Repository navigation
[Storage] Add snapshot/restore tests across RHCOS 9 and RHCOS 10 (CNV-96772) - #1
Draft
Ahmad-Hafe wants to merge 1822 commits into
Draft
Ahmad-Hafe wants to merge 1822 commits into
Ahmad-Hafe wants to merge 1822 commits into
Conversation
github-actions
Bot
added
sig-chaos
sig-infra
sig-iuo
sig-network
sig-scale
sig-storage
sig-virt
labels
Ahmad-Hafe
force-pushed
the
cnv-96772-snapshot-restore-mixed-rhcos
branch
18 times, most recently
from
October 8, 2026 22:25
747dd40 to
714b47e
Compare
…edHatQE#5544) ##### What this PR does / why we need it: De-quarantines `test_enable_fg_hco` (Polarion CNV-10091) and updates its assertions to match the current HCO behavior. HCO's `disableMDevConfiguration` feature gate no longer propagates `DisableMDEVConfiguration` to KubeVirt's `developerConfiguration.featureGates` list. Instead, HCO now syncs it to `mediatedDevicesConfiguration.enabled = false` (inverted boolean) in the KubeVirt CR. **Changes:** - Remove `@pytest.mark.xfail` quarantine decorator - Remove unused imports and constants (`DEVELOPER_CONFIGURATION`, `VALUE_STR`, `QUARANTINED`, `FEATUREGATE_NAME_KEY_STR`) - Simplify parametrize (remove extra params `kubevirt_featuregate_name` and `hco_featuregate`) - Add `MEDIATED_DEVICES_CONFIGURATION` constant - New assertions verify: 1. HCO `featureGates.disableMDevConfiguration == True` 2. KubeVirt `mediatedDevicesConfiguration.enabled == False` ##### Which issue(s) this PR fixes: ##### Special notes for reviewer: Should be backported to cnv-4.22 too. ##### jira-ticket: https://issues.redhat.com/browse/CNV-79304 <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Tests** * Updated upgrade/install test coverage to reflect the current feature-gate behavior. * Added validation for mediated device configuration being disabled when the related feature gate is enabled. * Simplified test parameter handling for clearer assertions. <!-- end of auto-generated comment: release notes by coderabbit.ai --> Signed-off-by: Ramón Lobillo <rlobillo@redhat.com> Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
##### What this PR does / why we need it: The Jira is resolved, so the memory/CPU override in fedora_vm_for_hotplug_scope_class is no longer needed. ##### Which issue(s) this PR fixes: ##### Special notes for reviewer: ##### jira-ticket: Signed-off-by: Jose Manuel Castano <joscasta@redhat.com> Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
…PRs (RedHatQE#5604) ## Summary - Add CodeRabbit **pre-merge checks** for STP scenario coverage (linked design-docs repo) and `special_infra` marker enforcement on new tests. - Add **`tests/**` path instructions** so STP gaps are posted as **inline review comments** on the `STP:` line, not only in PR conversation chat. - Update **AGENTS.md** with STP scenario coverage rules and `special_infra` marker requirements aligned with `pytest.ini`. ##### What this PR does / why we need it: CodeRabbit linked-repo STP analysis worked in chat but findings were hard to see on PRs. This change makes STP mismatch and missing `special_infra` marker violations visible in the pre-merge checks table and (for STP) as inline file comments, and documents the same rules in AGENTS.md for humans and AI reviewers. ##### Which issue(s) this PR fixes: ##### Special notes for reviewer: - Two new custom pre-merge checks (`STP scenario coverage`, `special_infra marker required`) join the existing `STP link required` check. - Triggering markers for `special_infra` are listed explicitly; `single_nic`, `remote_cluster`, `mixed_os_nodes`, `cclm`, and operator-only markers are exempt. - After merge, validate on an STD PR with partial STP coverage — expect ❌ in pre-merge checks and an inline comment on the STP line. ##### jira-ticket: NONE Made with [Cursor](https://cursor.com) <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Documentation** * Updated testing guidance to require traceability between referenced scenarios and test declarations. * Clarified alignment requirements for test preconditions, steps, and expected results. * Documented when special infrastructure markers are required for tests. * **Chores** * Added automated review checks to flag missing scenario coverage and required infrastructure markers. <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Signed-off-by: rnetser <rnetser@redhat.com>
…dHatQE#5118) ##### What this PR does / why we need it: replace pexpect.spawn with pty+subprocess to avoid forkpty - Replace pexpect.spawn with pty.openpty() + subprocess.Popen to avoid deprecated os.forkpty() in Python 3.12+ multi-threaded processes - Add _spawn_console() to create a pty pair and wrap the subprocess in pexpect.fdpexpect.fdspawn - Add _terminate_proc() to cleanly shut down the subprocess, call it in both the connect error path and disconnect cleanup - Simplify console_eof_sampler() — remove parameters, use _spawn_console() internally ##### Which issue(s) this PR fixes: DeprecationWarning on `test_vm_console` ##### Special notes for reviewer: The following warning was shown: `/tmp/pr-tests/.local/share/uv/python/cpython-3.14.5-linux-x86_64-gnu/lib/python3.14/pty.py:66: DeprecationWarning: This process (pid=110) is multi-threaded, use of forkpty() may lead to deadlocks in the child.` ##### jira-ticket: <!-- full-ticket-url needs to be provided. This would add a link to the pull request to the jira and close it when the pull request is merged If the task is not tracked by a Jira ticket, just write "NONE". --> NONE <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Refactor** * Updated the VM console connection to use a PTY-backed subprocess workflow for more reliable interactive sessions. * Improved connection teardown and EOF/sampling behavior, ensuring console sessions terminate cleanly and reducing orphaned processes. * **Tests** * Expanded console unit test coverage for PTY/subprocess setup, error cleanup during connect, graceful vs forced termination, and updated EOF sampling expectations. <!-- end of auto-generated comment: release notes by coderabbit.ai --> Signed-off-by: rkishner <rkishner@redhat.com> Co-authored-by: Cursor <cursoragent@cursor.com>
