Prefer autoscaling based on jobs rather than workflows if available (#114)

Adds the ability to autoscale on jobs in addition to workflows. We fall back to using workflow metrics if job details are not present.

Resolves #89

controllers/autoscaling.go

@@ -4,8 +4,9 @@ import (
 	"context"
 	"errors"
 	"fmt"
-	"github.com/summerwind/actions-runner-controller/api/v1alpha1"
 	"strings"
+
+	"github.com/summerwind/actions-runner-controller/api/v1alpha1"
 )
 
 func (r *HorizontalRunnerAutoscalerReconciler) determineDesiredReplicas(rd v1alpha1.RunnerDeployment, hra v1alpha1.HorizontalRunnerAutoscaler) (*int, error) {
@@ -44,6 +45,38 @@ func (r *HorizontalRunnerAutoscalerReconciler) determineDesiredReplicas(rd v1alp
 	}
 
 	var total, inProgress, queued, completed, unknown int
+	type callback func()
+	listWorkflowJobs := func(user string, repoName string, runID int64, fallback_cb callback) {
+		if runID == 0 {
+			fallback_cb()
+			return
+		}
+		jobs, _, err := r.GitHubClient.Actions.ListWorkflowJobs(context.TODO(), user, repoName, runID, nil)
+		if err != nil {
+			r.Log.Error(err, "Error listing workflow jobs")
+			fallback_cb()
+		} else if len(jobs.Jobs) == 0 {
+			fallback_cb()
+		} else {
+			for _, job := range jobs.Jobs {
+				switch job.GetStatus() {
+				case "completed":
+					// We add a case for `completed` so it is not counted in `unknown`.
+					// And we do not increment the counter for completed because
+					// that counter only refers to workflows. The reason for
+					// this is because we do not get a list of jobs for
+					// completed workflows in order to keep the number of API
+					// calls to a minimum.
+				case "in_progress":
+					inProgress++
+				case "queued":
+					queued++
+				default:
+					unknown++
+				}
+			}
+		}
+	}
 
 	for _, repo := range repos {
 		user, repoName := repo[0], repo[1]
@@ -52,20 +85,20 @@ func (r *HorizontalRunnerAutoscalerReconciler) determineDesiredReplicas(rd v1alp
 			return nil, err
 		}
 
-		for _, r := range list.WorkflowRuns {
+		for _, run := range list.WorkflowRuns {
 			total++
 
 			// In May 2020, there are only 3 statuses.
 			// Follow the below links for more details:
 			// - https://developer.github.com/v3/actions/workflow-runs/#list-repository-workflow-runs
 			// - https://developer.github.com/v3/checks/runs/#create-a-check-run
-			switch r.GetStatus() {
+			switch run.GetStatus() {
 			case "completed":
 				completed++
 			case "in_progress":
-				inProgress++
+				listWorkflowJobs(user, repoName, run.GetID(), func() { inProgress++ })
 			case "queued":
-				queued++
+				listWorkflowJobs(user, repoName, run.GetID(), func() { queued++ })
 			default:
 				unknown++
 			}

controllers/autoscaling_test.go

@@ -2,16 +2,17 @@ package controllers
 
 import (
 	"fmt"
+	"net/http/httptest"
+	"net/url"
+	"testing"
+
 	"github.com/summerwind/actions-runner-controller/api/v1alpha1"
 	"github.com/summerwind/actions-runner-controller/github"
 	"github.com/summerwind/actions-runner-controller/github/fake"
 	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
 	"k8s.io/apimachinery/pkg/runtime"
 	clientgoscheme "k8s.io/client-go/kubernetes/scheme"
-	"net/http/httptest"
-	"net/url"
 	"sigs.k8s.io/controller-runtime/pkg/log/zap"
-	"testing"
 )
 
 func newGithubClient(server *httptest.Server) *github.Client {
@@ -44,9 +45,11 @@ func TestDetermineDesiredReplicas_RepositoryRunner(t *testing.T) {
 		sReplicas    *int
 		sTime        *metav1.Time
 		workflowRuns string
+		workflowJobs map[int]string
 		want         int
 		err          string
 	}{
+		// Legacy functionality
 		// 3 demanded, max at 3
 		{
 			repo:         "test/valid",
@@ -122,6 +125,21 @@ func TestDetermineDesiredReplicas_RepositoryRunner(t *testing.T) {
 			workflowRuns: `{"total_count": 4, "workflow_runs":[{"status":"in_progress"}, {"status":"in_progress"}, {"status":"in_progress"}, {"status":"completed"}]}"`,
 			want:         3,
 		},
+
+		// Job-level autoscaling
+		// 5 requested from 3 workflows
+		{
+			repo:         "test/valid",
+			min:          intPtr(2),
+			max:          intPtr(10),
+			workflowRuns: `{"total_count": 4, "workflow_runs":[{"id": 1, "status":"queued"}, {"id": 2, "status":"in_progress"}, {"id": 3, "status":"in_progress"}, {"status":"completed"}]}"`,
+			workflowJobs: map[int]string{
+				1: `{"jobs": [{"status":"queued"}, {"status":"queued"}]}`,
+				2: `{"jobs": [{"status": "in_progress"}, {"status":"completed"}]}`,
+				3: `{"jobs": [{"status": "in_progress"}, {"status":"queued"}]}`,
+			},
+			want: 5,
+		},
 	}
 
 	for i := range testcases {
@@ -136,7 +154,7 @@ func TestDetermineDesiredReplicas_RepositoryRunner(t *testing.T) {
 		_ = v1alpha1.AddToScheme(scheme)
 
 		t.Run(fmt.Sprintf("case %d", i), func(t *testing.T) {
-			server := fake.NewServer(fake.WithListRepositoryWorkflowRunsResponse(200, tc.workflowRuns))
+			server := fake.NewServer(fake.WithListRepositoryWorkflowRunsResponse(200, tc.workflowRuns), fake.WithListWorkflowJobsResponse(200, tc.workflowJobs))
 			defer server.Close()
 			client := newGithubClient(server)
 
