feat(cli): Provide more details when scan job fails (#328)

Resolves: #70

pkg/kube/pod/manager.go

@@ -5,16 +5,14 @@ import (
 	"fmt"
 	"io"
 
-	"k8s.io/klog"
-
 	"github.com/aquasecurity/starboard/pkg/kube"
-
 	apps "k8s.io/api/apps/v1"
 	batch "k8s.io/api/batch/v1"
 	batchv1beta1 "k8s.io/api/batch/v1beta1"
 	core "k8s.io/api/core/v1"
 	meta "k8s.io/apimachinery/pkg/apis/meta/v1"
 	"k8s.io/client-go/kubernetes"
+	"k8s.io/klog"
 )
 
 type Manager struct {
@@ -123,6 +121,9 @@ func (pw *Manager) GetTerminatedContainersStatusesByJob(ctx context.Context, job
 
 func GetTerminatedContainersStatusesByPod(pod *core.Pod) map[string]*core.ContainerStateTerminated {
 	states := make(map[string]*core.ContainerStateTerminated)
+	if pod == nil {
+		return states
+	}
 	for _, status := range pod.Status.InitContainerStatuses {
 		if status.State.Terminated == nil {
 			continue
@@ -159,7 +160,10 @@ func (pw *Manager) GetPodByJob(ctx context.Context, job *batch.Job) (*core.Pod,
 	if err != nil {
 		return nil, err
 	}
-	return &podList.Items[0], nil
+	if podList != nil && len(podList.Items) > 0 {
+		return &podList.Items[0], nil
+	}
+	return nil, nil
 }
 
 func (pw *Manager) GetPodLogs(ctx context.Context, pod *core.Pod, container string) (io.ReadCloser, error) {
@@ -186,16 +190,3 @@ func (pw *Manager) logTerminatedContainersErrors(statuses map[string]*core.Conta
 		klog.Errorf("Container %s terminated with %s: %s", container, status.Reason, status.Message)
 	}
 }
-
-// GetImages gets a slice of images for the specified PodSpec.
-func GetImages(spec core.PodSpec) (images []string) {
-	for _, c := range spec.InitContainers {
-		images = append(images, c.Image)
-	}
-
-	for _, c := range spec.Containers {
-		images = append(images, c.Image)
-	}
-
-	return
-}

pkg/kube/runnable_job.go

@@ -57,7 +57,8 @@ func (r *runnableJob) Run(ctx context.Context) error {
 		defaultResyncDuration,
 		informers.WithNamespace(r.job.Namespace),
 	)
-	jobInformer := informerFactory.Batch().V1().Jobs()
+	jobsInformer := informerFactory.Batch().V1().Jobs()
+	eventsInformer := informerFactory.Core().V1().Events()
 
 	var err error
 
@@ -90,8 +91,8 @@ func (r *runnableJob) Run(ctx context.Context) error {
 
 	complete := make(chan error)
 
-	jobInformer.Informer().AddEventHandler(cache.ResourceEventHandlerFuncs{
-		UpdateFunc: func(oldObj, newObj interface{}) {
+	jobsInformer.Informer().AddEventHandler(cache.ResourceEventHandlerFuncs{
+		UpdateFunc: func(_, newObj interface{}) {
 			newJob, ok := newObj.(*batchv1.Job)
 			if !ok {
 				return
@@ -113,6 +114,28 @@ func (r *runnableJob) Run(ctx context.Context) error {
 		},
 	})
 
+	eventsInformer.Informer().AddEventHandler(cache.ResourceEventHandlerFuncs{
+		AddFunc: func(obj interface{}) {
+			event := obj.(*corev1.Event)
+
+			// TODO We might want to look into events associate with the pod controlled by the current scan job.
+			// For example, when a pod cannot be scheduled due to insufficient resource requests,
+			// the event will be attached to the pod, not the job.
+			if event.InvolvedObject.UID != r.job.UID {
+				return
+			}
+
+			if event.Type == corev1.EventTypeNormal {
+				klog.V(3).Infof("Event: %s (%s)", event.Message, event.Reason)
+			}
+
+			if event.Type == corev1.EventTypeWarning {
+				complete <- fmt.Errorf("warning event received: %s (%s)", event.Message, event.Reason)
+				return
+			}
+		},
+	})
+
 	informerFactory.Start(wait.NeverStop)
 	informerFactory.WaitForCacheSync(wait.NeverStop)
 

pkg/vulnerabilityreport/scanner.go

@@ -101,11 +101,6 @@ func (s *Scanner) Scan(ctx context.Context, workload kube.Object) ([]v1alpha1.Vu
 
 	klog.V(3).Infof("Scan job completed: %s/%s", job.Namespace, job.Name)
 
-	job, err = s.clientset.BatchV1().Jobs(job.Namespace).Get(ctx, job.Name, metav1.GetOptions{})
-	if err != nil {
-		return nil, fmt.Errorf("getting scan job: %w", err)
-	}
-
 	return s.getVulnerabilityReportsByScanJob(ctx, job, owner)
 }