Stacktrace when scanning k8s containers and no info about the problem

[ymettier]

Using

• Trivy-operator 0.21.1
• from official helm chart 0.23.1
• with values.yaml at the end of the report (see below)
• on Kubernetes 1.30.1

I have 2 problems. The first one, I cannot investigate because of the second one. So this issue is not about the first problem but about the second one.

How I reproduce the problem

Here is the second problem.

➜ kubectl -n trivy-system logs trivy-operator-8797d6d8b-78zhh
...
{"level":"error","ts":"2024-05-26T16:19:53Z","logger":"reconciler.scan job","msg":"Scan job container","job":"trivy-system/scan-vulnerabilityreport-687bbcc85b","container":"<REDACTED>","status.reason":"Error","status.message":"","stacktrace":"github.com/aquasecurity/trivy-operator/pkg/vulnerabilityreport/controller.(*ScanJobController).completedContainers\n\t/home/runner/work/trivy-operator/trivy-operator/pkg/vulnerabilityreport/controller/scanjob.go:353\ngithub.com/aquasecurity/trivy-operator/pkg/vulnerabilityreport/controller.(*ScanJobController).SetupWithManager.(*ScanJobController).reconcileJobs.func1\n\t/home/runner/work/trivy-operator/trivy-operator/pkg/vulnerabilityreport/controller/scanjob.go:80\nsigs.k8s.io/controller-runtime/pkg/reconcile.Func.Reconcile\n\t/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/reconcile/reconcile.go:113\nsigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).Reconcile\n\t/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:114\nsigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).reconcileHandler\n\t/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:311\nsigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).processNextWorkItem\n\t/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:261\nsigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).Start.func2.2\n\t/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:222"}

To make it easier to read, let's show the above message with printf :

github.com/aquasecurity/trivy-operator/pkg/vulnerabilityreport/controller.(*ScanJobController).completedContainers
	/home/runner/work/trivy-operator/trivy-operator/pkg/vulnerabilityreport/controller/scanjob.go:353
github.com/aquasecurity/trivy-operator/pkg/vulnerabilityreport/controller.(*ScanJobController).SetupWithManager.(*ScanJobController).reconcileJobs.func1
	/home/runner/work/trivy-operator/trivy-operator/pkg/vulnerabilityreport/controller/scanjob.go:80
sigs.k8s.io/controller-runtime/pkg/reconcile.Func.Reconcile
	/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/reconcile/reconcile.go:113
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).Reconcile
	/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:114
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).reconcileHandler
	/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:311
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).processNextWorkItem
	/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:261
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).Start.func2.2
	/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:222

In the source code I understand that

• the container scan was completed
• it failed for some reason
• the status.message points to itself in the log.Error
• the stacktrace also points to the log.Error

The issue

The issue is that I have a problem when the operator scans a container. I know it in the logs, but

• I have a stacktrace : I should not have a stacktrace because stacktraces are when the software is buggy
• neither the log, neither the stacktrace give information about what does not work.

I will try to investigate on my first problem, but for this one, could you make the logs more explicite please ?

my helm values

fullnameOverride: "trivy-operator"

operator:
  # -- builtInTrivyServer The flag enables the usage of built-in trivy server in cluster. It also overrides the following trivy params with built-in values
  # trivy.mode = ClientServer and serverURL = http://<serverServiceName>.<trivy operator namespace>:4975
  builtInTrivyServer: true
  privateRegistryScanSecretsNames: { "trivy-system": "<REDACTED>" }

serviceMonitor:
  enabled: true
  labels:
    k8s-app: "trivy-operator"

trivy:
  # -- mode is the Trivy client mode. Either Standalone or ClientServer. Depending
  # on the active mode other settings might be applicable or required.
  mode: ClientServer

  # -- whether to use a storage class for trivy server or emptydir (one mey want to use ephemeral storage)
  storageClassEnabled: true

  # -- storageClassName is the name of the storage class to be used for trivy server PVC. If empty, tries to find default storage class
  storageClassName: "<REDACTED>"

  # -- storageSize is the size of the trivy server PVC
  storageSize: "5Gi"

  # -- serverInsecure is the flag to enable insecure connection to the Trivy server.
  serverInsecure: false
  insecureRegistries:
    internalRegistry: <REDACTED>

nodeCollector:
  tolerations:
    - operator: Exists
      effect: NoSchedule

[ymettier]

I forgot to say that the pod running the operator is stable, up&running for days. No issue there.

The stacktrace I mention is not in the operator but probably in the pod that runs the vulnerability scan.

And the fix for issue 2101 (this issue) is probably just returning something else than a stacktrace in the logs because Trivy-operator is stable and working. Something that helps to find why the pod running the vulnerability scan failed (I have its name in the log, but only its name).

[chen-keinan]

@ymettier is it happen on specific image ? every time ? can you share it (if it public) ?

[ymettier]

I fixed my initial problem. I was switch from one private registry to another. Both are "insecure" but I forgot to add the new one in the configuration. My first problem is fixed.

About this issue, it happened on all images that were on my private registry. Every time of course.

In fact, the problem should appear at least every time that Trivy fails to retrieve an image.

You can reproduce my problem by creating a deployment/statefulset with some broken image :

apiVersion: apps/v1
kind: Deployment
metadata:
  name: issue-2101
spec:
  replicas: 1
  selector:
    matchLabels:
      app: issue-2101
  template:
    metadata:
      name: issue-2101
      labels:
        app: issue-2101
    spec:
      containers:
      - image: nowhere.nodomain.com:12345/nothing/noimage:latest
        imagePullPolicy: IfNotPresent
        name: issue-2101

How can you troubleshoot the problem with only the log with the stacktrace ?

[chen-keinan]

@ymettier can you please elaborate on problems:

• are you able to scan images with private registry at all?

[sheeeng]

To make it easier to read, let's show the above message with printf :

github.com/aquasecurity/trivy-operator/pkg/vulnerabilityreport/controller.(*ScanJobController).completedContainers
	/home/runner/work/trivy-operator/trivy-operator/pkg/vulnerabilityreport/controller/scanjob.go:353
github.com/aquasecurity/trivy-operator/pkg/vulnerabilityreport/controller.(*ScanJobController).SetupWithManager.(*ScanJobController).reconcileJobs.func1
	/home/runner/work/trivy-operator/trivy-operator/pkg/vulnerabilityreport/controller/scanjob.go:80
sigs.k8s.io/controller-runtime/pkg/reconcile.Func.Reconcile
	/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/reconcile/reconcile.go:113
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).Reconcile
	/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:114
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).reconcileHandler
	/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:311
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).processNextWorkItem
	/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:261
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).Start.func2.2
	/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:222

I was able to reproduce the same stacktrace with a minimal example as described in #2095 (comment).

[ymettier]

When the private registries are correctly configured : it works.
When an insecure registry is not configured : I have the stacktrace.

However, with the above deployment that helps to reproduce the problem, I guess that bad-configured private registries are just another way to reproduce the problem.

[urcus]

Hello,

I did encounter the same issue as, from your discussion i understand that the issue seems to came from my private registry configuration. But where ? I don't know (even after a lot of research / test)...
Is there a possibility to make the error the log more explicit than just the stacktrace ? this should maybe help people find where the issue comme from as "[email protected]/pkg/internal/controller/controller.go:222" is not really helpfull.

[ymettier]

I share @urcus 's comments as it was my "first problem" in the description of the issue.

However, this issue is expliciteliy not about this and it is solved and closed. A new issue should be open to get more explicit information when trivy fails.

(and thanks @chen-keinan for the fix! )