@@ -211,6 +229,7 @@ func TestDetermineDesiredReplicas_OrganizationalRunner(t *testing.T) {
 		sReplicas    *int
 		sTime        *metav1.Time
 		workflowRuns string
+		workflowJobs map[int]string
 		want         int
 		err          string
 	}{
@@ -316,6 +335,22 @@ func TestDetermineDesiredReplicas_OrganizationalRunner(t *testing.T) {
 			workflowRuns: `{"total_count": 2, "workflow_runs":[{"status":"in_progress"}, {"status":"completed"}]}"`,
 			err:          "validating autoscaling metrics: spec.autoscaling.metrics[].repositoryNames is required and must have one more more entries for organizational runner deployment",
 		},
+
+		// Job-level autoscaling
+		// 5 requested from 3 workflows
+		{
+			org:          "test",
+			repos:        []string{"valid"},
+			min:          intPtr(2),
+			max:          intPtr(10),
+			workflowRuns: `{"total_count": 4, "workflow_runs":[{"id": 1, "status":"queued"}, {"id": 2, "status":"in_progress"}, {"id": 3, "status":"in_progress"}, {"status":"completed"}]}"`,
+			workflowJobs: map[int]string{
+				1: `{"jobs": [{"status":"queued"}, {"status":"queued"}]}`,
+				2: `{"jobs": [{"status": "in_progress"}, {"status":"completed"}]}`,
+				3: `{"jobs": [{"status": "in_progress"}, {"status":"queued"}]}`,
+			},
+			want: 5,
+		},
 	}
 
 	for i := range testcases {
@@ -330,7 +365,7 @@ func TestDetermineDesiredReplicas_OrganizationalRunner(t *testing.T) {
 		_ = v1alpha1.AddToScheme(scheme)
 
 		t.Run(fmt.Sprintf("case %d", i), func(t *testing.T) {
-			server := fake.NewServer(fake.WithListRepositoryWorkflowRunsResponse(200, tc.workflowRuns))
+			server := fake.NewServer(fake.WithListRepositoryWorkflowRunsResponse(200, tc.workflowRuns), fake.WithListWorkflowJobsResponse(200, tc.workflowJobs))
 			defer server.Close()
 			client := newGithubClient(server)
 

github/fake/fake.go

@@ -4,7 +4,10 @@ import (
 	"fmt"
 	"net/http"
 	"net/http/httptest"
+	"strconv"
+	"strings"
 	"time"
+	"unicode"
 )
 
 const (
@@ -31,6 +34,24 @@ func (h *Handler) ServeHTTP(w http.ResponseWriter, req *http.Request) {
 	fmt.Fprintf(w, h.Body)
 }
 
+type MapHandler struct {
+	Status int
+	Bodies map[int]string
+}
+
+func (h *MapHandler) ServeHTTP(w http.ResponseWriter, req *http.Request) {
+	// Parse out int key from URL path
+	key, err := strconv.Atoi(strings.TrimFunc(req.URL.Path, func(r rune) bool { return !unicode.IsNumber(r) }))
+	if err != nil {
+		w.WriteHeader(400)
+	} else if body := h.Bodies[key]; len(body) == 0 {
+		w.WriteHeader(404)
+	} else {
+		w.WriteHeader(h.Status)
+		fmt.Fprintf(w, body)
+	}
+}
+
 type ServerConfig struct {
 	*FixedResponses
 }
@@ -45,7 +66,7 @@ func NewServer(opts ...Option) *httptest.Server {
 		o(&config)
 	}
 
-	routes := map[string]*Handler{
+	routes := map[string]http.Handler{
 		// For CreateRegistrationToken
 		"/repos/test/valid/actions/runners/registration-token": &Handler{
 			Status: http.StatusCreated,
@@ -126,6 +147,9 @@ func NewServer(opts ...Option) *httptest.Server {
 
 		// For auto-scaling based on the number of queued(pending) workflow runs
 		"/repos/test/valid/actions/runs": config.FixedResponses.ListRepositoryWorkflowRuns,
+
+		// For auto-scaling based on the number of queued(pending) workflow jobs
+		"/repos/test/valid/actions/runs/": config.FixedResponses.ListWorkflowJobs,
 	}
 
 	mux := http.NewServeMux()

github/fake/options.go

@@ -2,6 +2,7 @@ package fake
 
 type FixedResponses struct {
 	ListRepositoryWorkflowRuns *Handler
+	ListWorkflowJobs           *MapHandler
 }
 
 type Option func(*ServerConfig)
@@ -15,6 +16,15 @@ func WithListRepositoryWorkflowRunsResponse(status int, body string) Option {
 	}
 }
 
+func WithListWorkflowJobsResponse(status int, bodies map[int]string) Option {
+	return func(c *ServerConfig) {
+		c.FixedResponses.ListWorkflowJobs = &MapHandler{
+			Status: status,
+			Bodies: bodies,
+		}
+	}
+}
+
 func WithFixedResponses(responses *FixedResponses) Option {
 	return func(c *ServerConfig) {
 		c.FixedResponses = responses