Skip to content

net, tests, hot-plug: W/A guest-agent deadlock after hot plug#5466

Open
Anatw wants to merge 2 commits into
RedHatQE:mainfrom
Anatw:hot_plug_bug_wa
Open

net, tests, hot-plug: W/A guest-agent deadlock after hot plug#5466
Anatw wants to merge 2 commits into
RedHatQE:mainfrom
Anatw:hot_plug_bug_wa

Conversation

@Anatw

@Anatw Anatw commented Jul 2, 2026

Copy link
Copy Markdown
Contributor
What this PR does / why we need it:

CNV-77961 causes guest-agent to deadlock after hot-plugging an interface on VMs that start with only a primary (masquerade) interface. When this happens, lookup_iface_status() never sees the new interface in VMI status, and tests fail.

This PR adds a console-based fallback: when guest-agent fails to report the hot-plugged interface, we look it up directly inside the VM via ip -j addr show, matching by MAC address. Three call sites are covered:

  1. hot_plug_interface() - falls back to console when VMI status times out
  2. set_secondary_static_ip_address() - verifies IP via console when guest-agent doesn't report it
  3. get_guest_vm_interface_name_by_vmi_interface_name() - resolves interface name via console

All fallback paths are guarded by is_jira_open("CNV-77961") and will be removed once the bug is fixed.

The first commit avoids a redundant VMI lookup by passing guest_device_name through from hot_plug_interface() to set_secondary_static_ip_address().

Which issue(s) this PR fixes:

CNV-77961

Special notes for reviewer:

I wasn't able to reproduce the bug in our evs - I ran it dozens of time on both tlv2 env (bm02-tlv2) and rdu2 env (bm04-cnvqe-rdu2) and the bug didn't show. For this reason, I did the verification in two directions:

  • Natural run (no forced failures) - all hot-plug tests pass, confirming nothing was broken by the changes.
  • Forced W/A run - guest-agent was killed (systemctl mask --now qemu-guest-agent) before hot-plug to guarantee the W/A path fires. After the W/A completed, the agent was restored (systemctl unmask + start) so subsequent test steps that depend on guest-agent could proceed normally. All three fallback paths were verified this way.
    Our plan is to get this W/A merged as is and follow the logs to see it in action when the bug appears. If we'll see an issue with the implementation - we will revert this immediately.

Example for the W/A path in action (by forcing it using #5467):
test_connectivity_of_hot_plugged_jumbo_interface — PASSED (forced W/A, all 3 paths exercised)

  1. VM created, agent killed (forced W/A)
  2. Hot-plug bridge interface → agent timeout (2 min) → W/A path 1 kicked in
    • Console found eth1 by MAC 02:24:15:4d:78:3d ✓
  3. IP set (172.16.241.3/24) → agent didn't report it → W/A path 2 kicked in
    • Console verified IP on eth1 ✓
  4. Utility VM hot-plugged normally (agent alive)
  5. Jumbo ping (MTU 9000): 3/3 packets, 0% loss ✓