…QE#5546) ##### What this PR does / why we need it: Move `validate_metrics_value` from `tests/observability/utils.py` to `utilities/monitoring.py` for cross-directory reuse. Now accepts `str | int` expected_value (int 0 for absent metrics). Add unit tests. Update all observability callers. ##### Which issue(s) this PR fixes: Allows another packages to verify metrics tests. ##### Special notes for reviewer: ##### jira-ticket: NONE <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Refactor** * Consolidated metric polling and validation into a shared monitoring helper for consistent observability checks. * Updated observability-related tests (including upgrade scenarios) to use the shared helper. * Removed duplicated metric-validation logic from the test utilities. * **Tests** * Added unit tests covering successful metric matching, type-sensitive comparisons, and timeout/error behavior. <!-- end of auto-generated comment: release notes by coderabbit.ai --> Signed-off-by: Harel Meir <hmeir@redhat.com>
##### Which issue(s) this PR fixes: Removes the skip marker for the closed bug CNV-88747. ##### Special notes for reviewer: The previously blocked test passes with the Jira marker removed. ##### jira-ticket: CNV-88747 Signed-off-by: Yoss Segev <ysegev@redhat.com> Co-authored-by: Yossi Segev <ysegev@ysegev-thinkpadp1gen8.raanaii.csb>
##### What this PR does / why we need it: Move virt-domain fixtures from the monolithic `tests/conftest.py` into domain-scoped modules under `tests/fixtures/virt/`: - `instance_types.py` — instance type and VM preference fixtures - `migration.py` — migration policy and `migrated_vm_multiple_times` fixtures - `osinfo.py` — libosinfo DB download fixtures - `ssh.py` — SSH key generation fixtures - `templates.py` — `base_templates` fixture - `virtual_machines.py` — VM creation fixtures, `rhel_latest_os_params`, `vm_for_migration_test` Register new modules in `pytest_plugins` in root `conftest.py`. ##### Which issue(s) this PR fixes: ##### Special notes for reviewer: Part of series of `tests/conftest.py` refactor. ##### jira-ticket: https://redhat.atlassian.net/browse/CNV-80951 <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Tests** * Reorganized virtualization test setup into focused areas for VM lifecycle, migration, instance types, OS information, SSH access, templates, and VM actions. * Added reusable setup for VM creation and migration scenarios, SSH access, template retrieval, and process management. * Updated test fixture discovery and corrected plugin handling. <!-- end of auto-generated comment: release notes by coderabbit.ai --> Signed-off-by: vsibirsk <vsibirsk@redhat.com>
RedHatQE#6412) ##### What this PR does / why we need it: VirtPlatformAutopilotMachineConfigUpdateStaged is an informational alert so doesn't need a runbook. Add a skip list to `test_runbook_url` that excludes the alert ##### Which issue(s) this PR fixes: Failures in `test-pytest-cnv-5.0-observability-ocs` ##### Special notes for reviewer: ##### jira-ticket: https://redhat.atlassian.net/browse/CNV-98304 <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Tests** * Informational machine configuration update alerts are now validated without requiring a runbook URL. Other alerts continue to be checked for their runbook URLs, preserving existing validation behavior for those alerts. <!-- end of auto-generated comment: release notes by coderabbit.ai --> Signed-off-by: albarker-rh <albarker@redhat.com>
##### What this PR does / why we need it: - Implements the `TestVMTemplateUpgrade` test class, converting STD placeholder tests into working upgrade-continuity tests. - Adds conftest.py, constants.py, and utils.py for the vm_template test module, - add `VM_TEMPLATE_NODE_ID_PREFIX in upgrade_params.py. ##### Which issue(s) this PR fixes: The source VM is deleted after `VirtualMachineTemplateRequest` captures it, it doesn't need to persist across the upgrade ##### Special notes for reviewer: **STP P2 scenario exclusions (capability-disable and disable→reenable recovery):** These two scenarios are not covered in this PR because the VM Template capability is enabled by default in CNV 4.23 / OpenShift Virtualization 5.0+. Testing the disable/reenable flow is only meaningful on 4.22 where the feature flag can be toggled. A separate follow-up covering those scenarios on the 4.22 stream will be tracked in [CNV-97433](https://redhat.atlassian.net/browse/CNV-97433). ##### jira-ticket: <!-- full-ticket-url needs to be provided. This would add a link to the pull request to the jira and close it when the pull request is merged If the task is not tracked by a Jira ticket, just write "NONE". --> https://redhat.atlassian.net/browse/CNV-96840 <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Tests** * Expanded upgrade-continuity coverage for VM templates, template requests, and virtual machines. * Added checks that templates remain ready and can create new VMs after an upgrade. * Improved test reliability with shared setup, readiness checks, and ordered pre- and post-upgrade scenarios. <!