ACM-31164: Test and validate network policies for Image Based components

[shay23bra]

Summary

• Adds a NetworkPolicy for the image-based-install-operator pod, following the ACM 5.0 DDR pattern (Create-Once + Operand Ownership)
• Restricts ingress to only the ports the operator listens on (8080 metrics, 9443 webhook, 8000 image server)
• Restricts egress to DNS, Kubernetes API, spoke cluster APIs (port 6443), and HTTPS (port 443)

NetworkPolicy Details

Ingress:

• Same-namespace traffic (podSelector: {})
• OpenShift router → port 8000 (image server)
• Monitoring → port 8080 (metrics, exposed via ServiceMonitor on OCP)
• Any → port 9443 (webhook, called by kube-apiserver)
• Any → port 8000 (image server, called by Ironic/BMC)

Egress:

• DNS (openshift-dns, port 5353)
• Kubernetes API (port 6443)
• Spoke cluster API servers (0.0.0.0/0, port 6443)
• HTTPS (0.0.0.0/0, port 443, excluding AWS metadata)

Test plan

• Deploy NetworkPolicy on cluster with IBIO installed
• Verify IBIO pod remains healthy (no restarts, no network errors)
• Verify webhook (port 9443) responds from inside and outside namespace
• Verify image server (port 8000) responds from inside and outside namespace
• Verify random ports are blocked from inside and outside namespace
• Verify egress to DNS and Kubernetes API works
• Verify metrics endpoint (port 8080) is scrapable via ServiceMonitor on OCP
• Run end-to-end image-based installation flow on OCP

Test results (kind + Calico)

=== SUMMARY: 13 passed, 0 failed ===

Ingress allowed:  9443 (webhook), 8000 (image server), 8081 (health) ✓
Ingress blocked:  random ports denied from inside and outside namespace ✓
Cross-namespace:  8000 and 9443 reachable from default namespace ✓
Egress:           DNS resolution and Kubernetes API reachable ✓
Pod health:       running, all containers ready, no restarts, no network errors ✓

Note: metrics port 8080 binds to localhost in the manager — reachable via ServiceMonitor/Service on OCP, not via pod IP directly.

Related

• Jira: ACM-31164
• Also resolves: ACM-37746, ACM-36702
• Epic: ACM-31152 (network policies for all ACM components)
• DDR: ACM-DDR-XXX (Enablement for NetworkPolicies by ACM Operators)

Summary by CodeRabbit

• New Features
  • Added a Kubernetes NetworkPolicy for the image-based installation operator to restrict and allow required ingress/egress traffic.
• Chores
  • Included the new NetworkPolicy in the manager deployment configuration.
  • Updated the bundle metadata timestamp in the ClusterServiceVersion manifest.

[openshift-ci-robot]

@shay23bra: This pull request references ACM-31164 which is a valid jira issue.

Warning: The referenced jira issue has an invalid target version for the target branch this PR targets: expected the task to target either version "5.0." or "openshift-5.0.", but it targets "ACM 5.0.0" instead.

Details

In response to this:

Summary

• Adds a NetworkPolicy for the image-based-install-operator pod, following the ACM 5.0 DDR pattern (Create-Once + Operand Ownership)
• Restricts ingress to only the ports the operator listens on (8080 metrics, 9443 webhook, 8000 image server)
• Restricts egress to DNS, Kubernetes API, spoke cluster APIs (port 6443), and HTTPS (port 443)

NetworkPolicy Details

Ingress:

• Same-namespace traffic (podSelector: {})
• OpenShift router → port 8000 (image server)
• Monitoring → port 8080 (metrics)
• Any → port 9443 (webhook, called by kube-apiserver)
• Any → port 8000 (image server, called by Ironic/BMC)

Egress:

• DNS (openshift-dns, port 5353)
• Kubernetes API (port 6443)
• Spoke cluster API servers (0.0.0.0/0, port 6443)
• HTTPS (0.0.0.0/0, port 443, excluding AWS metadata)

Test plan

• Deploy NetworkPolicy on OCP cluster with IBIO installed
• Verify IBIO pod remains healthy (no restarts, no network errors)
• Verify webhook (port 9443) responds from outside namespace
• Verify image server (port 8000) responds from outside namespace
• Verify metrics endpoint (port 8080) is scrapable
• Verify random ports are blocked
• Run end-to-end image-based installation flow

Related

• Jira: ACM-31164
• Epic: ACM-31152 (network policies for all ACM components)
• DDR: ACM-DDR-XXX (Enablement for NetworkPolicies by ACM Operators)
• Also addresses: ACM-36702

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository.