Actual test logs:
2026-07-02T10:18:53.052667+00:00 timeout_sampler INFO Waiting for 300 seconds [0:05:00], retry every 10 seconds. (Function: utilities.console.connect Args: (<utilities.console.Console object at 0x7fb33673f9b0>,)) 2026-07-02T10:18:53.053362+00:00 timeout_sampler INFO Waiting for 300 seconds [0:05:00], retry every 5 seconds. (Function: pexpect.pty_spawn.spawn Kwargs: {'command': 'virtctl console jumbo-hot-plug-test-vm-1782987280-6881244 -n l2-bridge-test-bridge-nic-hot-plug', 'timeout': 30, 'encoding': 'utf-8'}) 2026-07-02T13:18:53.052230 tests.network.l2_bridge.libl2bridge WARNING CNV-77961: Guest agent did not report interface hot-plug-jumbo-iface on VM jumbo-hot-plug-test-vm-1782987280-6881244, falling back to console lookup by MAC 02:24:15:4d:78:3d. 2026-07-02T13:18:53.053109 utilities.console INFO Connect to jumbo-hot-plug-test-vm-1782987280-6881244 console ... 2026-07-02T13:18:54.276254 utilities.virt INFO Execute ip -j addr show on jumbo-hot-plug-test-vm-1782987280-6881244 2026-07-02T10:18:55.980250+00:00 timeout_sampler INFO Waiting for 300 seconds [0:05:00], retry every 10 seconds. (Function: utilities.console.connect Args: (<utilities.console.Console object at 0x7fb33673fce0>,)) 2026-07-02T10:18:55.981499+00:00 timeout_sampler INFO Waiting for 300 seconds [0:05:00], retry every 5 seconds. (Function: pexpect.pty_spawn.spawn Kwargs: {'command': 'virtctl console jumbo-hot-plug-test-vm-1782987280-6881244 -n l2-bridge-test-bridge-nic-hot-plug', 'timeout': 30, 'encoding': 'utf-8'}) 2026-07-02T13:18:55.978475 tests.network.l2_bridge.libl2bridge INFO CNV-77961: looking for MAC 02:24:15:4d:78:3d in guest jumbo-hot-plug-test-vm-1782987280-6881244, visible interfaces: [{'ifname': 'lo', 'address': '00:00:00:00:00:00'}, {'ifname': 'eth0', 'address': '02:24:15:4d:78:3c'}, {'ifname': 'eth1', 'address': '02:24:15:4d:78:3d'}] 2026-07-02T13:18:55.978900 tests.network.l2_bridge.libl2bridge INFO Console fallback found interface eth1 for hot-plug-jumbo-iface on VM jumbo-hot-plug-test-vm-1782987280-6881244. ... 2026-07-02T13:21:00.556203 utilities.console INFO jumbo-hot-plug-test-vm-1782987280-6881244: Got prompt \$ 2026-07-02T13:21:00.557031 utilities.virt INFO Execute ip -j -4 addr show eth1 on jumbo-hot-plug-test-vm-1782987280-6881244 2026-07-02T13:21:02.298828 tests.network.l2_bridge.libl2bridge WARNING CNV-77961: Verified IP 172.16.241.3 on eth1 via console (guest-agent not reporting on VM jumbo-hot-plug-test-vm-1782987280-6881244). 2026-07-02T13:21:02.299476 tests.network.l2_bridge.libl2bridge INFO jumbo-hot-plug-test-vm-1782987280-6881244/hot-plug-jumbo-iface set with IP address 172.16.241.3 2026-07-02T13:21:02.300469 conftest INFO Executing function fixture: hot_plugged_jumbo_interface_in_utility_vm ... PASSED2026-07-02T13:22:44.884117 utilities.network INFO ping returned PING 172.16.241.4 (172.16.241.4) 8972(9000) bytes of data.

jira-ticket:

https://issues.redhat.com/browse/CNV-77961

Anatw added 2 commits June 30, 2026 13:47
hot_plug_interface() already returns the full interface status including
interfaceName. Pass it through to set_secondary_static_ip_address()
instead of re-querying VMI status for the same value.

Signed-off-by: Anat Wax <awax@redhat.com>
Assisted-by: Claude <noreply@anthropic.com>
CNV-77961 causes guest-agent to stop reporting newly hot-plugged
interfaces, failing tests that depend on interface data from VMI status.
Work around the bug so tests continue to run until the fix is released.

Signed-off-by: Anat Wax <awax@redhat.com>
Assisted-by: Claude <noreply@anthropic.com>
@coderabbitai

coderabbitai Bot commented Jul 2, 2026

Copy link
Copy Markdown
Contributor

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info
⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: ASSERTIVE

Plan: Pro

Run ID: d73bbe00-16d5-4328-806d-d049f969d4ff

📥 Commits

Reviewing files that changed from the base of the PR and between 93d8ab2 and e8941cb.

📒 Files selected for processing (2)
  • tests/network/l2_bridge/libl2bridge.py
  • tests/network/l2_bridge/test_bridge_nic_hot_plug.py
👮 Files not reviewed due to content moderation or server errors (2)
  • tests/network/l2_bridge/test_bridge_nic_hot_plug.py
  • tests/network/l2_bridge/libl2bridge.py

📝 Walkthrough

[!WARNING]

Walkthrough skipped

File diffs could not be summarized.

🚥 Pre-merge checks | ❌ 1

❌ Failed checks (1 inconclusive)

Check name Status Explanation Resolution
Stp Link Required ❓ Inconclusive Custom check execution failed before a final verdict was produced. Retry the review run. If this persists, inspect pre-merge custom-check logs for infrastructure or agent runtime failures.
✨ Finishing Touches
🧪 Generate unit tests (beta)
  • Create PR with unit tests

Warning

Review ran into problems

🔥 Problems

Linked repositories: Your configuration references 1 linked repositories, but your current plan allows 0. Analyzed ``, skipped RedHatQE/openshift-virtualization-tests-design-docs.


Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

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

Anatw added a commit to Anatw/openshift-virtualization-tests that referenced this pull request Jul 2, 2026
For testing purposes only.
Kill guest-agent before hot-plug and omit guest_device_name to guarantee
all three W/A fallback paths are exercised. Not for review or merge.
Needed for PR
RedHatQE#5466

Signed-off-by: Anat Wax <awax@redhat.com>
Assisted-by: Claude <noreply@anthropic.com>
@openshift-virtualization-qe-bot-6

Copy link
Copy Markdown

Report bugs in Issues

Welcome! 🎉

This pull request will be automatically processed with the following features:

🔄 Automatic Actions

  • Reviewer Assignment: Reviewers are automatically assigned based on the OWNERS file in the repository root
  • Size Labeling: PR size labels (XS, S, M, L, XL, XXL) are automatically applied based on changes
  • Issue Creation: A tracking issue is created for this PR and will be closed when the PR is merged or closed
  • Branch Labeling: Branch-specific labels are applied to track the target branch
  • Auto-verification: Auto-verified users have their PRs automatically marked as verified
  • Labels: Enabled categories: branch, can-be-merged, cherry-pick, has-conflicts, hold, needs-rebase, size, verified, wip

📋 Available Commands

PR Status Management

  • /wip - Mark PR as work in progress (adds WIP: prefix to title)
  • /wip cancel - Remove work in progress status
  • /hold - Block PR merging (approvers only)
  • /hold cancel - Unblock PR merging
  • /verified - Mark PR as verified
  • /verified cancel - Remove verification status
  • /reprocess - Trigger complete PR workflow reprocessing (useful if webhook failed or configuration changed)
  • /regenerate-welcome - Regenerate this welcome message
  • /security-override - Set security check runs to pass (maintainers only)
  • /security-override cancel - Re-run security checks

Review & Approval

  • /lgtm - Approve changes (looks good to me)
  • /approve - Approve PR (approvers only)
  • /assign-reviewers - Assign reviewers based on OWNERS file
  • /assign-reviewer @username - Assign specific reviewer
  • /check-can-merge - Check if PR meets merge requirements

Testing & Validation

  • /retest tox - Run Python test suite with tox
  • /retest build-container - Rebuild and test container image
  • /retest verify-bugs-are-open - verify-bugs-are-open
  • /retest all - Run all available tests

Container Operations

  • /build-and-push-container - Build and push container image (tagged with PR number)
    • Supports additional build arguments: /build-and-push-container --build-arg KEY=value

Cherry-pick Operations

  • /cherry-pick <branch> - Schedule cherry-pick to target branch when PR is merged
    • Multiple branches: /cherry-pick branch1 branch2 branch3
  • /cherry-pick-retry <branch> - Retry a failed cherry-pick (merged PRs only)

Branch Management

  • /rebase - Rebase this PR branch onto its base branch

Custom Commands

  • /test-plan - Triggers CodeRabbit to analyze the PR's changed files and post a test execution plan
  • /rerun-smoke - Adds `retest-smoke` label on the PR to trigger smoke tests exectuion

Label Management

  • /<label-name> - Add a label to the PR
  • /<label-name> cancel - Remove a label from the PR

✅ Merge Requirements

This PR will be automatically approved when the following conditions are met:

  1. Approval: /approve from at least one approver
  2. LGTM Count: Minimum 2 /lgtm from reviewers
  3. Status Checks: All required status checks must pass
  4. No Blockers: No wip, hold, has-conflicts labels and PR must be mergeable (no conflicts)
  5. Verified: PR must be marked as verified

📊 Review Process

Approvers and Reviewers

Approvers:

  • EdDev
  • rnetser

Reviewers:

  • Anatw
  • EdDev
  • azhivovk
  • frenzyfriday
  • nirdothan
  • orelmisan
  • rnetser
  • servolkov
  • yossisegev
Available Labels
  • hold
  • verified
  • wip
  • lgtm
  • approve
AI Features
  • Cherry-Pick Conflict Resolution: Enabled (claude/claude-opus-4-6-1m)
Security Checks
  • Suspicious Path Detection: Monitors paths: .claude/, .vscode/, .cursor/, .devcontainer/, .pi/, .github/workflows/, .github/actions/
  • Committer Identity Check: Verifies last committer matches PR author
  • Mandatory: Security checks block merge (use /security-override to bypass — maintainers only)

