_job").Filter("policy_id", policyID)
+}
+
 func DeleteRepJob(id int64) error {
 	o := orm.NewOrm()
 	_, err := o.Delete(&models.RepJob{ID: id})

job/statemachine.go

@@ -39,12 +39,12 @@ type JobSM struct {
 // EnsterState transit the statemachine from the current state to the state in parameter.
 // It returns the next state the statemachine should tranit to.
 func (sm *JobSM) EnterState(s string) (string, error) {
-	log.Debugf("Trying to transit from State: %s, to State: %s", sm.CurrentState, s)
+	log.Debugf("Job id: %d, transiting from State: %s, to State: %s", sm.JobID, sm.CurrentState, s)
 	targets, ok := sm.Transitions[sm.CurrentState]
 	_, exist := targets[s]
 	_, isForced := sm.ForcedStates[s]
 	if !exist && !isForced {
-		return "", fmt.Errorf("Transition from %s to %s does not exist!", sm.CurrentState, s)
+		return "", fmt.Errorf("Job id: %d, transition from %s to %s does not exist!", sm.JobID, sm.CurrentState, s)
 	}
 	exitHandler, ok := sm.Handlers[sm.CurrentState]
 	if ok {
@@ -52,7 +52,7 @@ func (sm *JobSM) EnterState(s string) (string, error) {
 			return "", err
 		}
 	} else {
-		log.Debugf("No handler found for state:%s, skip", sm.CurrentState)
+		log.Debugf("Job id: %d, no handler found for state:%s, skip", sm.JobID, sm.CurrentState)
 	}
 	enterHandler, ok := sm.Handlers[s]
 	var next string = models.JobContinue
@@ -62,11 +62,11 @@ func (sm *JobSM) EnterState(s string) (string, error) {
 			return "", err
 		}
 	} else {
-		log.Debugf("No handler found for state:%s, skip", s)
+		log.Debugf("Job id: %d, no handler found for state:%s, skip", sm.JobID, s)
 	}
 	sm.PreviousState = sm.CurrentState
 	sm.CurrentState = s
-	log.Debugf("Transition succeeded, current state: %s", s)
+	log.Debugf("Job id: %d, transition succeeded, current state: %s", sm.JobID, s)
 	return next, nil
 }
 
@@ -75,10 +75,10 @@ func (sm *JobSM) EnterState(s string) (string, error) {
 // will enter error state if there's more than one possible path when next state is "_continue"
 func (sm *JobSM) Start(s string) {
 	n, err := sm.EnterState(s)
-	log.Debugf("next state from handler: %s", n)
+	log.Debugf("Job id: %d, next state from handler: %s", sm.JobID, n)
 	for len(n) > 0 && err == nil {
 		if d := sm.getDesiredState(); len(d) > 0 {
-			log.Debugf("Desired state: %s, will ignore the next state from handler")
+			log.Debugf("Job id: %d. Desired state: %s, will ignore the next state from handler", sm.JobID, d)
 			n = d
 			sm.setDesiredState("")
 			continue
@@ -87,19 +87,19 @@ func (sm *JobSM) Start(s string) {
 			for n = range sm.Transitions[sm.CurrentState] {
 				break
 			}
-			log.Debugf("Continue to state: %s", n)
+			log.Debugf("Job id: %d, Continue to state: %s", sm.JobID, n)
 			continue
 		}
 		if n == models.JobContinue && len(sm.Transitions[sm.CurrentState]) != 1 {
-			log.Errorf("Next state is continue but there are %d possible next states in transition table", len(sm.Transitions[sm.CurrentState]))
+			log.Errorf("Job id: %d, next state is continue but there are %d possible next states in transition table", sm.JobID, len(sm.Transitions[sm.CurrentState]))
 			err = fmt.Errorf("Unable to continue")
 			break
 		}
 		n, err = sm.EnterState(n)
-		log.Debugf("next state from handler: %s", n)
+		log.Debugf("Job id: %d, next state from handler: %s", sm.JobID, n)
 	}
 	if err != nil {
-		log.Warningf("The statemachin will enter error state due to error: %v", err)
+		log.Warningf("Job id: %d, the statemachin will enter error state due to error: %v", sm.JobID, err)
 		sm.EnterState(models.JobError)
 	}
 }
@@ -121,8 +121,17 @@ func (sm *JobSM) RemoveTransition(from string, to string) {
 	delete(sm.Transitions[from], to)
 }
 
-func (sm *JobSM) Stop() {
+func (sm *JobSM) Stop(id int64) {
-	sm.setDesiredState(models.JobStopped)
+	log.Debugf("Trying to stop the job: %d", id)
+	sm.lock.Lock()
+	defer sm.lock.Unlock()
+	//need to check if the sm switched to other job
+	if id == sm.JobID {
+		sm.desiredState = models.JobStopped
+		log.Debugf("Desired state of job %d is set to stopped", id)
+	} else {
+		log.Debugf("State machine has switched to job %d, so the action to stop job %d will be ignored", sm.JobID, id)
+	}
 }
 
 func (sm *JobSM) getDesiredState() string {
@@ -141,14 +150,19 @@ func (sm *JobSM) Init() {
 	sm.lock = &sync.Mutex{}
 	sm.Handlers = make(map[string]StateHandler)
 	sm.Transitions = make(map[string]map[string]struct{})
-
+	sm.ForcedStates = map[string]struct{}{
+		models.JobError:    struct{}{},
+		models.JobStopped:  struct{}{},
+		models.JobCanceled: struct{}{},
+	}
 }
-func (sm *JobSM) Reset(jid int64) error {
-	sm.JobID = jid
 