[openshift-ci]

Skipping CI for Draft Pull Request.
If you want CI signal for your change, please convert it to an actual PR.
You can still manually trigger a test run with /test all

[openshift-ci]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: shay23bra
Once this PR has been reviewed and has the lgtm label, please assign gamli75 for approval. For more information see the Code Review Process.

The full list of commands accepted by this bot can be found here.

Details

Needs approval from an approver in each of these files:

• OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Repository: openshift/coderabbit/.coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: 42ab03fe-f45e-4e21-8e22-fa49cb15d1c1

📥 Commits

Reviewing files that changed from the base of the PR and between bfd7e38 and 1a23edb.

📒 Files selected for processing (4)

• bundle/manifests/image-based-install-operator-networkpolicy.yaml
• bundle/manifests/image-based-install-operator.clusterserviceversion.yaml
• config/manager/image-based-install-operator-networkpolicy.yaml
• config/manager/kustomization.yaml

🚧 Files skipped from review as they are similar to previous changes (3)

• bundle/manifests/image-based-install-operator.clusterserviceversion.yaml
• config/manager/kustomization.yaml
• config/manager/image-based-install-operator-networkpolicy.yaml

---

Walkthrough

Adds and registers a Kubernetes NetworkPolicy for the image-based install operator, defining ingress and egress access for DNS, monitoring, API, HTTPS, and selected ports. Bundles the policy and updates the ClusterServiceVersion creation timestamp.

Changes

Operator network policy

Layer / File(s)	Summary
Define and register network policy config/manager/image-based-install-operator-networkpolicy.yaml, config/manager/kustomization.yaml	Defines ingress and egress rules for operator pods and adds the NetworkPolicy to manager resources.
Bundle policy and metadata bundle/manifests/image-based-install-operator-networkpolicy.yaml, bundle/manifests/image-based-install-operator.clusterserviceversion.yaml	Adds the NetworkPolicy to the bundle and updates the CSV createdAt annotation timestamp.

Estimated code review effort: 2 (Simple) | ~10 minutes

Suggested reviewers: leo8a, tsorya

🚥 Pre-merge checks | ✅ 15

✅ Passed checks (15 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title is relevant to the main change: adding network policies for Image Based components, though it emphasizes testing/validation rather than the change itself.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Stable And Deterministic Test Names	✅ Passed	No Ginkgo tests were added or modified; the PR only changes YAML manifests, so there are no test titles to evaluate.
Test Structure And Quality	✅ Passed	PR only changes YAML manifests; no Ginkgo test files or test logic changed, so these test-quality criteria don’t apply.
Microshift Test Compatibility	✅ Passed	No new Ginkgo tests were added; the PR only changes YAML/CSV manifests, so MicroShift compatibility is not implicated.
Single Node Openshift (Sno) Test Compatibility	✅ Passed	No new Ginkgo/e2e test code was added; the PR only changes YAML manifests, so the SNO compatibility check is not applicable.
Topology-Aware Scheduling Compatibility	✅ Passed	Changes only add a NetworkPolicy/kustomization and update a CSV timestamp; no affinity, nodeSelector, topology spread, toleration, or replica logic was introduced.
Ote Binary Stdout Contract	✅ Passed	PASS: PR only changes YAML manifests and a CSV timestamp; no Go entrypoints or suite setup were modified, so no new stdout writes were introduced.
Ipv6 And Disconnected Network Test Compatibility	✅ Passed	No new Ginkgo e2e tests were added; the PR only changes manifests and kustomization, so this IPv4/external-connectivity test check is not applicable.
No-Weak-Crypto	✅ Passed	PASS: The modified YAML/CSV files contain no MD5/SHA1/DES/RC4/3DES/Blowfish/ECB, custom crypto, or secret-comparison logic.
Container-Privileges	✅ Passed	PASS: Touched manifests add only a NetworkPolicy; no privileged/root/host* or SYS_ADMIN settings were introduced, and existing securityContext stays false.
No-Sensitive-Data-In-Logs	✅ Passed	Only NetworkPolicy/CSV manifest changes were added; no logging code or sensitive data appears in the diff.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

Warning

There were issues while running some tools. Please review the errors and either fix the tool's configuration or disable the tool if it's a critical failure.