💡 Tips

  • WIP Status: Use /wip when your PR is not ready for review
  • Verification: The verified label is removed on new commits unless the push is detected as a clean rebase
  • Cherry-picking: Cherry-pick labels are processed when the PR is merged
  • Container Builds: Container images are automatically tagged with the PR number
  • Permission Levels: Some commands require approver permissions
  • Auto-verified Users: Certain users have automatic verification and merge privileges

For more information, please refer to the project documentation or contact the maintainers.

@Anatw

Anatw commented Jul 2, 2026

Copy link
Copy Markdown
Contributor Author

/build-and-push-container

@Anatw

Anatw commented Jul 2, 2026

Copy link
Copy Markdown
Contributor Author

/verified

Both natural run + forced Q/A (using #5467)
Tested on bm03-tlv2
Openshift version: 4.22.0-rc.2
CNV version: 4.22.0
HCO image: brew.registry.redhat.io/rh-osbs/iib:1158778

@openshift-virtualization-qe-bot-3

Copy link
Copy Markdown
Contributor

@coderabbitai

Test execution plan request details

CRITICAL: You MUST post an inline review comment on the first changed line of the first file.
The inline comment should contain the full Test Execution Plan (smoke decision, gating decision, and specific affected tests).
Do NOT submit a blocking review event (REQUEST_CHANGES/APPROVE).
Post a single inline PR comment on Files Changed (non-blocking COMMENT flow).

As an expert software testing engineer, analyze all modified files in this PR and create a targeted test execution plan.
You will post an inline review comment with the test execution plan on the first changed file.
If you fail to run or post a comment, retry.

Analysis Requirements:

  1. Examine code changes in each modified file

  2. Identify affected code paths, functions, and classes

  3. Analyze pytest-specific elements: fixtures (scope, dependencies), parametrization, markers, conftest changes

  4. Trace test dependencies through imports, shared utilities, fixture inheritance, fixture teardown, and yield from cleanup in conftest

  5. Detect new tests introduced in the PR

  6. Utilities and libs impact (when utilities/ or libs/ changes):
    You MUST use shell scripts (rg, git diff) to trace the full impact.
    Follow these sub-steps in order:

    6a. Identify modified symbols: For each changed file under utilities/ or libs/,
    list every modified function or method.
    Example: git diff HEAD~1 --unified=0 -- utilities/hco.py | grep '^[+-]def '

    6b. Find direct callers: Search tests and conftest for each symbol from 6a.
    Example: rg -l 'get_hco_version' tests/

    6c. Trace fixture teardown and cleanup: Find fixtures that reach
    the modified symbol through yield from or context-manager wrappers.
    Example: rg -l 'yield from.*enable_common_boot|def.*enable_common_boot' tests/

    6d. Trace same-file callers: In each changed file, find other functions
    whose body calls a modified symbol (including code after yield
    in @contextmanager helpers).
    Example: rg 'get_hco_version|enable_common_boot' utilities/hco.py

    6e. Expand transitively: If function A calls modified B, then
    tests/fixtures that call A are affected — even when the test body
    never imports B directly.

    Do NOT limit impact to tests that import the modified symbol only.

  7. Smoke test impact: Intersect the affected set from step 6 with smoke-marked tests.
    Run: rg -l '@pytest.mark.smoke' tests/
    VERIFY the above command returned actual file paths before concluding False.
    Set True if either condition is met:

    • a smoke-marked file appears in the affected set from 6b-6e, OR
    • any conftest.py in the smoke test's parent-directory hierarchy (up to repo root)
      imports or calls a modified utilities/libs symbol — including autouse fixtures
      that depend on modified functions. ALL tests in that directory and below are affected.
      Example check: for each smoke_file, scan dirname(smoke_file)/conftest.py,
      dirname(dirname(smoke_file))/conftest.py, etc. for modified symbol imports
      and autouse fixtures that depend on modified symbols.
  8. Gating test impact: Intersect the affected set from step 6 with gating-marked tests.
    Run: rg -l '@pytest.mark.gating' tests/
    Set True if a gating-marked file also appears in the affected set from 6b-6e.
    Utilities/libs changes often affect gating tests without affecting smoke tests.
    Do NOT stop analysis after concluding Run smoke tests: False.