-- end of auto-generated comment: release notes by coderabbit.ai --> Signed-off-by: rkishner <rkishner@redhat.com>
…stop status (RedHatQE#6280) ##### What this PR does / why we need it: Wires the shared `dv_stop_status_restart_threshold` stop-status guard into the `wait_for_dv_success()` calls in `tests/storage/general/` whose effective timeout is >= 10 minutes: the `TIMEOUT_12MIN` importer wait in `conftest.py` and the default-timeout (10 min) `fedora_data_volume` wait in `test_storage_behavior.py`. Without the guard these waits block until the full timeout when an importer restarts excessively; reusing the existing shared helper (already used by `cdi_clone`/`golden_image`) makes them fail fast. ##### Which issue(s) this PR fixes: https://issues.redhat.com/browse/CNV-94468 ##### Special notes for reviewer: No new function was added — the existing shared `tests/storage/stop_status_utils.py:dv_stop_status_restart_threshold` is reused. Only calls with an effective timeout >= 10 minutes are guarded; the `TIMEOUT_5MIN` wait is intentionally left unchanged. No behavior change on the success path. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> ##### jira-ticket: https://issues.redhat.com/browse/CNV-94468 <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Tests** * Improved storage test reliability by waiting for data volume operations to reach the expected restart threshold. * Updated test setup and validation to use the relevant data volume context when monitoring operation completion. <!-- end of auto-generated comment: release notes by coderabbit.ai --> Signed-off-by: Adam Cinko <acinko@redhat.com> Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
##### What this PR does / why we need it: In DataVolume object, you can pass stop_status_func and stop_status_func_kwargs into .wait_for_dv_success method and make it fail sooner than the set timeout. The stop_status_func is passed in the sampler and checked every second so you can inject checking DV status or if the DV creation restarted 3 times, etc. and stop it preliminary based on real status, not timeout. This will reduce the time for some storage tests because the method won't have to solely wait for timeout to fail. This implements the DV Stop status function in `cdi_upload` module. ##### Which issue(s) this PR fixes: https://redhat.atlassian.net/browse/CNV-94181 ##### Special notes for reviewer: Signed-off-by: Claude Opus 4.6 [noreply@anthropic.com](mailto:noreply@anthropic.com) ##### jira-ticket: <!-- full-ticket-url needs to be provided. This would add a link to the pull request to the jira and close it when the pull request is merged If the task is not tracked by a Jira ticket, just write "NONE". --> <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Tests** * Updated automated checks for upload completion to account for DataVolumes that may restart after reaching a successful state. Coverage now validates this behavior across both upload workflows, helping ensure completion is assessed consistently. No end-user functionality has changed. <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Signed-off-by: Adam Cinko <acinko@redhat.com>
##### What this PR does / why we need it: Implement the two STD placeholders in tests/observability/metrics/rhel9_rhel10_cluster/test_migration_metrics.py, verifying that migration metrics are reported for cross-version live migration between RHCOS 9 and RHCOS 10 worker nodes. Each test uses a dedicated MigrationPolicy with allow_post_copy=False, avoiding CNV-84023 (postcopy fails on RHCOS 10.2) entirely, and checks migration metrics via subtests over a shared metric list plus a duration check. The two metrics affected by CNV-97013 (memory transfer rate, dirty memory rate) dynamically xfail while that bug stays open. Relocate RHCOS9_AFFINITY/RHCOS10_AFFINITY and set_vm_affinity() from the virt team's local utils.py to utilities/constants/cluster.py and utilities/virt.py, since they are now shared by both the virt and observability team test directories. assisted by: claude code claude-sonnet-5 ##### Which issue(s) this PR fixes: ##### Special notes for reviewer: coverage-omitted utilities/virt.py; both branches of get_or_create_golden_image_data_source and get_data_volume_template_dict_with_default_storage_class are exercised by the integration tests in this PR. ##### jira-ticket: https://redhat.atlassian.net/browse/CNV-96865 <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit - **Tests** - Added coverage for migration metrics during RHCOS 9-to-10 and 10-to-9 migrations of RHEL and Windows virtual machines. - Checks that migration start and end times are reported and that other migration metrics have positive values. - Waits for migration success before checking the reported end time. - Marks checks for metrics affected by a known issue as expected failures while that issue remains open. <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Signed-off-by: Ohad <orevah@redhat.com>
This PR contains the following updates: | Package | Type | Update | Change | |---|---|---|---| | [astral-sh/uv](https://redirect.github.com/astral-sh/uv) | uses-with | patch | `0.12.21` → `0.12.23` | --- ### Release Notes <details> <summary>astral-sh/uv (astral-sh/uv)</summary> ### [`v0.12.23`](https://redirect.github.com/astral-sh/uv/blob/HEAD/CHANGELOG.md#01223) [Compare Source](https://redirect.github.com/astral-sh/uv/compare/0.12.22...0.12.23) Released on 2026-10-03. ##### Python - Add CPython 3.15.0rc3 ([#​22164](https://redirect.github.com/astral-sh/uv/pull/22164)) ##### Preview features - Sync from `uv.lock` without a workspace manifest using `uv sync --frozen` with `frozen-lockfile` ([#​22018](https://redirect.github.com/astral-sh/uv/pull/22018)) - Export from `uv.lock` without a workspace manifest using `uv export --frozen` with `frozen-lockfile` ([#​22007](https://redirect.github.com/astral-sh/uv/pull/22007)) - Inspect dependency trees from `uv.lock` without a workspace manifest using `uv tree --frozen` with `frozen-lockfile` ([#​22016](https://redirect.github.com/astral-sh/uv/pull/22016)) - Inspect workspace metadata and optionally sync its environment from `uv.lock` without a workspace manifest using `uv workspace metadata --frozen` with `frozen-lockfile` ([#​22017](https://redirect.github.com/astral-sh/uv/pull/22017), [#​22018](https://redirect.github.com/astral-sh/uv/pull/22018)) ##### Bug fixes - Reject alternate sources for workspace members across conflicting dependency selections, avoiding lockfiles that cannot be installed ([#​22153](https://redirect.github.com/astral-sh/uv/pull/22153)) - Allow x86-64 Python interpreters running under emulation on Windows ARM64 to install compatible `win_amd64` wheels instead of building from source ([#​22099](https://redirect.github.com/astral-sh/uv/pull/22099)) ### [`v0.12.22`](https://redirect.github.com/astral-sh/uv/blob/HEAD/CHANGELOG.md#01222) [Compare Source](https://redirect.github.com/astral-sh/uv/compare/0.12.21...0.12.22) Released on 2026-10-01. ##### Python - Add CPython 3.10.22, 3.11.17, 3.12.15, 3.13.16, and 3.14.8 ([#​22147](https://redirect.github.com/astral-sh/uv/pull/22147)) ##### Enhancements - Accept uppercase release suffixes in wheel platform tags ([#​22113](https://redirect.github.com/astral-sh/uv/pull/22113)) - Record workspace-member default groups in lockfiles ([#​22010](https://redirect.github.com/astral-sh/uv/pull/22010), [#​22103](https://redirect.github.com/astral-sh/uv/pull/22103)) - Record workspace-member dependency-group Python requirements in lockfiles ([#​22044](https://redirect.github.com/astral-sh/uv/pull/22044), [#​22103](https://redirect.github.com/astral-sh/uv/pull/22103)) - Record default groups for non-project workspace roots in lockfiles ([#​22104](https://redirect.github.com/astral-sh/uv/pull/22104)) - Record dependency-group Python requirements for non-project workspace roots in lockfiles ([#​22104](https://redirect.github.com/astral-sh/uv/pull/22104)) - Format URLs and paths consistently in CLI messages ([#​21937](https://redirect.github.com/astral-sh/uv/pull/21937)) - Hide the unsupported `--offline` option from `uv publish` help ([#​22124](https://redirect.github.com/astral-sh/uv/pull/22124)) ##### Preview features - Honor `--no-default-groups` in `uv audit` ([#​22090](https://redirect.github.com/astral-sh/uv/pull/22090)) - Report a clear error when `uv audit` or `uv tool audit` runs offline and hide the unsupported option from help ([#​22114](https://redirect.github.com/astral-sh/uv/pull/22114)) ##### Configuration - Add `UV_PYTHON_ARCH` to select an interpreter architecture independently of its Python version ([#​22098](https://redirect.github.com/astral-sh/uv/pull/22098)) ##### Performance - Reduce uv's binary size by compressing embedded Python download metadata ([#​22126](https://redirect.github.com/astral-sh/uv/pull/22126)) ##### Bug fixes - Verify unchanged requirements against existing lockfile hashes when relocking ([#​22083](https://redirect.github.com/astral-sh/uv/pull/22083)) - Honor dependency-group Python requirements at non-project workspace roots ([#​22101](https://redirect.github.com/astral-sh/uv/pull/22101)) - Use each selected workspace member's recorded default groups during frozen sync ([#​22015](https://redirect.github.com/astral-sh/uv/pull/22015)) - Avoid false entry-point warnings for required workspace members ([#​22112](https://redirect.github.com/astral-sh/uv/pull/22112)) ##### Other changes - Raise the minimum supported Rust version for building uv to 1.97 and update the toolchain to Rust 1.99 ([#​22121](https://redirect.github.com/astral-sh/uv/pull/22121)) </details> --- ### Configuration 📅 **Schedule**: (UTC) - Branch creation - "after 18:00 on Friday" - Automerge - At any time (no schedule defined) 🚦 **Automerge**: Disabled by config. Please merge this manually once you are satisfied. ♻ **Rebasing**: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox. 🔕 **Ignore**: Close this PR and you won't be reminded about this update again. --- - [ ] <!-- rebase-check -->If you want to rebase/retry this PR, check this box --- This PR was generated by [Mend Renovate](https://mend.io/renovate/). View the [repository job log](https://developer.mend.io/github/RedHatQE/openshift-virtualization-tests). <!--renovate-debug:eyJjcmVhdGVkSW5WZXIiOiI0NC4xMjUuMSIsInVwZGF0ZWRJblZlciI6IjQ0LjEyNS4xIiwidGFyZ2V0QnJhbmNoIjoibWFpbiIsImxhYmVscyI6W119--> Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
This PR contains the following updates: | Update | Change | |---|---| | lockFileMaintenance | All locks refreshed | 🔧 This Pull Request updates lock files to use the latest dependency versions. --- ### Configuration 📅 **Schedule**: (UTC) - Branch creation - Between 12:00 AM and 03:59 AM, only on Monday (`* 0-3 * * 1`) - Automerge - At any time (no schedule defined) 🚦 **Automerge**: Disabled by config. Please merge this manually once you are satisfied. ♻ **Rebasing**: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox. 👻 **Immortal**: This PR will be recreated if closed unmerged. Get [config help](https://redirect.github.com/renovatebot/renovate/discussions) if that's undesired. --- - [ ] <!-- rebase-check -->If you want to rebase/retry this PR, check this box --- This PR was generated by [Mend Renovate](https://mend.io/renovate/). View the [repository job log](https://developer.mend.io/github/RedHatQE/openshift-virtualization-tests). <!--renovate-debug:eyJjcmVhdGVkSW5WZXIiOiI0NC4xMTIuMCIsInVwZGF0ZWRJblZlciI6IjQ0LjExMi4wIiwidGFyZ2V0QnJhbmNoIjoibWFpbiIsImxhYmVscyI6W119--> Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