🔧 Checkov (3.3.8)

bundle/manifests/image-based-install-operator-networkpolicy.yaml

Traceback (most recent call last):
File "/usr/local/bin/checkov", line 2, in
from checkov.main import Checkov
ModuleNotFoundError: No module named 'checkov'

bundle/manifests/image-based-install-operator.clusterserviceversion.yaml

Traceback (most recent call last):
File "/usr/local/bin/checkov", line 2, in
from checkov.main import Checkov
ModuleNotFoundError: No module named 'checkov'

config/manager/image-based-install-operator-networkpolicy.yaml

Traceback (most recent call last):
File "/usr/local/bin/checkov", line 2, in
from checkov.main import Checkov
ModuleNotFoundError: No module named 'checkov'

• 1 others

---

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

Actionable comments posted: 2

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@config/manager/image-based-install-operator-networkpolicy.yaml`:
- Around line 17-18: Update the ingress rule in the network policy to add
explicit port restrictions, allowing same-namespace pods to reach only the
documented operator service ports and excluding the manager health endpoint on
8081. Preserve the existing podSelector scope while configuring the allowed
ports in the policy’s ports section.
- Around line 35-38: Restrict the TCP/9443 ingress rule in the image-based
install operator network policy to kube-apiserver traffic by adding the
platform-appropriate API-server source CIDR or selector under from. Preserve
port 9443 and the existing webhook ingress intent while preventing access from
all other sources.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Repository: openshift/coderabbit/.coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: 11b550e5-21c1-43b0-8b94-c68da4b7a283

📥 Commits

Reviewing files that changed from the base of the PR and between 6621941 and a3cc0b7.

📒 Files selected for processing (2)

• config/manager/image-based-install-operator-networkpolicy.yaml
• config/manager/kustomization.yaml

[coderabbitai]

🔒 Security & Privacy | 🟠 Major | ⚡ Quick win

Restrict same-namespace ingress to the intended ports.

This rule allows every pod in the namespace to reach every port on the operator, including the manager health endpoint on 8081. Add explicit port rules so the policy actually limits ingress to the documented service ports.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@config/manager/image-based-install-operator-networkpolicy.yaml` around lines
17 - 18, Update the ingress rule in the network policy to add explicit port
restrictions, allowing same-namespace pods to reach only the documented operator
service ports and excluding the manager health endpoint on 8081. Preserve the
existing podSelector scope while configuring the allowed ports in the policy’s
ports section.

[coderabbitai]

🔒 Security & Privacy | 🟠 Major | ⚡ Quick win

Scope webhook ingress to the API-server source.

A rule with ports but no from permits TCP/9443 from all sources, not just the kube-apiserver described by the comment. Restrict it using the platform-appropriate API-server source CIDR or selector.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@config/manager/image-based-install-operator-networkpolicy.yaml` around lines
35 - 38, Restrict the TCP/9443 ingress rule in the image-based install operator
network policy to kube-apiserver traffic by adding the platform-appropriate
API-server source CIDR or selector under from. Preserve port 9443 and the
existing webhook ingress intent while preventing access from all other sources.

[giladravid16]

@shay23bra A few things:

1. You need to run make bundle
2. The ipBlocks you used also block all ipv6 network, I'm not sure if they're even needed
3. I think some rules that are not needed since they are part of a larger rule, for example the Kubernetes API egress rule is also allowed by the Spoke cluster API servers egress rule (unless ipBlock knows to block service and pod ips)

[shay23bra]

@giladravid16

Good catches, thanks!

1. Done.
2. Good point - added ::/0 for IPv6 dual-stack support.
   Regarding whether ipBlock is needed at all: we need it for spoke cluster APIs (port 6443) and external HTTPS (port 443) since those destinations are outside the cluster with unpredictable IPs. A namespaceSelector/podSelector rule can't match external hosts.
3. You're right, removed redundant rules.

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@config/manager/image-based-install-operator-networkpolicy.yaml`:
- Around line 46-54: Restrict the TCP/6443 egress rule in the image-based
install operator NetworkPolicy by removing the unrestricted IPv6 ::/0 entry and
replacing it with the applicable Kubernetes and spoke API-server IPv6 CIDRs or
the approved egress path. If dynamic addresses require broad access, document
the security trade-off and add coverage proving the permitted destinations
remain constrained.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Repository: openshift/coderabbit/.coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: ecde0169-7142-4ccd-8431-85041cf96e24

📥 Commits

Reviewing files that changed from the base of the PR and between a3cc0b7 and bfd7e38.

📒 Files selected for processing (2)

• bundle/manifests/image-based-install-operator.clusterserviceversion.yaml
• config/manager/image-based-install-operator-networkpolicy.yaml

[shay23bra]

/retest

[shay23bra]

/retest

[shay23bra]

/retest-required

[openshift-ci]

@shay23bra: The following tests failed, say /retest to rerun all failed tests or /retest-required to rerun all mandatory failed tests:

Test name	Commit	Details	Required	Rerun command
ci/prow/e2e-ibio-v4v6	1a23edb	link	true	/test e2e-ibio-v4v6
ci/prow/ibio-reinstall-v4v6	1a23edb	link	false	/test ibio-reinstall-v4v6

Full PR test history. Your PR dashboard.

Details

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. I understand the commands that are listed here.