Output rules:
Do NOT include analysis step numbers (1-8) in your visible output.

Your deliverable:
Your inline informational comment will be based on the following requirements:

Test Execution Plan

  • Run smoke tests: True / False — If True, state the dependency path (test → fixture → changed symbol). True ONLY with a verified path.
  • Run gating tests: True / False — If True, state the dependency path. True if any gating-marked test is in the affected set.
  • Affected tests to run (required when utilities/, libs/, or shared conftest changes — list concrete paths even when smoke is False)

Use these formats:

  • path/to/test_file.py - When the entire test file needs verification
  • path/to/test_file.py::TestClass::test_method - When specific test(s) needed
  • path/to/test_file.py::test_function - When specific test(s) needed
  • -m marker - When a marker covers multiple affected tests (e.g. -m gating only if ALL gating tests in scope need run)
  • Tag each listed test or group with its marker when not obvious, e.g. (gating) or (smoke)

Real test commands (MANDATORY when changes affect session/runtime code):

When the affected code runs at session/collection time (conftest fixtures, pytest plugins,
config hooks, session-scoped setup) or modifies runtime behavior that unit tests mock away,
you MUST include concrete pytest commands the PR author must run on a real cluster
to verify the change works end-to-end. Include:

  • A command for the error/fix path (the scenario the PR fixes)
  • A command for the happy path (regression: the normal case still works)
  • Use lightweight tests (e.g., --collect-only for startup failures,
    a single small test for runtime behavior)
    If the PR only changes test logic (not utilities/libs/conftest), the affected test
    paths themselves serve as the real test commands — no separate section needed.

Example output for a session-startup fix:

**Real tests (cluster required)**
Error path (the fix):
`pytest tests/storage/.../test_foo.py --storage-class-matrix=nonexistent-sc --collect-only`
Expected: ValueError with clear message, not IndexError

Happy path (regression):
`pytest tests/storage/.../test_foo.py --storage-class-matrix=<valid-sc> -k test_bar`
Expected: session starts normally

Guidelines:

  • Include tests affected directly OR via fixture setup/teardown, yield from cleanup, or transitive utility call chains (caller calls modified helper)
  • Use a full file path only if ALL tests in that file require verification
  • Use file path + test name when only specific tests use an affected fixture or utility wrapper (preferred for partial file impact)
  • If a test marker can cover multiple files/tests, provide the marker
  • Balance coverage vs over-testing - Keep descriptions minimal
  • Example: if leaf helper foo() changes, include tests whose fixture teardown calls wrapper bar() where bar() calls foo(), even when the test body only imports an unrelated symbol from the same utilities module

Hardware-Related Checks (SR-IOV, GPU, DPDK):

When PR modifies fixtures for hardware-specific resources:

  • Collection Safety: Fixtures MUST have existence checks (return None when hardware unavailable)
  • Test Plan: MUST verify both WITH and WITHOUT hardware:
    • Run affected tests on cluster WITH hardware
    • Verify collection succeeds on cluster WITHOUT hardware

CRITICAL WORKFLOW COMPLETION RULES:

When responding to this test execution plan request, you MUST follow these rules EXACTLY:

  1. YOUR ONLY DELIVERABLE: Post one non-blocking inline comment containing the test execution plan on the first changed line
  2. THEN STOP IMMEDIATELY - Do NOT generate any additional response
  3. FALLBACK ONLY: If inline comment API calls fail after retrying, post as a regular PR comment
  4. SILENCE = SUCCESS: After successfully submitting the review, your task is complete. No confirmation needed.

ABSOLUTE PROHIBITIONS (violating these creates empty/meaningless reviews):

  • ❌ Do NOT post acknowledgment messages like "Test execution plan posted", "Review posted successfully", "I've successfully posted"
  • ❌ Do NOT mention review IDs, URLs, or confirmation of posting in the PR thread
  • ❌ Do NOT add any follow-up comments after submitting the review
  • ❌ Do NOT reply to confirm task completion
  • ❌ Do NOT explain what you did - just do it and stop

Remember: The pull request review is visible to users. Additional comments are redundant noise.