##### What this PR does / why we need it: The descheduler may churn through other eviction candidates before selecting a virt-launcher, so allow more cycles for a VM to migrate off the overloaded node. ##### Which issue(s) this PR fixes: ##### Special notes for reviewer: ##### jira-ticket: <!-- full-ticket-url needs to be provided. This would add a link to the pull request to the jira and close it when the pull request is merged If the task is not tracked by a Jira ticket, just write "NONE". --> <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Tests** * Updated migration checks to allow more time for the descheduler to select a virtual machine for migration, accommodating cases where it processes other eviction candidates first. No user-facing functionality changed. <!-- end of auto-generated comment: release notes by coderabbit.ai --> Signed-off-by: Denys Shchedrivyi <dshchedr@redhat.com> Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
…RedHatQE#6426) ##### What this PR does / why we need it: Move written_files_to_mounted_hotplugged_disks from @pytest.mark.usefixtures into the test parameter list for test_vm_storage_class_migration_with_hotplugged_volume. This ensures the fixture is set up before storage migration runs, so later data-integrity checks are valid. ##### Which issue(s) this PR fixes: ##### Special notes for reviewer: ##### jira-ticket: https://redhat.atlassian.net/browse/CNV-98245 <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Tests** * Updated the hotplugged-volume migration test to request its fixture directly; test inputs and verification remain unchanged. <!-- end of auto-generated comment: release notes by coderabbit.ai --> Signed-off-by: Noam Berko Laor <nberkola@redhat.com>
…edHatQE#6436) ##### What this PR does / why we need it: Remove redundant test Basically, it is an ordinary node-drain migration to the single remaining node, so it is an extra case that adds no distinct coverage. It also requires an additional node drain - which is nice to avoid, especially on compact clusters. ##### Which issue(s) this PR fixes: ##### Special notes for reviewer: ##### jira-ticket: <!-- full-ticket-url needs to be provided. This would add a link to the pull request to the jira and close it when the pull request is merged If the task is not tracked by a Jira ticket, just write "NONE". --> <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Tests** * Removed the test covering VM migration when multiple nodes are unschedulable. <!-- end of auto-generated comment: release notes by coderabbit.ai --> Signed-off-by: Denys Shchedrivyi <dshchedr@redhat.com> Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
…#6358) ##### What this PR does / why we need it: Network session setup (`network_sanity` → `hosts_common_available_ports` → `nodes_active_nics`) crashes with `KeyError: 'bridge'` when a NodeNetworkState interface is typed `linux-bridge` or `ovs-bridge` but has no `bridge` section. That is valid nmstate data (empty OVN bridges such as `br-ex`/`br-int`, or a bridge with no ports). The fixture treated those keys as required, so the entire network lane failed in setup before any test ran. Treat missing `bridge`, `link-aggregation`, and `ipv4` sections as empty so NIC discovery continues. Physical NICs are still marked occupied only when NNS actually lists them as linux-bridge / ovs-bridge / bond ports (the existing filter). ##### Which issue(s) this PR fixes: ##### Special notes for reviewer: Test-automation bug only; not a CNV/nmstate product defect. Fixture-only change in `tests/fixtures/network/node_nics.py` (coverage-omitted path); behavior is exercised by `network_sanity` setup. ##### jira-ticket: NONE <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit - **Bug Fixes** - Network interface test fixtures now handle missing bridge, link-aggregation, port, and IPv4 details without raising errors. Missing optional settings are treated as empty, allowing network checks to continue when those details are absent. Existing behavior is unchanged when network settings are provided. This improves the reliability of checks for configurations with incomplete optional network information. <!-- end of auto-generated comment: release notes by coderabbit.ai --> Signed-off-by: Harshitha MS <harshitha.ms@ibm.com>
…ion (RedHatQE#6307) ##### What this PR does / why we need it: KubeVirt v1.8 (kubevirt/kubevirt#16404) added `direct` and `extended` sub-fields to the Hyper-V TLBFlush feature. These options improve Windows guest performance. This PR adds automation coverage for the new sub-features: - Enables `tlbflush.direct` and `tlbflush.extended` in the Fedora common template `HYPERV_DICT` - Verifies both sub-features in `check_vm_xml_hyperv()` libvirt XML checks, following the existing `stimer.direct` pattern - Adds unit tests for both failure paths - Updates all existing unit test mock data to include the new sub-fields ##### Which issue(s) this PR fixes: ##### Special notes for reviewer: ##### jira-ticket: NONE <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit - **Bug Fixes** - Hyper-V validation now checks the feature settings configured for each virtual machine, including TLB flush direct and extended modes. - Checks detect missing features and subfeatures, verify enabled states and spinlock retry values, and correctly match the configured timer name. - **Tests** - Expanded coverage for feature subsets, missing or disabled settings, and Fedora and Windows virtual machine configurations. <!-- end of auto-generated comment: release notes by coderabbit.ai --> Signed-off-by: Samuel Albershtein <salbersh@redhat.com>
