Back to Checks
kubernetes

Scheduler pod has profiling disabled

scheduler_profiling
Service: schedulerSeverity: medium
OfficialBuilt-inVerified
Categories:cluster-security
Resource Type:Pod

Kubernetes Scheduler profiling configuration, specifically whether scheduler containers run with --profiling=false to keep the profiling API disabled.

Prowler logoProwlerprowler-provider-kubernetes

Risk

With profiling enabled, the pprof endpoints can expose runtime internals (stack traces, memory, goroutines), aiding reconnaissance and credential discovery, harming confidentiality.

Extra CPU/heap usage can be abused for DoS, impacting availability.

Run this check with Prowler

Run in Prowler Cloud
Run
prowler kubernetes --checks scheduler_profiling

Recommendation

Disable by default: set --profiling=false on the Scheduler.

If profiling is required, enable it only temporarily, restrict access with network policies, bind to loopback, and log/monitor usage. Apply least privilege and defense in depth to limit exposure.

Remediation

Other

1SSH to the control-plane node
2Edit the scheduler manifest: `sudo vi /etc/kubernetes/manifests/kube-scheduler.yaml`
3In `spec.containers[].command`, add this flag (or change existing to false):
4```
5--profiling=false
6```
7Save the file; kubelet will automatically restart the scheduler with profiling disabled

Check Code

from prowler.lib.check.models import Check, Check_Report_Kubernetes
from prowler.providers.kubernetes.services.scheduler.scheduler_client import (
    scheduler_client,
)


class scheduler_profiling(Check):
    def execute(self) -> Check_Report_Kubernetes:
        findings = []
        for pod in scheduler_client.scheduler_pods:
            report = Check_Report_Kubernetes(metadata=self.metadata(), resource=pod)
            report.status = "FAIL"
            report.status_extended = (
                f"Scheduler has profiling enabled in pod {pod.name}."
            )
            for container in pod.containers.values():
                if "--profiling=false" in str(container.command):
                    report.status = "PASS"
                    report.status_extended = (
                        f"Scheduler does not have profiling enabled in pod {pod.name}."
                    )
            findings.append(report)
        return findings

References