CRITICAL — Verification rules for smoke and gating decisions:
- For each True decision, you must have traced a concrete path: test file → fixture → utility function → modified symbol.
- Do NOT assume impact based on scope, file proximity, or infrastructure-sounding names.
Only set True when you trace a concrete path through steps 6a-6e, then 7 or 8.
WRONG: "utilities/hco.py changed → run smoke" (no traced dependency path)
WRONG: "Run smoke tests: False" with no affected-test list when gating tests use the changed utility
RIGHT: "Run smoke tests: False, Run gating tests: True" plus tests/.../test_foo.py::TestBar::test_baz (gating)
RIGHT: "Smoke test X uses fixture Y → wrapper Z → changed function W in utilities/hco.py"

@coderabbitai

coderabbitai Bot commented Jul 2, 2026

Copy link
Copy Markdown
Contributor

Tip

For best results, initiate chat on the files or code changes.

🧩 Analysis chain

🏁 Script executed:

#!/bin/bash
set -euo pipefail
gh pr view 5466 --repo RedHatQE/openshift-virtualization-tests --json files,title,number -q '.files[].path' | head -50
echo "---diff---"
gh pr diff 5466 --repo RedHatQE/openshift-virtualization-tests | head -400

Length of output: 9633


🏁 Script executed:

#!/bin/bash
set -euo pipefail
echo "--- markers in test file ---"
rg -n '`@pytest.mark`' tests/network/l2_bridge/test_bridge_nic_hot_plug.py
echo "--- callers of functions across repo ---"
rg -n 'hot_plug_interface\(|set_secondary_static_ip_address\(|get_guest_vm_interface_name_by_vmi_interface_name\(|hot_plug_interface_and_set_address\(' --type=py -g '!libl2bridge.py'
echo "--- smoke/gating files under l2_bridge ---"
rg -n '`@pytest.mark.smoke`|`@pytest.mark.gating`' tests/network/l2_bridge/
echo "--- conftest in dir hierarchy ---"
fd conftest.py tests/network/l2_bridge tests/network
echo "--- first changed line in first file for inline comment context ---"
sed -n '1,10p' tests/network/l2_bridge/libl2bridge.py

Length of output: 2516


🏁 Script executed:

#!/bin/bash
echo "--- callers of functions across repo (retry) ---"
rg -n 'hot_plug_interface\(|set_secondary_static_ip_address\(|get_guest_vm_interface_name_by_vmi_interface_name\(|hot_plug_interface_and_set_address\(' --type py | grep -v 'tests/network/l2_bridge/libl2bridge.py'
echo "--- done callers ---"
echo "--- smoke/gating markers repo-wide referencing l2_bridge ---"
rg -n '`@pytest.mark.smoke`|`@pytest.mark.gating`' tests/network/l2_bridge/ || echo "none found"
echo "--- conftest files ---"
fd conftest.py tests/network/l2_bridge tests/network || true
echo "--- first changed line ---"
sed -n '1,5p' tests/network/l2_bridge/libl2bridge.py

Length of output: 1004


🏁 Script executed:

#!/bin/bash
echo "=== callers repo-wide ==="
rg -n 'hot_plug_interface\(|set_secondary_static_ip_address\(|get_guest_vm_interface_name_by_vmi_interface_name\(|hot_plug_interface_and_set_address\(' --type py
echo "=== smoke/gating markers in l2_bridge dir ==="
rg -n '`@pytest.mark.smoke`|`@pytest.mark.gating`' tests/network/l2_bridge/
echo "=== conftest files ==="
fd conftest.py tests/network/l2_bridge tests/network
echo "=== first lines of file ==="
sed -n '1,5p' tests/network/l2_bridge/libl2bridge.py

Length of output: 836


🏁 Script executed:

#!/bin/bash
rg -n 'hot_plug_interface\(' --type py || true
echo "---"
rg -n 'set_secondary_static_ip_address\(' --type py || true
echo "---"
rg -n 'get_guest_vm_interface_name_by_vmi_interface_name\(' --type py || true
echo "---"
rg -n 'hot_plug_interface_and_set_address\(' --type py || true

Length of output: 431


🏁 Script executed:

#!/bin/bash
pwd
ls
fd -e py | wc -l
rg -n "def hot_plug_interface" .