…edHatQE#6430) ##### What this PR does / why we need it: On FIPS-enabled clusters, the `ocp-virt-validation-checkup-rhel9` image uses RHEL 9.8 as base with the system Python interpreter. The system Python links `hashlib` against RHEL's system OpenSSL, which enforces FIPS mode when the cluster has FIPS enabled (`/proc/sys/crypto/fips_enabled = 1`). In FIPS mode, `hashlib.md5()` without `usedforsecurity=False` raises `UnsupportedDigestmodError`, causing all SSH connectivity checks into VMs to fail: ``` Function: rrmngmnt.host.run_command Kwargs: {'command': ['exit'], 'tcp_timeout': 60} Last exception: UnsupportedDigestmodError: [digital envelope routines] unsupported. ``` **Root cause:** `paramiko.pkey.md5` (imported from `hashlib`) is called by: - `PKey.get_fingerprint()` — used for display/logging of host key fingerprints in all host key policies (`AutoAddPolicy`, `RejectPolicy`, `WarningPolicy`) and during SSH client auth - `PKey._read_private_key_pem()` — used for password-protected PEM key decryption CI (Jenkins) uses the `openshift-virtualization-tests:latest` image (Fedora 41 + `uv` standalone Python) which bundles its own OpenSSL that does not inherit the host FIPS policy — hence no failure there. **Upstream status:** Known paramiko bug ([issue RedHatQE#396](paramiko/paramiko#396), open since 2014). Multiple PRs submitted ([RedHatQE#1928](paramiko/paramiko#1928), [RedHatQE#2189](paramiko/paramiko#2189), [RedHatQE#2496](paramiko/paramiko#2496), [RedHatQE#2624](paramiko/paramiko#2624)) — none merged. No fix in `paramiko 5.0.0`. **Fix:** Patch `paramiko.pkey.md5` at import time via `vars()` to use `usedforsecurity=False`. This tells OpenSSL FIPS that the MD5 usage is not for security purposes (fingerprint display/logging only). **Validated on a FIPS-enabled cluster:** `test_clone_from_block_to_fs_using_dv_template` fails without the patch and passes with it on the `ocp-virt-validation-checkup-rhel9` image. ##### Which issue(s) this PR fixes: CNV-98463 ##### Special notes for reviewer: `vars(paramiko.pkey).update(...)` is used instead of direct attribute assignment to avoid mypy `attr-defined`/`method-assign` errors on third-party module patching, without requiring `# type: ignore` (prohibited by AGENTS.md) or a broad `ignore_errors` mypy override. ##### jira-ticket: CNV-98463 🤖 Generated with [Claude Code](https://claude.com/claude-code) <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Chores** * Improved compatibility for automated test runs in FIPS-enabled environments. Application behavior and user-facing functionality are unchanged; this update affects test setup only and does not alter security protections. No new end-user features or other behavior changes are included. <!-- end of auto-generated comment: release notes by coderabbit.ai --> Signed-off-by: Ramón Lobillo <rlobillo@redhat.com> Co-authored-by: Claude Sonnet 4.6 (1M context) <noreply@anthropic.com>
…mespace clone (RedHatQE#6195) ##### What this PR does / why we need it: From QE Storage lead I got comment it may be better to have more granular permissions: "view" + "cdi.kubevirt.io:clone-sourcer", as this is what we will recommend to customers. It would be good to have the same in the tests. Granular built-in ClusterRoles that together grant the least-privilege permissions to clone golden images across namespaces (the combination recommended to customers): read access to the source images (``view``) plus permission to clone from the source PVCs (``cdi.kubevirt.io:clone-source``). ##### Which issue(s) this PR fixes: [CNV-96366](https://redhat.atlassian.net/browse/CNV-96366) ##### Special notes for reviewer: Co-Authored: Opus 4.8 <noreply@anthropic.com> ##### jira-ticket: NONE <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Chores** * Updated validation environment access and resource lookups for Windows image testing. * No user-facing functionality changes. <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Signed-off-by: Adam Cinko <acinko@redhat.com> Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
##### What this PR does / why we need it: In DataVolume object, you can pass stop_status_func and stop_status_func_kwargs into .wait_for_dv_success method and make it fail sooner than the set timeout. The stop_status_func is passed in the sampler and checked every second so you can inject checking DV status or if the DV creation restarted 3 times, etc. and stop it preliminary based on real status, not timeout. This will reduce the time for some storage tests because the method won't have to solely wait for timeout to fail. This implements the DV Stop status function in `cdi_import` module. ##### Which issue(s) this PR fixes: https://redhat.atlassian.net/browse/CNV-94181 ##### Special notes for reviewer: Signed-off-by: Claude Opus 4.6 <noreply@anthropic.com> ##### jira-ticket: <!-- full-ticket-url needs to be provided. This would add a link to the pull request to the jira and close it when the pull request is merged If the task is not tracked by a Jira ticket, just write "NONE". --> <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit - **Tests** - Updated automated checks to account for DataVolume stop-and-restart thresholds while waiting for successful HTTP and public registry imports. - Applied the updated wait behavior to secure-image, authenticated, blank-disk, and low-capacity import scenarios, as well as relevant VM and pod workflows. - Preserved existing timeout settings for the affected waits. <!