-	sm.AddTransition(models.JobPending, models.JobRunning, StatusUpdater{DummyHandler{JobID: sm.JobID}, models.JobRunning})
+func (sm *JobSM) Reset(jid int64) error {
-	sm.Handlers[models.JobError] = StatusUpdater{DummyHandler{JobID: sm.JobID}, models.JobError}
+	//To ensure the new jobID is visible to the thread to stop the SM
-	sm.Handlers[models.JobStopped] = StatusUpdater{DummyHandler{JobID: sm.JobID}, models.JobStopped}
+	sm.lock.Lock()
+	sm.JobID = jid
+	sm.desiredState = ""
+	sm.lock.Unlock()
 
 	//init parms
 	job, err := dao.GetRepJob(sm.JobID)
@@ -172,7 +186,7 @@ func (sm *JobSM) Reset(jid int64) error {
 		Operation:   job.Operation,
 	}
 	if policy.Enabled == 0 {
-		//handler will cancel this job
+		//worker will cancel this job
 		return nil
 	}
 	target, err := dao.GetRepTarget(policy.TargetID)
@@ -188,16 +202,23 @@ func (sm *JobSM) Reset(jid int64) error {
 	//init states handlers
 	sm.Logger = utils.Logger{sm.JobID}
 	sm.CurrentState = models.JobPending
+
+	sm.AddTransition(models.JobPending, models.JobRunning, StatusUpdater{DummyHandler{JobID: sm.JobID}, models.JobRunning})
+	sm.Handlers[models.JobError] = StatusUpdater{DummyHandler{JobID: sm.JobID}, models.JobError}
+	sm.Handlers[models.JobStopped] = StatusUpdater{DummyHandler{JobID: sm.JobID}, models.JobStopped}
+
 	if sm.Parms.Operation == models.RepOpTransfer {
 		/*
 			sm.AddTransition(models.JobRunning, "pull-img", ImgPuller{DummyHandler: DummyHandler{JobID: sm.JobID}, img: sm.Parms.Repository, logger: sm.Logger})
-			//only handle on target for now
+			//only handle one target for now
 			sm.AddTransition("pull-img", "push-img", ImgPusher{DummyHandler: DummyHandler{JobID: sm.JobID}, targetURL: sm.Parms.TargetURL, logger: sm.Logger})
 			sm.AddTransition("push-img", models.JobFinished, StatusUpdater{DummyHandler{JobID: sm.JobID}, models.JobFinished})
 		*/
+
 		if err = addImgOutTransition(sm); err != nil {
 			return err
 		}
+
 	}
 	return nil
 }

job/workerpool.go

@@ -7,7 +7,24 @@ import (
 	"github.com/vmware/harbor/utils/log"
 )
 
-var WorkerPool chan *Worker
+type workerPool struct {
+	workerChan chan *Worker
+	workerList []*Worker
+}
+
+var WorkerPool *workerPool
+
+func (wp *workerPool) StopJobs(jobs []int64) {
+	log.Debugf("Works working on jobs: %v will be stopped", jobs)
+	for _, id := range jobs {
+		for _, w := range wp.workerList {
+			if w.SM.JobID == id {
+				log.Debugf("found a worker whose job ID is %d, will try to stop it", id)
+				w.SM.Stop(id)
+			}
+		}
+	}
+}
 
 type Worker struct {
 	ID      int
@@ -19,7 +36,7 @@ type Worker struct {
 func (w *Worker) Start() {
 	go func() {
 		for {
-			WorkerPool <- w
+			WorkerPool.workerChan <- w
 			select {
 			case jobID := <-w.RepJobs:
 				log.Debugf("worker: %d, will handle job: %d", w.ID, jobID)
@@ -43,10 +60,10 @@ func (w *Worker) Stop() {
 func (w *Worker) handleRepJob(id int64) {
 	err := w.SM.Reset(id)
 	if err != nil {
-		log.Errorf("Worker %d, failed to re-initialize statemachine, error: %v", w.ID, err)
+		log.Errorf("Worker %d, failed to re-initialize statemachine for job: %d, error: %v", w.ID, id, err)
 		err2 := dao.UpdateRepJobStatus(id, models.JobError)
 		if err2 != nil {
-			log.Errorf("Failed to update job status to ERROR, error:%v", err2)
+			log.Errorf("Failed to update job status to ERROR, job: %d, error:%v", id, err2)
 		}
 		return
 	}
@@ -70,9 +87,13 @@ func NewWorker(id int) *Worker {
 }
 
 func InitWorkerPool() {
-	WorkerPool = make(chan *Worker, config.MaxJobWorkers())
+	WorkerPool = &workerPool{
+		workerChan: make(chan *Worker, config.MaxJobWorkers()),
+		workerList: make([]*Worker, 0, config.MaxJobWorkers()),
+	}
 	for i := 0; i < config.MaxJobWorkers(); i++ {
 		worker := NewWorker(i)
+		WorkerPool.workerList = append(WorkerPool.workerList, worker)
 		worker.Start()
 		log.Debugf("worker %d started", worker.ID)
 	}
@@ -84,7 +105,7 @@ func Dispatch() {
 		case job := <-JobQueue:
 			go func(jobID int64) {
 				log.Debugf("Trying to dispatch job: %d", jobID)
-				worker := <-WorkerPool
+				worker := <-WorkerPool.workerChan
 				worker.RepJobs <- jobID
 			}(job)
 		}

jobservice/router.go

@@ -8,4 +8,5 @@ import (
 
 func initRouters() {
 	beego.Router("/api/jobs/replication", &api.ReplicationJob{})
+	beego.Router("/api/jobs/replication/actions", &api.ReplicationJob{}, "post:HandleAction")
 }

jobservice/stop.json

@@ -0,0 +1,4 @@
+{
+    "policy_id":1,
+    "action":"stop"
+}