Length of output: 654


🏁 Script executed:

#!/bin/bash
echo "=== callers of hot_plug_interface( ==="
rg -n 'hot_plug_interface\(' tests utilities libs
echo "=== callers of set_secondary_static_ip_address( ==="
rg -n 'set_secondary_static_ip_address\(' tests utilities libs
echo "=== callers of get_guest_vm_interface_name_by_vmi_interface_name( ==="
rg -n 'get_guest_vm_interface_name_by_vmi_interface_name\(' tests utilities libs
echo "=== callers of hot_plug_interface_and_set_address( ==="
rg -n 'hot_plug_interface_and_set_address\(' tests utilities libs
echo "=== smoke/gating in l2_bridge ==="
rg -n '`@pytest.mark.smoke`|`@pytest.mark.gating`' tests/network/l2_bridge
echo "=== conftest files ==="
fd conftest.py tests/network/l2_bridge tests/network

Length of output: 4477


🏁 Script executed:

#!/bin/bash
set -euo pipefail
REPO=RedHatQE/openshift-virtualization-tests
PR=5466
HEAD_SHA=$(gh pr view $PR --repo $REPO --json headRefOid -q .headRefOid)
echo "HEAD_SHA=$HEAD_SHA"

BODY=$(cat <<'EOF'
**Test Execution Plan**

- **Run smoke tests: False** — No `@pytest.mark.smoke` tests exist under `tests/network/l2_bridge/`.
- **Run gating tests: False** — No `@pytest.mark.gating` tests exist under `tests/network/l2_bridge/`; all callers of the changed functions (`hot_plug_interface`, `hot_plug_interface_and_set_address`, `set_secondary_static_ip_address`, `get_guest_vm_interface_name_by_vmi_interface_name`) are confined to `tests/network/l2_bridge/test_bridge_nic_hot_plug.py` and no conftest in the `tests/network/l2_bridge/` hierarchy wraps them with an autouse fixture.
- **Affected tests to run:**
  - `tests/network/l2_bridge/test_bridge_nic_hot_plug.py` — all fixtures/tests use `hot_plug_interface`, `hot_plug_interface_and_set_address`, or `set_secondary_static_ip_address` (via `hot_plugged_interface_with_address`, `hot_plugged_second_interface_with_address`, and the `TestBridgeNicHotPlug`/secondary-interface test classes), so the whole file needs verification.
  - Prioritize: `tests/network/l2_bridge/test_bridge_nic_hot_plug.py::test_connectivity_of_hot_plugged_jumbo_interface` (jumbo/MTU path, `special_infra`) and the secondary-static-IP tests (`hot_plugged_second_interface_with_address` consumers), since these exercise all three new console-fallback paths guarded by `is_jira_open("CNV-77961")`.

**Real tests (cluster required)**
Fix path (W/A, requires forcing guest-agent failure per PR notes, e.g. via `systemctl mask --now qemu-guest-agent`):
`pytest tests/network/l2_bridge/test_bridge_nic_hot_plug.py::TestBridgeNicHotPlug -k jumbo`
Expected: hot-plug succeeds via console MAC lookup, IP set verified via console, warnings logged referencing `CNV-77961`.

Happy path (regression, guest-agent healthy):
`pytest tests/network/l2_bridge/test_bridge_nic_hot_plug.py -m ipv4`
Expected: guest-agent reports interfaces/IPs normally; console fallback code paths are not triggered.
EOF
)

gh api \
  --method POST \
  -H "Accept: application/vnd.github+json" \
  "repos/$REPO/pulls/$PR/reviews" \
  -f commit_id="$HEAD_SHA" \
  -f event="COMMENT" \
  -f body=" " \
  -F "comments[][path]=tests/network/l2_bridge/libl2bridge.py" \
  -F "comments[][line]=2" \
  -F "comments[][side]=RIGHT" \
  -f "comments[][body]=$BODY"

Length of output: 3027


🏁 Script executed:

#!/bin/bash
set -euo pipefail
REPO=RedHatQE/openshift-virtualization-tests
PR=5466

BODY=$(cat <<'EOF'
**Test Execution Plan** (inline comment on `tests/network/l2_bridge/libl2bridge.py` line 2 was attempted but rejected by GitHub with `403 Resource not accessible by integration`; posting as a regular comment instead)