-- end of auto-generated comment: release notes by coderabbit.ai --> Signed-off-by: Adam Cinko <acinko@redhat.com>
…edHatQE#6206) ##### What this PR does / why we need it: - Add support for --upgrade_custom=cnv flag to retrieve CNV version explorer URL - Mirrors existing logic for --upgrade eus/cnv flags - Add unit test for upgrade_custom=cnv scenario ##### Which issue(s) this PR fixes: Fixes custom upgrade lane version explorer URL retrieval ##### Special notes for reviewer: ##### jira-ticket: https://redhat.atlassian.net/browse/CNV-96592 <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Bug Fixes** * CNV Version Explorer URLs are now recognized for standard CNV upgrades and custom CNV upgrades, alongside existing install and EUS upgrade configurations. * **Tests** * Added coverage for standard CNV upgrades and custom CNV and EUS upgrade options. <!-- end of auto-generated comment: release notes by coderabbit.ai --> Signed-off-by: Harel Meir <hmeir@redhat.com> Co-authored-by: Cursor <cursoragent@cursor.com>
##### What this PR does / why we need it: update main to point to 5.1 ##### Which issue(s) this PR fixes: ##### Special notes for reviewer: ##### jira-ticket: <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Chores** * Updated the package version to 5.1. * Refreshed issue references and test tracking for the 5.1 release. * **User Impact** * No user-facing behavior changes are included in this update. <!-- end of auto-generated comment: release notes by coderabbit.ai --> Signed-off-by: vsibirsk <vsibirsk@redhat.com>
##### What this PR does / why we need it: This standardizes traceability for every new product test: each test must carry its own exact `STP: <URL>` when an STP exists, or a `Jira: <URL>` fallback when none exists. Module- or class-only links do not cover a new test. Manual placeholders are classified with `@pytest.mark.manual`, have effective `__test__ = False`, and remove both controls when automated; manual tests are registered. The STD guide, `AGENTS.md`, and CodeRabbit rules are synchronized, including enclosing suppression-change checks and unchanged placeholder siblings. Utility/tooling tests are exempt. The follow-up commit expands the change from the four policy/configuration files to exactly nine files by adding 40 direct `@pytest.mark.manual` markers to five existing STD files: - `tests/network/upgrade/test_udn.py`: 2 markers (UDN network upgrade) - `tests/storage/cbt/test_cbt.py`: 10 markers (CBT) - `tests/storage/storage_migration/test_storage_class_migration_cleanup.py`: 16 markers (migration cleanup) - `tests/storage/storage_migration/test_storage_class_online_offline_migration.py`: 8 markers (online/offline migration) - `tests/virt/node/gpu/vgpu/test_rhel_vm_with_mig_vgpu.py`: 4 markers (MIG vGPU) These are direct decorator-only classifications of existing tests. Their docstrings, test bodies, and effective `__test__ = False` behavior are unchanged; there are no new test definitions and no legacy traceability rewrite. There is no utility behavior change, so no new utility unit tests are needed. ##### Which issue(s) this PR fixes: ##### Special notes for reviewer: The exact live diff is four policy/configuration files (`.coderabbit.yaml`, `AGENTS.md`, `docs/SOFTWARE_TEST_DESCRIPTION.md`, and `pytest.ini`) plus the five existing STD files listed above: exactly nine changed files and 40 added manual markers. This policy maintenance and decorator-only STD classification is not a new feature. No issue was created, per the user decision. Existing legacy links are preserved. The exact follow-up commit is `8c8daa485923d352f2266aff7df94cc3d8146dd4`, with tree `e0fba76b24285f91e46e469f5816200aca60cc7e`; hooks ran. Validation passed in isolation against the upstream baseline: `uv run --with pre-commit pre-commit run --all-files`, live schema validation, gitleaks, full tox across six environments, 888 utility tests at 98.95% coverage, and 16 STD-example checks. The `--with` tools were not project-installed and no hook bypass was used. No cluster execution is claimed. Six focused reviews found the change clean; no additional review round was requested. No new utility tests are needed because utility behavior is unchanged. ##### jira-ticket: NONE <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Documentation** * Updated the software test guide with clearer requirements for linking product tests to their specifications or issue records. * Clarified how manual test placeholders are marked, kept out of automated test runs, and updated when automation is added. * Specified which test categories are covered by these requirements and how existing test links should be preserved. * **Chores** * Updated pre-merge checks to validate test traceability and manual-placeholder requirements. <!-- end of auto-generated comment: release notes by CodeRabbit --> --------- Signed-off-by: rnetser <rnetser@redhat.com> Co-authored-by: PI (gpt-5.6-luna) <noreply@pi.dev>
…d Windows 11 WSL2 (RedHatQE#6089) ##### What this PR does / why we need it: Windows 10 has reached End of Life.(Refer - https://support.microsoft.com/en-us/windows/deployment/updates-lifecycle/windows-10-support-has-ended-on-october-14-2025) This removes win-10 from the test matrix and replaces all Windows 10 test references with Windows 11. - Removed `win-10` from `windows_os_list` (amd64 and multiarch configs) - Removed the `win-10` entry from `WINDOWS_OS_MAPPING` in `utilities/os_utils.py` - Removed `WINDOWS_10`, `WINDOWS_10_TEMPLATE_LABELS`, `WINDOWS_10_WSL` constants - Replaced Windows 10 with Windows 11 in: GPU passthrough, vGPU, OOM, memory-load migration, WSL2, tablet device, and guest tools tests - Consolidated WSL2 longevity params to Windows 11 only (changed from list to dict since only one entry remains) - Removed the Win 10 network interface name conditional in longevity MTU setup (now always uses "Ethernet Instance 0") - Updated unit tests accordingly Image constants (`WIN10_IMG`, etc.) and base template existence checks for `windows10` are intentionally retained - the images still exist on infrastructure and KubeVirt still ships the templates. ##### Which issue(s) this PR fixes: ##### Special notes for reviewer: ##### jira-ticket: https://issues.redhat.com/browse/CNV-77964 <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit - **Changes** - Windows 10 is no longer included in supported Windows test configurations or image options. - Windows 11 is used for WSL2, GPU passthrough, migration, guest tools, memory-load, OOM, and longevity scenarios. - Windows test configurations cover Windows Server 2019, Windows 11, Windows Server 2022, and Windows Server 2025. - **Tests** - Updated operating-system mapping and matrix checks to reflect the revised Windows coverage. - **Documentation** - Updated the IBM Cloud Windows test example to use Windows 11. <!-- end of auto-generated comment: release notes by coderabbit.ai --> Signed-off-by: akeskar <akeskar@redhat.com>
##### What this PR does / why we need it: These tests inherited their STP or Jira link from the module or class docstring and now repeat it in their own test docstring, satisfying the per-test traceability rule merged in RedHatQE#6473. ##### Which issue(s) this PR fixes: ##### Special notes for reviewer: This is documentation-only (docstring traceability lines; no test logic, assertions, markers or fixtures changed). Previously undocumented tests received accurate STD docstrings derived from their actual behavior. This PR is split by team folder; sibling PRs exist for the other teams. Validation: mechanical verification confirmed every inserted link matches its enclosing module/class URL and Jira lines retain the inline `# <skip-jira-utils-check>`; all branches compile; pre-commit passed 16/16 hooks on all five branches; full tox passed on representative branches with utilities unit tests 888 passed at 98.95% coverage. ##### jira-ticket: NONE Signed-off-by: rnetser <rnetser@redhat.com>
Ahmad-Hafe
force-pushed
the
cnv-96772-snapshot-restore-mixed-rhcos
branch
from
October 8, 2026 22:32
714b47e to
9bc6b96
Compare
… (CNV-96772) Software Test Description for snapshot/restore test scenarios across RHCOS versions. Defines test structure, docstrings, preconditions, steps, and expected results for: - test_snapshot_rhcos9_restore_rhcos10: Snapshot on RHCOS 9 → Restore on RHCOS 10 - test_snapshot_rhcos10_restore_rhcos9: Snapshot on RHCOS 10 → Restore on RHCOS 9 Includes fixtures and test framework setup. STP: https://github.com/RedHatQE/openshift-virtualization-tests-design-docs/blob/main/stps/sig-virt/dual-stream-cluster-rhcos9-rhcos10/storage.md Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com> Signed-off-by: Ahmad Hafe <ahafe@redhat.com>
Ahmad-Hafe
force-pushed
the
cnv-96772-snapshot-restore-mixed-rhcos
branch
2 times, most recently
from
October 8, 2026 22:34
d2923bf to
a744d94
Compare
Ahmad-Hafe
pushed a commit
that referenced
this pull request
Oct 8, 2026
…ages (#1…" (RedHatQE#2128) This reverts commit 78770b8.
Ahmad-Hafe
force-pushed
the
cnv-96772-snapshot-restore-mixed-rhcos
branch
from
October 8, 2026 22:36
a744d94 to
a54c84b
Compare
Ahmad-Hafe
pushed a commit
that referenced
this pull request
Oct 14, 2026
…#6418) ##### What this PR does / why we need it: `test_restore_multiple_namespaces` failed while the restored guest was still booting. `Console` treated the kernel banner `#1 SMP` as a shell prompt, skipped login, and typed the file check into the boot console. Pytest then reported only the `disconnect()` timeout. This was introduced by RedHatQE#6049, which correctly accepts an already-logged-in session but reused the bare `#` and `$` prompt patterns. The default prompt now requires a trailing space (`# ` / `$ `), so a real prompt still matches and `#1 SMP` does not. The login wait is 2 minutes unless the caller passes `timeout`. A missing prompt fails once. `EOF` while `virtctl` is not up still retries. ##### Which issue(s) this PR fixes: ##### Special notes for reviewer: ##### jira-ticket: NONE Made with [Cursor](https://cursor.com) <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Bug Fixes** * Console commands retain a 30-second default timeout, while login and shell-prompt waits default to two minutes. Virtual-machine console readiness uses the two-minute login timeout. * Shell prompt detection is more precise, reducing the chance that startup messages are mistaken for a ready prompt. * Connection attempts retry at ten-second intervals for up to five minutes. Connection failures trigger cleanup. * Disconnect failures are reported after successful sessions without replacing an error already raised during the session. <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Signed-off-by: Dalia Frank <dafrank@redhat.com>
Ahmad-Hafe
force-pushed
the
cnv-96772-snapshot-restore-mixed-rhcos
branch
from
October 14, 2026 12:13
a54c84b to
d2923bf
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Add two test stubs for snapshot/restore scenarios across RHCOS 9 and RHCOS 10. Tests marked with @pytest.mark.mixed_os_nodes. Jira: https://redhat.atlassian.net/browse/CNV-96772