- **Run smoke tests: False** — No `@pytest.mark.smoke` tests exist under `tests/network/l2_bridge/`.
- **Run gating tests: False** — No `@pytest.mark.gating` tests exist under `tests/network/l2_bridge/`; all callers of the changed functions (`hot_plug_interface`, `hot_plug_interface_and_set_address`, `set_secondary_static_ip_address`, `get_guest_vm_interface_name_by_vmi_interface_name`) are confined to `tests/network/l2_bridge/test_bridge_nic_hot_plug.py`, and no conftest in the `tests/network/l2_bridge/` hierarchy wraps them with an autouse fixture.
- **Affected tests to run:**
  - `tests/network/l2_bridge/test_bridge_nic_hot_plug.py` — all fixtures/tests use `hot_plug_interface`, `hot_plug_interface_and_set_address`, or `set_secondary_static_ip_address` (via `hot_plugged_interface_with_address`, `hot_plugged_second_interface_with_address`, and the `TestBridgeNicHotPlug`/secondary-interface test classes), so the whole file needs verification.
  - Prioritize: `tests/network/l2_bridge/test_bridge_nic_hot_plug.py::test_connectivity_of_hot_plugged_jumbo_interface` (`special_infra`, jumbo/MTU path) and the secondary-static-IP tests (`hot_plugged_second_interface_with_address` consumers), since these exercise all three new console-fallback paths guarded by `is_jira_open("CNV-77961")`.

**Real tests (cluster required)**
Fix path (W/A, requires forcing guest-agent failure per PR notes, e.g. `systemctl mask --now qemu-guest-agent`):
`pytest tests/network/l2_bridge/test_bridge_nic_hot_plug.py::TestBridgeNicHotPlug -k jumbo`
Expected: hot-plug succeeds via console MAC lookup, IP set verified via console, warnings logged referencing `CNV-77961`.

Happy path (regression, guest-agent healthy):
`pytest tests/network/l2_bridge/test_bridge_nic_hot_plug.py -m ipv4`
Expected: guest-agent reports interfaces/IPs normally; console fallback code paths are not triggered.
EOF
)

gh pr comment $PR --repo $REPO --body "$BODY"

Length of output: 2623


@openshift-virtualization-qe-bot-3: I attempted to post the Test Execution Plan as both an inline review comment and a regular PR comment via the GitHub API, but both attempts were rejected with 403/GraphQL: Resource not accessible by integration — my current credentials lack pull_requests: write scope in this sandbox. Posting the plan here instead:

Test Execution Plan

  • Run smoke tests: False — No @pytest.mark.smoke tests exist under tests/network/l2_bridge/.
  • Run gating tests: False — No @pytest.mark.gating tests exist under tests/network/l2_bridge/. All callers of the changed functions (hot_plug_interface, hot_plug_interface_and_set_address, set_secondary_static_ip_address, get_guest_vm_interface_name_by_vmi_interface_name) are confined to tests/network/l2_bridge/test_bridge_nic_hot_plug.py, and no conftest in the tests/network/l2_bridge/ hierarchy (checked all conftest.py files up to tests/network/conftest.py) wraps them via an autouse fixture.
  • Affected tests to run:
    • tests/network/l2_bridge/test_bridge_nic_hot_plug.py — entire file needs verification, since every fixture/test that hot-plugs an interface or sets a secondary IP goes through the modified functions (hot_plugged_interface_with_address, hot_plugged_second_interface_with_address, and the TestBridgeNicHotPlug classes).
    • Prioritize test_connectivity_of_hot_plugged_jumbo_interface (special_infra, jumbo/MTU path) and the secondary-static-IP consumers of hot_plugged_second_interface_with_address, since these exercise all three new is_jira_open("CNV-77961") console-fallback paths.

Real tests (cluster required)

Fix path (W/A, force guest-agent failure per PR notes, e.g. systemctl mask --now qemu-guest-agent):
pytest tests/network/l2_bridge/test_bridge_nic_hot_plug.py::TestBridgeNicHotPlug -k jumbo
Expected: hot-plug succeeds via console MAC lookup, IP verified via console, CNV-77961 warnings logged.

Happy path (regression, guest-agent healthy):
pytest tests/network/l2_bridge/test_bridge_nic_hot_plug.py -m ipv4
Expected: guest-agent reports interfaces/IPs normally; console fallback paths are not triggered.

@openshift-virtualization-qe-bot-3

Copy link
Copy Markdown
Contributor

New container for quay.io/openshift-cnv/openshift-virtualization-tests:pr-5466 published

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants