Compare commits

..
22 Commits
Author SHA1 Message Date
e91b3171ea Release v6.6.0 (#1261)
* Release v6.6.0

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* bump Makefile and README example VERSION to 6.6.0

* docs: regenerate for 6.6.0 and fix stable-6 links

- Regenerate module documentation with collection_prep
- Update helm module docs for reuse_values/reset_values changes
- Add wait_for_jobs, copy_timeout, metadata_only parameters
- Restore ansible_turbo_mode.rst (not auto-generated)
- Fix README.md links to point to stable-6 branch

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Sonnet 4.5 <noreply@anthropic.com>
Co-authored-by: Mandar Kulkarni <mandkulk@redhat.com>
2026-09-16 12:28:23 +01:00
693e80389b fix(helm): validate reuse_values against reset_values and replace (#1230) (#1262)
* fix(helm): validate reuse_values against reset_values and replace

reset_values defaults to true and helm ignores --reuse-values whenever
--reset-values is passed, so the documented reuse_values example was a
no-op. Warn rather than flip the default, which would silently change
what existing playbooks deploy.

helm install accepts neither --reuse-values nor --reset-then-reuse-values,
so pairing either with replace could only ever make helm report an unknown
flag. Fail up front instead, and keep both flags inside the upgrade branch.

Fixes #1229

* fix(docstring): neither nor replacing



* fix(helm): warn instead of failing on replace + reuse_values

reuse_values has been accepted since 3.0.0, and a converged release never
reaches deploy(), so these combinations pass in playbooks today. Failing
in main() would turn them red on upgrade. The flags are already dropped
from the install path, so warning leaves a valid command behind.

Splits the changelog fragment per user-visible change, and covers the
helm diff path, whose 'not replace' guards the hard failure had made
unreachable.

Changes based on the @abikouo's feedback

---------


(cherry picked from commit 10584acdbb)

Co-authored-by: Yuriy Novostavskiy <yuriy@novostavskiy.kyiv.ua>
Co-authored-by: Don Naro <dnaro@redhat.com>
2026-09-16 11:47:18 +01:00
53cb65a740 :fix(helm_repository): normalize both sides of the repo URL comparison (#1236) (#1258)
helm stores repository URLs verbatim, so a repository registered with a
trailing slash keeps it in repositories.yaml. Only the URL coming from the
task was normalized, so a repository that already matched was reported as a
name conflict.

Fixes #1235

(cherry picked from commit 094240ab77)

Co-authored-by: Yuriy Novostavskiy <yuriy@novostavskiy.kyiv.ua>
2026-09-15 11:03:49 -07:00
a2803df8e8 Add deprecation warnings for the status.values return value (#1241) (#1256)
* chore: add deprecation warning for status.values

* docs: add changelog entries for status.values deprecation

(cherry picked from commit f42c78ce83)

Co-authored-by: Don Naro <dnaro@redhat.com>
2026-09-15 16:44:53 +01:00
1016f80aee Add ansible-core version check to k8s lookup turbo mode (#1253) (#1254)
* feat(lookup): add ansible-core version check for turbo mode in k8s lookup

Add ansible-core version checking to the k8s lookup plugin to match the
behavior already implemented in module_utils/ansiblemodule.py.

When ENABLE_TURBO_MODE is set and ansible-core >= 2.19.0:
- Emit deprecation warning for ENABLE_TURBO_MODE
- Emit version warning explaining turbo mode is unsupported
- Disable turbo mode and fall back to standard LookupBase

This prevents confusing tracebacks when cloud.common (which declares
requires_ansible '>=2.15.0,<2.19') is used on newer ansible-core versions
where it is not fully compatible.

Changes:
- Import ansible_version and LooseVersion
- Add TURBO_UNSUPPORTED_CORE = "2.19.0" constant
- Check version before enabling turbo mode
- Disable turbo mode and warn if version >= 2.19.0

Addresses reviewer feedback on PR #1251 to add version testing to the
lookup plugin matching the module implementation.



* test(lookup): add unit tests for k8s turbo mode version check

Add comprehensive unit tests for the ansible-core version checking logic
in the k8s lookup plugin's turbo mode handling.

Tests validate that when ENABLE_TURBO_MODE is set:
- Turbo mode is disabled on ansible-core >= 2.19.0 (tests 2.19.0 and 2.21.4)
- Turbo mode is enabled on ansible-core < 2.19.0 (tests 2.18.0)
- Appropriate deprecation and version warnings are shown
- Standard LookupBase is used when turbo mode is disabled

Tests also verify that when ENABLE_TURBO_MODE is not set:
- No deprecation or warning messages are shown
- Turbo mode remains disabled
- Standard LookupBase is used

Implementation follows pytest best practices:
- Autouse fixture for module cleanup between tests
- Parametrized tests to avoid duplication (2.19.0 and 2.21.4 share logic)
- Mock assertion methods (assert_called, assert_not_called)
- Module-level import of StandardLookupBase to avoid repetition

Simplified and maintainable: 71 lines vs initial 139-line implementation
(49% reduction through deduplication and better abstractions).



* style: format with black

* fixup! test(lookup): add unit tests for k8s turbo mode version check

* docs: add changelog fragment for k8s lookup turbo version check

Add changelog entry documenting that the k8s lookup plugin now ignores
ENABLE_TURBO_MODE on ansible-core >= 2.19.0 and falls back to standard
LookupBase instead of failing with a traceback.

This complements the existing deprecate-enable-turbo-mode.yml fragment
which covers the module implementation, by documenting the lookup plugin
fix specifically.



* fixup! fixup! test(lookup): add unit tests for k8s turbo mode version check

* fixup! fixup! fixup! test(lookup): add unit tests for k8s turbo mode version check

---------


(cherry picked from commit c46180bb6f)

Co-authored-by: Don Naro <dnaro@redhat.com>
Co-authored-by: Claude Sonnet 4.5 <noreply@anthropic.com>
2026-09-15 16:42:50 +01:00
90e868aa37 ci: add the PE certification checker workflow (#1231) (#1249)
* ci: add the PE certification checker workflow

* ci: add stable branches to certification workflow

(cherry picked from commit 732925b986)

Co-authored-by: Don Naro <dnaro@redhat.com>
2026-09-15 14:25:49 +01:00
864c7addd1 k8s_cp - wait for remote tar to finish before closing the stream (#1217) (#1252)
* fix(k8s_cp): wait for remote tar to finish before closing the stream

The module closed the exec connection right after writing the last chunk,
killing tar mid-extract, and never checked its exit status, so a partial
or failed copy was reported as success (#776).

The v4 exec protocol cannot half-close stdin, and busybox tar blocks for
EOF rather than stopping at the end-of-archive marker, so bound the stream
with `head -c N | tar ...`. tar then gets a clean EOF, exits on its own,
and its status is read from the error channel. Containers without /bin/sh
and head keep the previous behaviour and get a warning.

Also add copy_timeout (default 300s, #621) and stream the archive from the
temp file instead of buffering every chunk in memory.

* test(k8s_cp): split combined assert into separate statements

Addresses review feedback on #1217.

* chore(changelog): link the PR and use single-line fragment entries

* fix(linter): reformat with black

* fix(linter): sort imports in test_copy.py

isort --profile black puts ansible_collections in the same third-party
section as pytest, so the blank line between them is a failure.

* fix(k8s_cp): report what tar said when the copy fails early

An early close or a stall is the symptom; tar has usually already
written the cause (a read-only mount, a missing directory) to stderr,
and drain() has it. Discarding it left only the generic truncation or
timeout message.

* refactor(k8s_cp): move the stderr suffix into _fail

The nested fail() helper added a branch inside a closure, which counts
double for nesting and took _stream_tar_to_pod over Sonar's cognitive
complexity limit (14 -> 16). _fail already had the response; give it
the stderr too.

* test(k8s_cp): tempfile for the repeated-copy test directory



---------


(cherry picked from commit 8a9dd54fe7)

Co-authored-by: Yuriy Novostavskiy <yuriy@novostavskiy.kyiv.ua>
Co-authored-by: Bikouo Aubin <79859644+abikouo@users.noreply.github.com>
2026-09-15 10:53:45 +01:00
021ffbfec2 Add a deprecation notice for ENABLE_TURBO_MODE (#1242) (#1250)
* Add a deprecation notice for ENABLE_TURBO_MODE

* minor fix

* fix(turbo): ensure deprecation warning reaches user and skip turbo on unsupported ansible-core

Move the ENABLE_TURBO_MODE deprecation warning to module import time,
since AnsibleTurboModule.__init__() exits the process before returning,
making the previous per-instance deprecate() call unreachable. Also
skip turbo mode on ansible-core >= 2.19.0, where cloud.common is not
supported and the import succeeding could otherwise fail later with
an unrelated traceback, and warn instead of silently falling back
when cloud.common is not installed.

Assisted-by: Claude Sonnet 5

* docs(turbo): clarify that turbo mode is ignored on ansible-core 2.19+

Turbo mode is now bypassed with a warning rather than failing on
unsupported ansible-core, so drop the "may lead to fails" wording from
the README note and the matching statement in the turbo mode guide.

Assisted-by: Claude Opus 5


---------


(cherry picked from commit 031bc475fa)

Co-authored-by: Mandar Kulkarni <mandkulk@redhat.com>
Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
2026-09-15 09:19:32 +01:00
b40a401ec2 fix(helm): skip Helm's OCI registry progress messages on stdout (#1223) (#1247)
* fix(helm): skip OCI registry progress messages on stdout

Helm 4.2.1 moved the registry client's progress messages from stderr to
stdout (helm/helm#32056), so on an OCI reference they now precede the real
output. helm parses the stdout of `helm show chart` and helm_template
returns the stdout of `helm template`, and neither can detect this from the
return code because both commands exit 0.

Helm rewrites '+' to '_' in OCI tags, which makes it also print a free-text
explanation. That is not valid YAML, so a chart version carrying semver
build metadata failed outright. Without the '+' the messages parse as a
mapping instead, silently adding Pulled/Digest keys to the chart info.

Match helm's message verbatim rather than seeking the start of the real
output: the payload has three shapes and can be empty, whereas the messages
are always at the front and byte-identical. A reworded message then fails
loudly instead of consuming a line of real output.

Fixes: #1222

* test(helm): cover OCI charts whose version has build metadata

Pushes a chart versioned 1.0.0+build.1 to the local registry and asserts
that helm installs it idempotently with the version reported verbatim, and
that helm_template returns only rendered manifests. A second version without
build metadata covers the case where the registry messages parse as YAML.

Pinned to Helm 4.2.4: the target is only meaningful on >= 4.2.1, which is
where the messages moved to stdout. The rest of the helm matrix tops out at
v4.0.0, which is why this went unnoticed.

Verified to fail on the parent commit and pass on the fix.

* test(helm): drop Go template from chart fixture

The yamllint sanity test parses chart templates with libyaml, where
`{{ default "test" .Values.myValue }}` is an unhashable mapping key. The
existing chart fixtures carry a yamllint!skip entry in every
tests/sanity/ignore-*.txt for exactly this, but the target never renders
the chart with values, so the template served no purpose. Making the
fixture plain YAML keeps the ignore lists from growing.

* test(ci): empty commit to re-trigger CI

(cherry picked from commit c098ecc8c8)

Co-authored-by: Yuriy Novostavskiy <yuriy@novostavskiy.kyiv.ua>
2026-09-14 14:33:17 +01:00
e620b92b49 [PR #1240/cb00cf3c backport][stable-6] Address stale deprecations (#1245)
* Address stale deprecations (#1240)

* chore: update deprecation for wait_timeout

* chore: update deprecation for return_code

* docs(changelogs): add stale deprecations to changelogs

* docs: ignore pylint deprecation warnings

* chore: update deprecation for merge_type

* docs: update changelogs for merge_type

* docs: ignore more pylint deprecation warnings

(cherry picked from commit cb00cf3c9e)

* fix: remove unnecessary ignore statements

---------

Co-authored-by: Don Naro <dnaro@redhat.com>
2026-09-14 13:49:58 +01:00
40123060c1 chore(test): add ansible 2.21 to integration tests matrix (#1234) (#1243)
* chore(test): add ansible 2.21 to integration tests matrix

current version of Ansible (as the time of this commit) is
2.21 that is included in the Ansible 14 release.

To ensure testing coverage for actual version if ansible,
version 2.21 should be included in the integration tests matrix.

Updated matrix includes:
- stable-2.18 as the latest version support turbo mode
- stable-2.21 as the actual version
- milestone as the development version

* chore(test): exclude 2.18/turbo off from integration tests matrix

stable-2.18 is used mostly as the last version with turbo mode,
so no reason to test it with turbo mode disabled as it tested with
the later version of ansible.

Excluding to decrease CI time/cost

(cherry picked from commit 7df5e01974)

Co-authored-by: Yuriy Novostavskiy <yuriy@novostavskiy.kyiv.ua>
2026-09-14 13:21:20 +01:00
d44a8501e6 test(sanity): align python_requires with documented 3.9+ support (#1226) (#1227)
tests/config.yml declared ">=3.6", so ansible-test ran the module import
test on target Python versions this collection never supported, while
README states 3.9+. Narrow the declaration and add ignore entries for the
supported target versions that were missing.

(cherry picked from commit d9e0218dba)

Co-authored-by: Yuriy Novostavskiy <yuriy@novostavskiy.kyiv.ua>
2026-09-10 17:27:53 +01:00
c3f7a453fd fix(meta): add missing modules to helm and k8s action groups, document them (#1216) (#1221)
* fix(meta): add missing modules to helm and k8s action groups

`helm_plugin` and `helm_plugin_info` both build their argument spec from
`HELM_AUTH_ARG_SPEC`, and `k8s_taint` builds its own from `AUTH_ARG_SPEC`,
but none of the three were listed in the action groups. Anyone setting
`module_defaults` on `group/kubernetes.core.helm` or
`group/kubernetes.core.k8s` silently got no defaults for them.

Deliberately left out of the groups, since they do not accept the shared
authentication options and would make `module_defaults` fail for every
user of the group:

  - `helm_pull` and `helm_template` take only `binary_path`
  - `helm_registry_auth` takes `binary_path` plus its own required `host`,
    which is an OCI registry URL rather than a Kubernetes API server
  - `kubeconfig` writes a kubeconfig file locally and takes no cluster
    connection options

Also documents both action groups in the scenario guide, including the
exclusions above and the distinction from the legacy `k8s` group that
ansible-core defines, which is a different group and does not cover this
collection's modules.

Closes #577

* test(meta): guard action group membership against argspec drift

The action groups in `meta/runtime.yml` were maintained by hand, so
`helm_plugin`, `helm_plugin_info` and `k8s_taint` were able to inherit the
shared authentication arg specs without anyone adding them to a group.
Nothing failed, users just silently got no `module_defaults`.

Asserts the invariant in both directions: a module belongs to a group if
and only if it builds its argument spec from that group's shared constant.
The reverse direction matters too, since a member that does not accept the
shared options makes `module_defaults` fail with "Unsupported parameters"
for every user of the group.

Membership is read statically with `ast`, so the test needs no imports and
runs without a cluster or the `kubernetes` package.

The four modules that are deliberately excluded are listed with their
reasons, so restoring one has to be a decision rather than an oversight.

* fix(tests): satisfy pylint disallowed-name and set an explicit license

ansible-test's pylint profile rejects `_` as a name, which the sanity job
flagged three times. Rather than renaming the throwaways, give the entries
in SHARED_ARG_SPECS named keys so there is nothing to discard when reading
them back.

The header was copied from a sibling test that carries Apache-2.0 because
it derives from kubernetes/kubernetes, so it named the wrong copyright
holder. It now names the author, and keeps Apache-2.0 deliberately: the
collection requirements allow code outside plugins/ that imports no
GPL-3.0-or-later code to use any GPL-3.0-compatible license, and this file
imports only ast, os, pytest and yaml, since it reads group membership
statically rather than importing the modules. A comment records that
reasoning so the constraint is visible if an import is ever added.

* fix(doc): update changelog fragment with the manintainer suggestion

* fix(doc): show helm group defaults in the action group example

abikouo asked for the module_defaults example to cover the helm action
group as well, with tasks that actually consume those defaults. Set
binary_path, kubeconfig and context on group/kubernetes.core.helm and
add helm_repository, helm and helm_info tasks that rely on them; the
namespace in the k8s tasks now matches the release namespace so the
playbook reads as one flow.



---------


(cherry picked from commit 0bc99117af)

Co-authored-by: Yuriy Novostavskiy <yuriy@novostavskiy.kyiv.ua>
Co-authored-by: Bikouo Aubin <79859644+abikouo@users.noreply.github.com>
2026-09-10 17:24:42 +01:00
5c06a554a8 Use collections.abc (#1057) (#1220)
* use collections.abc

ansible.module_utils.common._collections_compat shall be deprecated in
ansible-core 2.24

Use collections.abc

Link: https://forum.ansible.com/t/deprecation-warnings-in-ansible-core-2-20/44741


* Apply suggestion from @abikouo

* Fix linter errors

---------




(cherry picked from commit c643719085)

Signed-off-by: Plamerdi Makela <plamerdi447@gmail.com>
Co-authored-by: Plamerdi Makela <plamerdi447@gmail.com>
Co-authored-by: Bikouo Aubin <79859644+abikouo@users.noreply.github.com>
Co-authored-by: Bianca Henderson <bianca@redhat.com>
2026-09-10 17:23:52 +01:00
20b9c3fb7f Fix ansible-lint no-changed-when and execution-environment schema warnings (#1196) (#1237)
* Fix ansible-lint no-changed-when and execution-environment schema warnings

Add `changed_when: false` to example command tasks in the kubectl
connection plugin to resolve no-changed-when warnings from ansible-lint.

Add `version: 3` to meta/execution-environment.yml to satisfy the
schema[execution-environment] rule requiring the version property.



* Add changelog fragment for ansible-lint warning fixes



* Fix changelog fragment YAML formatting

Quote strings to avoid YAML parsing ambiguity in the trivial section.



---------


(cherry picked from commit a7922e498c)

Co-authored-by: Daniel Brennand <52419383+dbrennand@users.noreply.github.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-09-10 15:26:41 +01:00
7b17313ace Support wait_for_jobs flag for helm release upgrade/install (#1140) (#1214)
* Support wait_for_jobs flag for helm release upgrade/install

* Update plugins/modules/helm.py



* Apply suggestion from @abikouo

* fix(helm): correct wait_for_jobs implementation and add changelog

- Fix double-append of --wait-for-jobs flag
- Fix version check to run before appending the flag
- Fetch helm_version before use to avoid undefined variable
- Fix YAML indentation for version_added in docs
- Add changelog fragment



* test: add integration tests for wait_for_jobs

- Test that --wait-for-jobs is appended when wait=true and wait_for_jobs=true
- Test that wait_for_jobs fails with appropriate message on helm < 3.5.0
- Test that wait_for_jobs is ignored when wait is not set
- Add test-chart-job helm chart containing a simple Job resource



* fix: harden test Job pod spec

Disable service account token automounting and add memory limit to
the test Job container.



* fix: add resource requests and pin image version in test Job

Add CPU, memory, and ephemeral-storage requests. Pin busybox image
to version 1.37.



* style: apply linter formatting to helm argument_spec



* Update sanity ignore file

* fix: quote Helm template expression to pass yamllint sanity



* Fix linters errors

---------





(cherry picked from commit e475d06403)

Co-authored-by: Daniel Simko <danosimko1@gmail.com>
Co-authored-by: Bikouo Aubin <79859644+abikouo@users.noreply.github.com>
Co-authored-by: Yuriy Novostavskiy <yuriy@novostavskiy.kyiv.ua>
Co-authored-by: aubin bikouo <abikouo@redhat.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-08-11 12:11:51 -04:00
569bae6593 Remove remaining ansible.module_utils.six import from kubectl connection plugin (#1197) (#1200)
* Replace ansible.module_utils.six with Python stdlib equivalents

ansible.module_utils.six is deprecated in ansible-core 2.21 and is
scheduled for removal in 2.24. Every use of it in this collection has a
direct Python 3 standard library equivalent, so switch to those.

* Update changelogs/fragments/20260723-remove-module-utils-six.yaml



---------


(cherry picked from commit dd4ca7a8b0)

Co-authored-by: Zhiwei Liang <zhiwei.liang@zliang.me>
Co-authored-by: Bikouo Aubin <79859644+abikouo@users.noreply.github.com>
2026-08-11 12:54:54 +02:00
da42348358 ci: exclude test code from SonarCloud duplication detection (#1207) (#1213)
Duplication is scored on new code only, so a PR adding cases to an already
repetitive test module inherits debt accrued across earlier PRs rather than
introducing it; tests/unit/modules/test_module_helm.py sits at 47.9% on main
before any change.

sonar.exclusions filters the source set, not the copy/paste detector, so test
code needs its own exclusion. Production code stays covered.

(cherry picked from commit f31ecce59c)

Co-authored-by: Yuriy Novostavskiy <yuriy@novostavskiy.kyiv.ua>
2026-08-10 12:01:19 -04:00
79e98c3d6e feat(waiter): add native wait support for Job resources (#1202) (#1211)
Add a `job_complete` predicate that waits for Job resources to reach
either `Complete` or `Failed` condition, removing the need for custom
wait_condition workarounds.

Also rename shadowed `match` variable to `match_condition` in
`custom_condition` to avoid masking the Python built-in.

Fixes ansible-collections/kubernetes.core#1201


(cherry picked from commit 3e58e3a9ad)

Co-authored-by: Bikouo Aubin <79859644+abikouo@users.noreply.github.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-08-10 16:49:59 +02:00
477bb79305 feat(helm): add cleanup_on_fail option (#1206) (#1210)
* feat(helm): add cleanup_on_fail option

atomic rolls back to the last successful revision but can leave behind
resources the failed upgrade created; helm accepts both flags together so
the rollback also drops those orphans.

Rejected with replace=true, since that deploys via helm install, which has
no --cleanup-on-fail. No version gate: the flag exists in helm 3.0.0, the
module's minimum, and was not renamed in v4.

* refactor(tests): share helm command-building test helpers

The new cleanup_on_fail tests copied setUp and _run_with_version verbatim
from the atomic tests, which SonarQube flagged as duplication on new code.
Extract both to module level so the two classes share them.

Scoped to the classes involved in that duplication; the other test classes
in this file keep their own fixtures.

* test(helm): align cleanup_on_fail assertions with review feedback

Adjust the Helm unit tests to assert the expected flags for
cleanup_on_fail on Helm v3 and v4, matching the behavior clarified
in review feedback.



---------


(cherry picked from commit d4ef10fc80)

Co-authored-by: Yuriy Novostavskiy <yuriy@novostavskiy.kyiv.ua>
Co-authored-by: abikouo <abikouo@users.noreply.github.com>
2026-08-10 16:41:33 +02:00
fd65a4f09e Handle empty template output gracefully (#1044) (#1205)
* Handle empty template output gracefully

* Add changelog fragment for empty template output fix

* refactor: reduce cognitive complexity of ActionModule.run

Extract compound boolean condition to a variable to reduce the
cognitive complexity from 16 to the allowed maximum of 15.



* refactor: extract duplicated error handling into _fail_with_error

Replace two identical error-handling blocks in ActionModule.run() with
a shared _fail_with_error static method to address SonarQube code
duplication finding.



* refactor: reduce code duplication in action plugin and tests

Revert _fail_with_error extraction that increased duplication on new
code. Consolidate duplicate integration test tasks for conditional
template into a single looped task.



---------



(cherry picked from commit ec8942dd90)

Co-authored-by: W. H. Wang <mattwang44@gmail.com>
Co-authored-by: aubin bikouo <abikouo@redhat.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-08-10 16:29:44 +02:00
b9749c94f9 Support for metadata-only fetches in k8s_info module (#1030) (#1203)
* Support for metadata-only fetches in k8s_info module

* Fix integration test assert

* Add missing changelog fragment

* Fix linters and sanity checks

* Fix linters

* Apply suggestion from @abikouo

* fix: resolve linter errors after rebase

Fix import sorting and apply black formatting to k8s_info.py and
args_common.py to pass tox -e linters.



---------




(cherry picked from commit 08a761fc17)

Co-authored-by: Federico Chiacchiaretta <federico.chia@gmail.com>
Co-authored-by: Bikouo Aubin <79859644+abikouo@users.noreply.github.com>
Co-authored-by: aubin bikouo <abikouo@redhat.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-08-07 11:03:57 +02:00
70 changed files with 2778 additions and 214 deletions
+20
View File
@@ -0,0 +1,20 @@
---
name: Run collection certification checks
permissions:
contents: read
on:
pull_request:
branches:
- main
- stable-*
workflow_dispatch:
concurrency:
group: cert-ver-${{ github.head_ref || github.run_id }}
cancel-in-progress: true
jobs:
call:
uses: ansible-collections/partner-certification-checker/.github/workflows/certification-reusable.yml@f4bf7ba4d3f008b2db8fc6a812fa3f89fdf42c2e # v4.1.0
+11 -2
View File
@@ -56,15 +56,24 @@ jobs:
fail-fast: false
matrix:
# Ref must match a branch/tag on github.com/ansible/ansible (e.g. stable-2.18, not 2.18).
ansible-version: ["stable-2.18", "milestone"]
#
ansible-version: ["stable-2.18", "stable-2.21", "milestone"]
enable-turbo-mode: [true, false]
python-version: ["3.12", "3.13"]
exclude:
# stable-2.18 is used mainly because turbo mode, no reason to test it with turbo mode disabled
- ansible-version: "stable-2.18"
enable-turbo-mode: false
# turbo mode is supported only by ansible<2.19
- ansible-version: "milestone"
enable-turbo-mode: true
# stable-2.18 supports 3.12; milestone (devel) requires >=3.13
- ansible-version: "stable-2.21"
enable-turbo-mode: true
# stable-2.18 supports 3.12
- ansible-version: "stable-2.18"
python-version: "3.13"
- ansible-version: "stable-2.21"
python-version: "3.12"
- ansible-version: "milestone"
python-version: "3.12"
workflow-id: ${{ fromJson(needs.splitter.outputs.test_jobs) }}
+42
View File
@@ -4,6 +4,48 @@ Kubernetes Collection Release Notes
.. contents:: Topics
v6.6.0
======
Minor Changes
-------------
- Remove the remaining ``ansible.module_utils.six`` import to avoid deprecation warnings, replacing it with the Python standard library equivalent (https://github.com/ansible-collections/kubernetes.core/pull/1197).
- helm - Add ``wait_for_jobs`` option to wait for all Jobs to complete before marking a Helm release as successful. Requires Helm >= 3.5.0 (https://github.com/ansible-collections/kubernetes.core/pull/1140).
- helm - add the ``cleanup_on_fail`` option, mapping to the ``--cleanup-on-fail`` flag, to allow deletion of new resources created during a failed upgrade. It complements ``atomic`` and can be combined with it, but cannot be used with ``replace``, since that deploys through ``helm install`` which does not accept the flag (https://github.com/ansible-collections/kubernetes.core/pull/1206).
- helm - warn when ``reuse_values`` or ``reset_then_reuse_values`` is requested in a combination helm ignores, which happens by default because ``reset_values`` defaults to ``true``, and whenever either is combined with ``replace``. The option descriptions and the examples now state that ``reset_values`` has to be set to ``false`` for either option to take effect (https://github.com/ansible-collections/kubernetes.core/pull/1230).
- k8s lookup - warn when ``ENABLE_TURBO_MODE`` is set but the ``cloud.common`` collection is not installed, instead of silently falling back to the standard lookup base (https://github.com/ansible-collections/kubernetes.core/pull/1242).
- k8s_cp - Add the ``copy_timeout`` option (default 300 seconds) bounding how long the module will spend streaming an archive to a pod and waiting for the remote ``tar`` to finish, so that a stalled copy fails instead of hanging indefinitely (https://github.com/ansible-collections/kubernetes.core/pull/1217).
- k8s_cp - Stream the tar archive to the pod chunk by chunk instead of building the whole archive in memory first, which cuts peak memory use when copying large files (https://github.com/ansible-collections/kubernetes.core/pull/1217).
- k8s_cp - When copying to a pod, the module now also uses ``/bin/sh`` and ``head`` in the container, when present, to confirm the copy completed. Containers without them keep the previous behaviour and get a warning that completion could not be verified (https://github.com/ansible-collections/kubernetes.core/pull/1217).
- k8s_info - Support for metadata-only fetches in k8s_info module (https://github.com/ansible-collections/kubernetes.core/pull/1030)
- meta - Add ``helm_plugin`` and ``helm_plugin_info`` to the ``helm`` action group, and ``k8s_taint`` to the ``k8s`` action group, so that ``module_defaults`` set on those groups applies to them (https://github.com/ansible-collections/kubernetes.core/pull/1216).
- use `collections.abc` instead of deprecated `ansible.module_utils.common._collections_compat` (https://github.com/ansible-collections/kubernetes.core/pull/1057).
- waiter - Add ``job_complete`` predicate to support waiting for Job resources to reach ``Complete`` or ``Failed`` condition (https://github.com/ansible-collections/kubernetes.core/issues/1201).
Deprecated Features
-------------------
- Ansible Turbo mode (``ENABLE_TURBO_MODE``) has been deprecated and will be removed in release 8.0.0, as it depends on the ``cloud.common`` collection, which is being retired (https://github.com/ansible-collections/kubernetes.core/pull/1242).
- helm - the ``status.values`` return value has been deprecated and will be removed in version 8.0.0. Use ``status.release_values`` instead (https://github.com/ansible-collections/kubernetes.core/issues/1239).
- helm - the ``wait_timeout`` parameter has been deprecated and will be removed in version 7.0.0. Use ``timeout`` instead (https://github.com/ansible-collections/kubernetes.core/issues/1239).
- helm_info - the ``status.values`` return value has been deprecated and will be removed in version 8.0.0. Use ``status.release_values`` instead (https://github.com/ansible-collections/kubernetes.core/issues/1239).
- k8s_exec - the ``return_code`` return value has been deprecated and will be removed in version 7.0.0. Use ``rc`` instead (https://github.com/ansible-collections/kubernetes.core/issues/1239).
- k8s_service - the ``merge_type=json`` option has been deprecated and will be removed in version 7.0.0. Use ``kubernetes.core.k8s_json_patch`` module instead (https://github.com/ansible-collections/kubernetes.core/issues/1239).
Bugfixes
--------
- Ansible Turbo mode - ignore ``ENABLE_TURBO_MODE`` on ansible-core 2.19.0 and later, where the ``cloud.common`` collection is not supported, and fall back to the standard ``AnsibleModule`` instead of failing with ``byte indices must be integers or slices, not str`` (https://github.com/ansible-collections/kubernetes.core/pull/1242).
- helm - do not pass the upgrade-only ``--reuse-values`` and ``--reset-then-reuse-values`` flags to ``helm install``, which rejected them with ``unknown flag`` whenever ``reuse_values`` or ``reset_then_reuse_values`` was combined with ``replace`` (https://github.com/ansible-collections/kubernetes.core/pull/1230).
- helm - fix the ``reuse_values`` example, which set only ``reuse_values=true``. ``reset_values`` defaults to ``true`` and helm ignores ``--reuse-values`` whenever ``--reset-values`` is passed, so following the example reset the release values instead of reusing them (https://github.com/ansible-collections/kubernetes.core/pull/1230).
- helm - skip Helm's OCI registry progress messages when parsing the output of ``helm show chart``. As of Helm 4.2.1 these are printed to stdout instead of stderr (https://github.com/helm/helm/pull/32056), which made installing a chart from an OCI registry fail with a YAML scanner error when the chart version contained a ``+`` (https://github.com/ansible-collections/kubernetes.core/pull/1223).
- helm_repository - normalize both sides of the repository URL comparison, so that a repository already registered with a trailing slash is recognized as matching instead of failing with ``Repository already have a repository named <name>`` (https://github.com/ansible-collections/kubernetes.core/pull/1236).
- helm_template - strip Helm's OCI registry progress messages from the returned ``stdout``, which is documented to contain only the rendered templates (https://github.com/ansible-collections/kubernetes.core/pull/1223).
- k8s lookup - ignore ``ENABLE_TURBO_MODE`` on ansible-core 2.19.0 and later, where the ``cloud.common`` collection is not supported. Fall back to the standard ``LookupBase`` instead of failing with a traceback (https://github.com/ansible-collections/kubernetes.core/pull/1253).
- k8s_cp - Fix silent truncation when copying to a pod. The module closed the exec connection immediately after writing the last chunk of the tar archive, which killed the remote ``tar`` before it had finished extracting, and never checked the process exit status, so a partial copy was reported as a success. The archive is now bounded with ``head -c`` so ``tar`` gets a clean EOF and exits on its own, and the module waits for that exit and fails on a non-zero status (https://github.com/ansible-collections/kubernetes.core/pull/1217).
- k8s_info - Handle empty template output gracefully by returning ``changed=false`` instead of failing when Jinja2 template renders to an empty string (https://github.com/ansible-collections/kubernetes.core/issues/1042).
v6.5.0
======
+1 -1
View File
@@ -1,5 +1,5 @@
# Also needs to be updated in galaxy.yml
VERSION = 6.5.0
VERSION = 6.6.0
TEST_ARGS ?= ""
PYTHON_VERSION ?= `python -c 'import platform; print(".".join(platform.python_version_tuple()[0:2]))'`
+28 -26
View File
@@ -22,7 +22,7 @@ Click on the name of a plugin or module to view that content's documentation:
### Connection plugins
Name | Description
--- | ---
[kubernetes.core.kubectl](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.kubectl_connection.rst)|Execute tasks in pods running on Kubernetes.
[kubernetes.core.kubectl](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.kubectl_connection.rst)|Execute tasks in pods running on Kubernetes.
### K8s filter plugins
Name | Description
@@ -32,33 +32,33 @@ kubernetes.core.k8s_config_resource_name|Generate resource name for the given re
### Lookup plugins
Name | Description
--- | ---
[kubernetes.core.k8s](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.k8s_lookup.rst)|Query the K8s API
[kubernetes.core.kustomize](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.kustomize_lookup.rst)|Build a set of kubernetes resources using a 'kustomization.yaml' file.
[kubernetes.core.k8s](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.k8s_lookup.rst)|Query the K8s API
[kubernetes.core.kustomize](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.kustomize_lookup.rst)|Build a set of kubernetes resources using a 'kustomization.yaml' file.
### Modules
Name | Description
--- | ---
[kubernetes.core.helm](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.helm_module.rst)|Manages Kubernetes packages with the Helm package manager
[kubernetes.core.helm_info](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.helm_info_module.rst)|Get information from Helm package deployed inside the cluster
[kubernetes.core.helm_plugin](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.helm_plugin_module.rst)|Manage Helm plugins
[kubernetes.core.helm_plugin_info](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.helm_plugin_info_module.rst)|Gather information about Helm plugins
[kubernetes.core.helm_pull](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.helm_pull_module.rst)|download a chart from a repository and (optionally) unpack it in local directory.
[kubernetes.core.helm_registry_auth](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.helm_registry_auth_module.rst)|Helm registry authentication module
[kubernetes.core.helm_repository](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.helm_repository_module.rst)|Manage Helm repositories.
[kubernetes.core.helm_template](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.helm_template_module.rst)|Render chart templates
[kubernetes.core.k8s](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.k8s_module.rst)|Manage Kubernetes (K8s) objects
[kubernetes.core.k8s_cluster_info](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.k8s_cluster_info_module.rst)|Describe Kubernetes (K8s) cluster, APIs available and their respective versions
[kubernetes.core.k8s_cp](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.k8s_cp_module.rst)|Copy files and directories to and from pod.
[kubernetes.core.k8s_drain](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.k8s_drain_module.rst)|Drain, Cordon, or Uncordon node in k8s cluster
[kubernetes.core.k8s_exec](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.k8s_exec_module.rst)|Execute command in Pod
[kubernetes.core.k8s_info](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.k8s_info_module.rst)|Describe Kubernetes (K8s) objects
[kubernetes.core.k8s_json_patch](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.k8s_json_patch_module.rst)|Apply JSON patch operations to existing objects
[kubernetes.core.k8s_log](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.k8s_log_module.rst)|Fetch logs from Kubernetes resources
[kubernetes.core.k8s_rollback](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.k8s_rollback_module.rst)|Rollback Kubernetes (K8S) Deployments and DaemonSets
[kubernetes.core.k8s_scale](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.k8s_scale_module.rst)|Set a new size for a Deployment, ReplicaSet, Replication Controller, or Job.
[kubernetes.core.k8s_service](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.k8s_service_module.rst)|Manage Services on Kubernetes
[kubernetes.core.k8s_taint](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.k8s_taint_module.rst)|Taint a node in a Kubernetes/OpenShift cluster
[kubernetes.core.kubeconfig](https://github.com/ansible-collections/kubernetes.core/blob/main/docs/kubernetes.core.kubeconfig_module.rst)|Generate, update, and optionally write Kubernetes kubeconfig files
[kubernetes.core.helm](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.helm_module.rst)|Manages Kubernetes packages with the Helm package manager
[kubernetes.core.helm_info](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.helm_info_module.rst)|Get information from Helm package deployed inside the cluster
[kubernetes.core.helm_plugin](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.helm_plugin_module.rst)|Manage Helm plugins
[kubernetes.core.helm_plugin_info](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.helm_plugin_info_module.rst)|Gather information about Helm plugins
[kubernetes.core.helm_pull](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.helm_pull_module.rst)|download a chart from a repository and (optionally) unpack it in local directory.
[kubernetes.core.helm_registry_auth](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.helm_registry_auth_module.rst)|Helm registry authentication module
[kubernetes.core.helm_repository](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.helm_repository_module.rst)|Manage Helm repositories.
[kubernetes.core.helm_template](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.helm_template_module.rst)|Render chart templates
[kubernetes.core.k8s](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.k8s_module.rst)|Manage Kubernetes (K8s) objects
[kubernetes.core.k8s_cluster_info](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.k8s_cluster_info_module.rst)|Describe Kubernetes (K8s) cluster, APIs available and their respective versions
[kubernetes.core.k8s_cp](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.k8s_cp_module.rst)|Copy files and directories to and from pod.
[kubernetes.core.k8s_drain](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.k8s_drain_module.rst)|Drain, Cordon, or Uncordon node in k8s cluster
[kubernetes.core.k8s_exec](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.k8s_exec_module.rst)|Execute command in Pod
[kubernetes.core.k8s_info](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.k8s_info_module.rst)|Describe Kubernetes (K8s) objects
[kubernetes.core.k8s_json_patch](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.k8s_json_patch_module.rst)|Apply JSON patch operations to existing objects
[kubernetes.core.k8s_log](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.k8s_log_module.rst)|Fetch logs from Kubernetes resources
[kubernetes.core.k8s_rollback](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.k8s_rollback_module.rst)|Rollback Kubernetes (K8S) Deployments and DaemonSets
[kubernetes.core.k8s_scale](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.k8s_scale_module.rst)|Set a new size for a Deployment, ReplicaSet, Replication Controller, or Job.
[kubernetes.core.k8s_service](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.k8s_service_module.rst)|Manage Services on Kubernetes
[kubernetes.core.k8s_taint](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.k8s_taint_module.rst)|Taint a node in a Kubernetes/OpenShift cluster
[kubernetes.core.kubeconfig](https://github.com/ansible-collections/kubernetes.core/blob/stable-6/docs/kubernetes.core.kubeconfig_module.rst)|Generate, update, and optionally write Kubernetes kubeconfig files
<!--end collection content-->
@@ -100,7 +100,7 @@ You can also include it in a `requirements.yml` file and install it via `ansible
---
collections:
- name: kubernetes.core
version: 6.5.0
version: 6.6.0
```
### Installing the Kubernetes Python Library
@@ -181,7 +181,9 @@ For documentation on how to use individual modules and other content included in
## Ansible Turbo Mode Tech Preview
> **Note:** Ansible Turbo mode is supported only with ``ansible-core`` versions **lower than 2.19**. Starting from ``ansible-core`` 2.19, the ``cloud.common`` collection is no longer supported, and therefore Ansible Turbo mode is also not supported in ``kubernetes.core``. If ``ENABLE_TURBO_MODE`` is set to ``1`` (or any truthy value) with ``ansible-core >= 2.19.0``, it may lead to fails.
> **Deprecated:** Ansible Turbo mode (`ENABLE_TURBO_MODE`) is deprecated and will be removed in a future major release (planned for 8.0.0), since it depends on the `cloud.common` collection, which is being retired. Setting `ENABLE_TURBO_MODE` will keep emitting a deprecation warning until support is removed, at which point the variable will have no effect.
> **Note:** Ansible Turbo mode is supported only with ``ansible-core`` versions **lower than 2.19**. Starting from ``ansible-core`` 2.19, the ``cloud.common`` collection is no longer supported, and therefore Ansible Turbo mode is also not supported in ``kubernetes.core``. If ``ENABLE_TURBO_MODE`` is set to ``1`` (or any truthy value) with ``ansible-core >= 2.19.0``, Turbo mode is ignored and a warning is emitted; modules run without it.
The ``kubernetes.core`` collection supports Ansible Turbo mode as a tech preview via the ``cloud.common`` collection (requires `ansible-core < 2.19`). By default, this feature is disabled. To enable Turbo mode for modules, set the environment variable `ENABLE_TURBO_MODE=1` on the managed node. For example:
+109
View File
@@ -1251,3 +1251,112 @@ releases:
name: kubeconfig
namespace: ''
release_date: '2026-07-08'
6.6.0:
changes:
bugfixes:
- Ansible Turbo mode - ignore ``ENABLE_TURBO_MODE`` on ansible-core 2.19.0 and
later, where the ``cloud.common`` collection is not supported, and fall back
to the standard ``AnsibleModule`` instead of failing with ``byte indices must
be integers or slices, not str`` (https://github.com/ansible-collections/kubernetes.core/pull/1242).
- helm - do not pass the upgrade-only ``--reuse-values`` and ``--reset-then-reuse-values``
flags to ``helm install``, which rejected them with ``unknown flag`` whenever
``reuse_values`` or ``reset_then_reuse_values`` was combined with ``replace``
(https://github.com/ansible-collections/kubernetes.core/pull/1230).
- helm - fix the ``reuse_values`` example, which set only ``reuse_values=true``.
``reset_values`` defaults to ``true`` and helm ignores ``--reuse-values``
whenever ``--reset-values`` is passed, so following the example reset the
release values instead of reusing them (https://github.com/ansible-collections/kubernetes.core/pull/1230).
- helm - skip Helm's OCI registry progress messages when parsing the output
of ``helm show chart``. As of Helm 4.2.1 these are printed to stdout instead
of stderr (https://github.com/helm/helm/pull/32056), which made installing
a chart from an OCI registry fail with a YAML scanner error when the chart
version contained a ``+`` (https://github.com/ansible-collections/kubernetes.core/pull/1223).
- helm_repository - normalize both sides of the repository URL comparison, so
that a repository already registered with a trailing slash is recognized as
matching instead of failing with ``Repository already have a repository named
<name>`` (https://github.com/ansible-collections/kubernetes.core/pull/1236).
- helm_template - strip Helm's OCI registry progress messages from the returned
``stdout``, which is documented to contain only the rendered templates (https://github.com/ansible-collections/kubernetes.core/pull/1223).
- k8s lookup - ignore ``ENABLE_TURBO_MODE`` on ansible-core 2.19.0 and later,
where the ``cloud.common`` collection is not supported. Fall back to the standard
``LookupBase`` instead of failing with a traceback (https://github.com/ansible-collections/kubernetes.core/pull/1253).
- k8s_cp - Fix silent truncation when copying to a pod. The module closed the
exec connection immediately after writing the last chunk of the tar archive,
which killed the remote ``tar`` before it had finished extracting, and never
checked the process exit status, so a partial copy was reported as a success.
The archive is now bounded with ``head -c`` so ``tar`` gets a clean EOF and
exits on its own, and the module waits for that exit and fails on a non-zero
status (https://github.com/ansible-collections/kubernetes.core/pull/1217).
- k8s_info - Handle empty template output gracefully by returning ``changed=false``
instead of failing when Jinja2 template renders to an empty string (https://github.com/ansible-collections/kubernetes.core/issues/1042).
deprecated_features:
- Ansible Turbo mode (``ENABLE_TURBO_MODE``) has been deprecated and will be
removed in release 8.0.0, as it depends on the ``cloud.common`` collection,
which is being retired (https://github.com/ansible-collections/kubernetes.core/pull/1242).
- helm - the ``status.values`` return value has been deprecated and will be
removed in version 8.0.0. Use ``status.release_values`` instead (https://github.com/ansible-collections/kubernetes.core/issues/1239).
- helm - the ``wait_timeout`` parameter has been deprecated and will be removed
in version 7.0.0. Use ``timeout`` instead (https://github.com/ansible-collections/kubernetes.core/issues/1239).
- helm_info - the ``status.values`` return value has been deprecated and will
be removed in version 8.0.0. Use ``status.release_values`` instead (https://github.com/ansible-collections/kubernetes.core/issues/1239).
- k8s_exec - the ``return_code`` return value has been deprecated and will be
removed in version 7.0.0. Use ``rc`` instead (https://github.com/ansible-collections/kubernetes.core/issues/1239).
- k8s_service - the ``merge_type=json`` option has been deprecated and will
be removed in version 7.0.0. Use ``kubernetes.core.k8s_json_patch`` module
instead (https://github.com/ansible-collections/kubernetes.core/issues/1239).
minor_changes:
- Remove the remaining ``ansible.module_utils.six`` import to avoid deprecation
warnings, replacing it with the Python standard library equivalent (https://github.com/ansible-collections/kubernetes.core/pull/1197).
- helm - Add ``wait_for_jobs`` option to wait for all Jobs to complete before
marking a Helm release as successful. Requires Helm >= 3.5.0 (https://github.com/ansible-collections/kubernetes.core/pull/1140).
- helm - add the ``cleanup_on_fail`` option, mapping to the ``--cleanup-on-fail``
flag, to allow deletion of new resources created during a failed upgrade.
It complements ``atomic`` and can be combined with it, but cannot be used
with ``replace``, since that deploys through ``helm install`` which does not
accept the flag (https://github.com/ansible-collections/kubernetes.core/pull/1206).
- helm - warn when ``reuse_values`` or ``reset_then_reuse_values`` is requested
in a combination helm ignores, which happens by default because ``reset_values``
defaults to ``true``, and whenever either is combined with ``replace``. The
option descriptions and the examples now state that ``reset_values`` has to
be set to ``false`` for either option to take effect (https://github.com/ansible-collections/kubernetes.core/pull/1230).
- k8s lookup - warn when ``ENABLE_TURBO_MODE`` is set but the ``cloud.common``
collection is not installed, instead of silently falling back to the standard
lookup base (https://github.com/ansible-collections/kubernetes.core/pull/1242).
- k8s_cp - Add the ``copy_timeout`` option (default 300 seconds) bounding how
long the module will spend streaming an archive to a pod and waiting for the
remote ``tar`` to finish, so that a stalled copy fails instead of hanging
indefinitely (https://github.com/ansible-collections/kubernetes.core/pull/1217).
- k8s_cp - Stream the tar archive to the pod chunk by chunk instead of building
the whole archive in memory first, which cuts peak memory use when copying
large files (https://github.com/ansible-collections/kubernetes.core/pull/1217).
- k8s_cp - When copying to a pod, the module now also uses ``/bin/sh`` and ``head``
in the container, when present, to confirm the copy completed. Containers
without them keep the previous behaviour and get a warning that completion
could not be verified (https://github.com/ansible-collections/kubernetes.core/pull/1217).
- k8s_info - Support for metadata-only fetches in k8s_info module (https://github.com/ansible-collections/kubernetes.core/pull/1030)
- meta - Add ``helm_plugin`` and ``helm_plugin_info`` to the ``helm`` action
group, and ``k8s_taint`` to the ``k8s`` action group, so that ``module_defaults``
set on those groups applies to them (https://github.com/ansible-collections/kubernetes.core/pull/1216).
- use `collections.abc` instead of deprecated `ansible.module_utils.common._collections_compat`
(https://github.com/ansible-collections/kubernetes.core/pull/1057).
- waiter - Add ``job_complete`` predicate to support waiting for Job resources
to reach ``Complete`` or ``Failed`` condition (https://github.com/ansible-collections/kubernetes.core/issues/1201).
fragments:
- 1253-k8s-lookup-turbo-version-check.yml
- 20250108-fix-empty-template-output.yaml
- 20251027-metadata-only-fetches.yaml
- 20260109-fix-deprecated-calls.yaml
- 20260723-remove-module-utils-six.yaml
- 20260805-add-job-wait-condition.yaml
- 20260807-helm-cleanup-on-fail.yaml
- 20260807-helm-wait-for-jobs.yaml
- 20260812-action-groups.yaml
- 20260813-k8s-cp-truncation.yaml
- 20260817-helm-oci-registry-progress-on-stdout.yml
- 20260826-helm-value-reuse-options.yaml
- 20260908-helm-repository-trailing-slash.yaml
- 20260911-deprecate-status-values.yaml
- 20260911-update-deprecation-notices.yaml
- deprecate-enable-turbo-mode.yml
- fix-ansible-lint-warnings.yaml
release_date: '2026-09-16'
+8 -1
View File
@@ -12,12 +12,19 @@ Following document provides overview of Ansible Turbo mode in ``kubernetes.core`
:depth: 1
.. warning::
Ansible Turbo mode is **deprecated** and will be removed in a future major release
(planned for 8.0.0), as it depends on the ``cloud.common`` collection, which is being
retired. Setting ``ENABLE_TURBO_MODE`` now emits a deprecation warning; once support is
removed, the variable will have no effect. Start planning your migration away from
Turbo mode.
.. warning::
Ansible Turbo mode is supported only with ``ansible-core`` versions **lower than 2.19**.
Starting from ``ansible-core`` 2.19, the ``cloud.common`` collection is no longer supported,
and therefore Ansible Turbo mode is also not supported in ``kubernetes.core``.
If ``ENABLE_TURBO_MODE`` is set to ``1`` (or any truthy value) with ``ansible-core >= 2.19.0``,
it may lead to failures.
Turbo mode is ignored and a warning is emitted; modules run without it.
Synopsis
--------
@@ -42,6 +42,92 @@ Basic authentication is also supported using the ``username`` and ``password`` o
To disable SSL certificate verification, set ``verify_ssl`` to false.
Sharing options across tasks with action groups
===============================================
Repeating the same authentication options on every task gets verbose quickly. The collection
defines two `action groups <https://docs.ansible.com/ansible/latest/playbook_guide/playbooks_module_defaults.html#module-defaults-groups>`_
so you can set them once with ``module_defaults``:
``kubernetes.core.k8s``
Modules that talk to the Kubernetes API and accept the standard authentication options
(``kubeconfig``, ``context``, ``host``, ``api_key``, ``validate_certs``, and so on):
``k8s``, ``k8s_cluster_info``, ``k8s_cp``, ``k8s_drain``, ``k8s_exec``, ``k8s_info``,
``k8s_json_patch``, ``k8s_log``, ``k8s_rollback``, ``k8s_scale``, ``k8s_service``, and
``k8s_taint``.
``kubernetes.core.helm``
Modules that shell out to the ``helm`` binary and accept its shared authentication
options (``binary_path``, ``kubeconfig``, ``context``, ``host``, ``api_key``,
``ca_cert``, ``validate_certs``): ``helm``, ``helm_info``, ``helm_plugin``,
``helm_plugin_info``, and ``helm_repository``.
Prefix the group name with ``group/`` to use it:
.. code-block:: yaml
- hosts: localhost
module_defaults:
group/kubernetes.core.k8s:
kubeconfig: /path/to/kubeconfig
context: staging
group/kubernetes.core.helm:
binary_path: /path/to/helm
kubeconfig: /path/to/kubeconfig
context: staging
tasks:
- name: Create a namespace
kubernetes.core.k8s:
definition:
apiVersion: v1
kind: Namespace
metadata:
name: monitoring
- name: Read it back
kubernetes.core.k8s_info:
kind: Namespace
name: monitoring
- name: Add the prometheus chart repository
kubernetes.core.helm_repository:
name: prometheus-community
repo_url: https://prometheus-community.github.io/helm-charts
- name: Deploy the kube-prometheus-stack chart
kubernetes.core.helm:
name: kube-prometheus-stack
chart_ref: prometheus-community/kube-prometheus-stack
release_namespace: monitoring
- name: Collect the releases in that namespace
kubernetes.core.helm_info:
name: kube-prometheus-stack
release_namespace: monitoring
Every task picks up ``kubeconfig`` and ``context`` from its group without repeating them, the
``helm`` modules additionally pick up ``binary_path``, and an option set on an individual task
still wins over the group default.
.. note::
Always write the group name fully qualified as ``group/kubernetes.core.k8s``. A bare
``group/k8s`` refers to the legacy ``k8s`` group that ansible-core defined back when the
Kubernetes modules shipped in core itself. The two are not the same group, and the legacy
one does not cover the modules in this collection.
Some modules are deliberately **not** in these groups because they do not accept the shared
authentication options, and including them would make ``module_defaults`` fail for everyone
using the group:
- ``helm_pull`` and ``helm_template`` accept only ``binary_path``.
- ``helm_registry_auth`` accepts ``binary_path`` and its own ``host``, which is an OCI
registry URL rather than a Kubernetes API server.
- ``kubeconfig`` writes a kubeconfig file locally and takes no cluster connection options.
Pass options to those modules directly, or give them their own per-module ``module_defaults``
entry.
Reporting an issue
==================
+1 -1
View File
@@ -488,7 +488,7 @@ Common return values are documented `here <https://docs.ansible.com/projects/ans
<td>always</td>
<td>
<div>Dict of Values used to deploy</div>
<div>This return value has been deprecated and will be removed in a release after 2027-01-08. Use RV(status.release_values) instead.</div>
<div>This return value has been deprecated and will be removed in version 8.0.0. Use RV(status.release_values) instead.</div>
<br/>
</td>
</tr>
+71 -5
View File
@@ -157,6 +157,29 @@ Parameters
<div>Chart version to install. If this is not specified, the latest version is installed.</div>
</td>
</tr>
<tr>
<td colspan="2">
<div class="ansibleOptionAnchor" id="parameter-"></div>
<b>cleanup_on_fail</b>
<a class="ansibleOptionLink" href="#parameter-" title="Permalink to this option"></a>
<div style="font-size: small">
<span style="color: purple">boolean</span>
</div>
<div style="font-style: italic; font-size: small; color: darkgreen">added in 6.6.0</div>
</td>
<td>
<ul style="margin: 0; padding: 0"><b>Choices:</b>
<li><div style="color: blue"><b>no</b>&nbsp;&larr;</div></li>
<li>yes</li>
</ul>
</td>
<td>
<div>Allow deletion of new resources created during a failed upgrade.</div>
<div>Maps to the <code>--cleanup-on-fail</code> flag, which is available since helm 3.0.0 and was not renamed in Helm v4.</div>
<div>Complements O(atomic) rather than replacing it, and the two can be combined so that a rollback also drops resources orphaned by the failed upgrade.</div>
<div>Helm only accepts this flag on <code>helm upgrade</code>, so it cannot be used together with O(replace), which deploys through <code>helm install</code>.</div>
</td>
</tr>
<tr>
<td colspan="2">
<div class="ansibleOptionAnchor" id="parameter-"></div>
@@ -522,7 +545,9 @@ Parameters
</td>
<td>
<div>When upgrading package, reset the values to the ones built into the chart, apply the last release&#x27;s values and merge in any overrides from parameters O(release_values), O(values_files) or O(set_values).</div>
<div>If O(reset_values) or O(reuse_values) is set to V(True), this is ignored.</div>
<div>Helm ignores this when <code>--reset-values</code> or <code>--reuse-values</code> is passed. Since O(reset_values) defaults to V(true), it has to be set to V(false), and O(reuse_values) left unset or V(false), for this option to take any effect.</div>
<div><code>helm install</code> does not accept <code>--reset-then-reuse-values</code>, so this option is ignored when O(replace) is set.</div>
<div>The module warns whenever one of those combinations makes this option a no-op.</div>
<div>This feature requires helm diff &gt;= 3.9.12.</div>
</td>
</tr>
@@ -544,6 +569,7 @@ Parameters
</td>
<td>
<div>When upgrading package, reset the values to the ones built into the chart.</div>
<div>Defaults to V(true), which makes helm ignore O(reuse_values) and O(reset_then_reuse_values). Set it to V(false) to use either of them.</div>
</td>
</tr>
<tr>
@@ -563,8 +589,10 @@ Parameters
</ul>
</td>
<td>
<div>When upgrading package, specifies wether to reuse the last release&#x27;s values and merge in any overrides from parameters <em>release_values</em>, <em>values_files</em> or <em>set_values</em>.</div>
<div>If <em>reset_values</em> is set to <code>True</code>, this is ignored.</div>
<div>When upgrading package, specifies whether to reuse the last release&#x27;s values and merge in any overrides from parameters O(release_values), O(values_files) or O(set_values).</div>
<div>Helm ignores this when <code>--reset-values</code> is passed. Since O(reset_values) defaults to V(true), it has to be set to V(false) for this option to take any effect.</div>
<div><code>helm install</code> does not accept <code>--reuse-values</code>, so this option is ignored when O(replace) is set.</div>
<div>The module warns whenever one of those combinations makes this option a no-op.</div>
</td>
</tr>
<tr>
@@ -811,6 +839,28 @@ Parameters
<div>When <em>release_state</em> is set to <code>absent</code>, will wait until all the resources are deleted before returning. It will wait for as long as <em>wait_timeout</em>. This feature requires helm&gt;=3.7.0. Added in version 2.3.0.</div>
</td>
</tr>
<tr>
<td colspan="2">
<div class="ansibleOptionAnchor" id="parameter-"></div>
<b>wait_for_jobs</b>
<a class="ansibleOptionLink" href="#parameter-" title="Permalink to this option"></a>
<div style="font-size: small">
<span style="color: purple">boolean</span>
</div>
<div style="font-style: italic; font-size: small; color: darkgreen">added in 6.6.0</div>
</td>
<td>
<ul style="margin: 0; padding: 0"><b>Choices:</b>
<li><div style="color: blue"><b>no</b>&nbsp;&larr;</div></li>
<li>yes</li>
</ul>
</td>
<td>
<div>When <em>release_state</em> is set to <code>present</code>, and <em>wait</em> is set to <code>True</code>, wait until all jobs are in a completed state before marking the release as successful.</div>
<div>Ignored when used without O(wait).</div>
<div>Requires Helm &gt;= 3.5.0</div>
</td>
</tr>
<tr>
<td colspan="2">
<div class="ansibleOptionAnchor" id="parameter-"></div>
@@ -824,7 +874,7 @@ Parameters
</td>
<td>
<div>Timeout when wait option is enabled (helm2 is a number of seconds, helm3 is a duration).</div>
<div>The use of <em>wait_timeout</em> to wait for kubernetes commands to complete has been deprecated and will be removed after 2022-12-01.</div>
<div>The use of <em>wait_timeout</em> to wait for kubernetes commands to complete has been deprecated and will be removed in version 7.0.0. Use O(timeout) instead.</div>
</td>
</tr>
</table>
@@ -935,6 +985,7 @@ Examples
release_namespace: default
force: True
wait: True
wait_for_jobs: True
replace: True
update_repo_cache: True
disable_hook: True
@@ -954,16 +1005,31 @@ Examples
enabled: True
# Deploy latest version
# 'reset_values' defaults to true and helm ignores '--reuse-values' when it is set,
# so it has to be turned off explicitly.
- name: Deploy latest version of Grafana chart using reuse_values
kubernetes.core.helm:
name: test
chart_ref: stable/grafana
release_namespace: monitoring
reuse_values: true
reset_values: false
values:
replicas: 2
version: 3e8ec0b2dffa40fb97d5342e4af887de95faa8c61a62480dd7f8aa03dffcf533
# Same for 'reset_then_reuse_values', which helm ignores when either
# '--reset-values' or '--reuse-values' is set.
- name: Deploy latest version of Grafana chart using reset_then_reuse_values
kubernetes.core.helm:
name: test
chart_ref: stable/grafana
release_namespace: monitoring
reset_then_reuse_values: true
reset_values: false
values:
replicas: 2
Return Values
@@ -1152,7 +1218,7 @@ Common return values are documented `here <https://docs.ansible.com/projects/ans
<td>always</td>
<td>
<div>Dict of Values used to deploy.</div>
<div>This return value has been deprecated and will be removed in a release after 2027-01-08. Use RV(status.release_values) instead.</div>
<div>This return value has been deprecated and will be removed in version 8.0.0. Use RV(status.release_values) instead.</div>
<br/>
</td>
</tr>
+20
View File
@@ -152,6 +152,25 @@ Parameters
<div>The name of a context found in the config file. Can also be specified via K8S_AUTH_CONTEXT environment variable.</div>
</td>
</tr>
<tr>
<td colspan="2">
<div class="ansibleOptionAnchor" id="parameter-"></div>
<b>copy_timeout</b>
<a class="ansibleOptionLink" href="#parameter-" title="Permalink to this option"></a>
<div style="font-size: small">
<span style="color: purple">integer</span>
</div>
<div style="font-style: italic; font-size: small; color: darkgreen">added in 6.6.0</div>
</td>
<td>
<b>Default:</b><br/><div style="color: blue">300</div>
</td>
<td>
<div>Maximum time in seconds to spend streaming the archive to the pod and waiting for the remote <code>tar</code> process to finish extracting it.</div>
<div>Raise this when copying large files over a slow connection.</div>
<div>This option is ignored when <em>state</em> is set to <code>from_pod</code>.</div>
</td>
</tr>
<tr>
<td colspan="2">
<div class="ansibleOptionAnchor" id="parameter-"></div>
@@ -512,6 +531,7 @@ Notes
.. note::
- the tar binary is required on the container when copying from local filesystem to pod.
- a POSIX shell at ``/bin/sh`` and ``head`` are also used, when present, to bound the archive sent to the container so that the module can confirm the copy finished. Without them the copy still runs but completion cannot be verified, and a warning is emitted.
- the (init) container has to be started before you copy files or directories to it.
- To avoid SSL certificate validation errors when ``validate_certs`` is *True*, the full certificate chain for the API server must be provided via ``ca_cert`` or in the kubeconfig file.
+1 -1
View File
@@ -535,7 +535,7 @@ Common return values are documented `here <https://docs.ansible.com/projects/ans
</td>
<td></td>
<td>
<div>The command status code. This attribute is deprecated and will be removed in a future release. Please use rc instead.</div>
<div>The command status code. This attribute is deprecated and will be removed in version 7.0.0. Use RV(result.rc) instead.</div>
<br/>
</td>
</tr>
+29
View File
@@ -280,6 +280,29 @@ Parameters
<div>List of label selectors to use to filter results</div>
</td>
</tr>
<tr>
<td colspan="2">
<div class="ansibleOptionAnchor" id="parameter-"></div>
<b>metadata_only</b>
<a class="ansibleOptionLink" href="#parameter-" title="Permalink to this option"></a>
<div style="font-size: small">
<span style="color: purple">boolean</span>
</div>
<div style="font-style: italic; font-size: small; color: darkgreen">added in 6.6.0</div>
</td>
<td>
<ul style="margin: 0; padding: 0"><b>Choices:</b>
<li><div style="color: blue"><b>no</b>&nbsp;&larr;</div></li>
<li>yes</li>
</ul>
</td>
<td>
<div>Request metadata only response from the Kubernetes API server.</div>
<div>This feature is useful for clients that only need to check for the existence of an object,</div>
<div>or that only need to read its metadata.</div>
<div>It can significantly reduce the size of the response from the API server.</div>
</td>
</tr>
<tr>
<td colspan="2">
<div class="ansibleOptionAnchor" id="parameter-"></div>
@@ -716,6 +739,12 @@ Examples
retries: 60
delay: 15
- name: Get a list of all pods metadata only
kubernetes.core.k8s_info:
kind: Pod
metadata_only: true
register: pod_metadata_list
Return Values
@@ -247,6 +247,7 @@ Parameters
<div>See <a href='https://kubernetes.io/docs/tasks/run-application/update-api-object-kubectl-patch/#use-a-json-merge-patch-to-update-a-deployment'>https://kubernetes.io/docs/tasks/run-application/update-api-object-kubectl-patch/#use-a-json-merge-patch-to-update-a-deployment</a></div>
<div>If more than one <code>merge_type</code> is given, the merge_types will be tried in order</div>
<div>This defaults to <code>[&#x27;strategic-merge&#x27;, &#x27;merge&#x27;]</code>, which is ideal for using the same parameters on resource kinds that combine Custom Resources and built-in resources.</div>
<div><em>merge_type=json</em> has been deprecated and will be removed in version 7.0.0. Use <span class='module'>kubernetes.core.k8s_json_patch</span> instead.</div>
</td>
</tr>
<tr>
@@ -383,6 +383,7 @@ Examples
# and requires python to be installed in the image
- name: Run a command in a pod
ansible.builtin.command: echo "Hello, World!"
changed_when: false
- name: Run a command in a pod using local kubectl with inventory variables
# Example inventory:
@@ -407,6 +408,7 @@ Examples
# and requires python to be installed in the image
- name: Run a command in a pod
ansible.builtin.command: echo "Hello, World!"
changed_when: false
- name: Run a command in a pod using dynamic inventory
hosts: localhost
@@ -440,6 +442,7 @@ Examples
# be aware that the command is executed as the user that started the container
# and requires python to be installed in the image
ansible.builtin.command: echo "Hello, World!"
changed_when: false
delegate_to: "{{ my_app_pod_name }}"
+1 -1
View File
@@ -25,7 +25,7 @@ tags:
- openshift
- okd
- cluster
version: 6.5.0
version: 6.6.0
build_ignore:
- .DS_Store
- "*.tar.gz"
+1
View File
@@ -1,4 +1,5 @@
---
version: 3
dependencies:
python: requirements.txt
system: meta/ee-bindep.txt
+8
View File
@@ -1,10 +1,17 @@
---
requires_ansible: '>=2.16.0'
# Only modules that accept the shared authentication options belong in these
# groups: HELM_AUTH_ARG_SPEC for helm, AUTH_ARG_SPEC for k8s. Adding a module
# that does not accept them makes module_defaults fail for every user of the
# group. This excludes helm_pull, helm_registry_auth, helm_template and
# kubeconfig, which only take a subset (or none) of those options.
action_groups:
helm:
- helm
- helm_info
- helm_plugin
- helm_plugin_info
- helm_repository
k8s:
- k8s
@@ -18,6 +25,7 @@ action_groups:
- k8s_rollback
- k8s_scale
- k8s_service
- k8s_taint
plugin_routing:
inventory:
+11 -2
View File
@@ -35,7 +35,11 @@ def _from_yaml_to_definition(buffer):
import yaml
except ImportError:
raise AnsibleError("Failed to import the required Python library (PyYAML).")
return list(yaml.safe_load_all(buffer))
# Handle empty template output
if buffer is None:
return []
return [doc for doc in yaml.safe_load_all(buffer) if doc is not None]
ENV_KUBECONFIG_PATH_SEPARATOR = ";" if platform.system() == "Windows" else ":"
@@ -380,10 +384,15 @@ class ActionModule(ActionBase):
template = self._task.args.get("template", None)
if template:
self.load_template(template, new_module_args, task_vars)
# If template rendered to empty definition, return early with no changes
if not new_module_args.get("definition"):
result["changed"] = False
return self._ensure_invocation(result)
local_path = self._task.args.get("local_path")
state = self._task.args.get("state", None)
if local_path and state == "to_pod" and not remote_transport:
is_local_pod_upload = local_path and state == "to_pod" and not remote_transport
if is_local_pod_upload:
new_module_args["local_path"] = self.get_file_realpath(local_path)
# Execute the k8s_* module.
+4 -1
View File
@@ -196,6 +196,7 @@ EXAMPLES = r"""
# and requires python to be installed in the image
- name: Run a command in a pod
ansible.builtin.command: echo "Hello, World!"
changed_when: false
- name: Run a command in a pod using local kubectl with inventory variables
# Example inventory:
@@ -220,6 +221,7 @@ EXAMPLES = r"""
# and requires python to be installed in the image
- name: Run a command in a pod
ansible.builtin.command: echo "Hello, World!"
changed_when: false
- name: Run a command in a pod using dynamic inventory
hosts: localhost
@@ -253,6 +255,7 @@ EXAMPLES = r"""
# be aware that the command is executed as the user that started the container
# and requires python to be installed in the image
ansible.builtin.command: echo "Hello, World!"
changed_when: false
delegate_to: "{{ my_app_pod_name }}"
"""
@@ -262,11 +265,11 @@ import os.path
import shutil
import subprocess
import tempfile
from shlex import quote as shlex_quote
from ansible.errors import AnsibleError, AnsibleFileNotFound
from ansible.module_utils.common.text.converters import to_bytes
from ansible.module_utils.parsing.convert_bool import boolean
from ansible.module_utils.six.moves import shlex_quote
from ansible.parsing.yaml.loader import AnsibleLoader
from ansible.plugins.connection import BUFSIZE, ConnectionBase
from ansible.utils.display import Display
+33 -1
View File
@@ -209,28 +209,60 @@ RETURN = """
"""
import os
from collections.abc import KeysView
from ansible.errors import AnsibleError
from ansible.module_utils.common._collections_compat import KeysView
from ansible.module_utils.ansible_release import __version__ as ansible_version
from ansible.module_utils.common.validation import check_type_bool
from ansible.utils.display import Display
from ansible_collections.kubernetes.core.plugins.module_utils.k8s.client import (
get_api_client,
)
from ansible_collections.kubernetes.core.plugins.module_utils.k8s.resource import (
create_definitions,
)
from ansible_collections.kubernetes.core.plugins.module_utils.version import (
LooseVersion,
)
display = Display()
# cloud.common declares requires_ansible '>=2.15.0,<2.19'. On newer ansible-core the
# import still succeeds, so the version has to be checked explicitly: relying on
# ImportError alone lets turbo mode fail later with an unrelated traceback.
TURBO_UNSUPPORTED_CORE = "2.19.0"
try:
enable_turbo_mode = check_type_bool(os.environ.get("ENABLE_TURBO_MODE"))
except TypeError:
enable_turbo_mode = False
if enable_turbo_mode:
display.deprecated(
"ENABLE_TURBO_MODE is deprecated, as it relies on the cloud.common "
"collection which is being retired. Setting this environment "
"variable will have no effect once support is removed.",
version="8.0.0",
collection_name="kubernetes.core",
)
if LooseVersion(ansible_version) >= LooseVersion(TURBO_UNSUPPORTED_CORE):
display.warning(
"ENABLE_TURBO_MODE is ignored on ansible-core %s: Ansible Turbo mode requires "
"the cloud.common collection, which supports ansible-core < %s only. "
"Continuing without Turbo mode." % (ansible_version, TURBO_UNSUPPORTED_CORE)
)
enable_turbo_mode = False
if enable_turbo_mode:
try:
from ansible_collections.cloud.common.plugins.plugin_utils.turbo.lookup import (
TurboLookupBase as LookupBase,
)
except ImportError:
display.warning(
"ENABLE_TURBO_MODE is set but the cloud.common collection is not installed. "
"Continuing without Turbo mode."
)
from ansible.plugins.lookup import LookupBase # noqa: F401
else:
from ansible.plugins.lookup import LookupBase # noqa: F401
+30
View File
@@ -5,13 +5,39 @@ __metaclass__ = type
import os
from ansible.module_utils.ansible_release import __version__ as ansible_version
from ansible.module_utils.common.validation import check_type_bool
from ansible.module_utils.common.warnings import deprecate, warn
from ansible_collections.kubernetes.core.plugins.module_utils.version import (
LooseVersion,
)
# cloud.common declares requires_ansible '>=2.15.0,<2.19'. On newer ansible-core the
# import still succeeds, so the version has to be checked explicitly: relying on
# ImportError alone lets turbo mode fail later with an unrelated traceback.
TURBO_UNSUPPORTED_CORE = "2.19.0"
try:
enable_turbo_mode = check_type_bool(os.environ.get("ENABLE_TURBO_MODE"))
except TypeError:
enable_turbo_mode = False
if enable_turbo_mode:
deprecate(
"ENABLE_TURBO_MODE is deprecated, as it relies on the cloud.common collection "
"which is being retired. Setting this environment variable will have no effect "
"once support is removed.",
version="8.0.0",
collection_name="kubernetes.core",
)
if LooseVersion(ansible_version) >= LooseVersion(TURBO_UNSUPPORTED_CORE):
warn(
"ENABLE_TURBO_MODE is ignored on ansible-core %s: Ansible Turbo mode requires "
"the cloud.common collection, which supports ansible-core < %s only. "
"Continuing without Turbo mode." % (ansible_version, TURBO_UNSUPPORTED_CORE)
)
enable_turbo_mode = False
if enable_turbo_mode:
try:
from ansible_collections.cloud.common.plugins.module_utils.turbo.module import ( # noqa: F401
@@ -20,6 +46,10 @@ if enable_turbo_mode:
AnsibleModule.collection_name = "kubernetes.core"
except ImportError:
warn(
"ENABLE_TURBO_MODE is set but the cloud.common collection is not installed. "
"Continuing without Turbo mode."
)
from ansible.module_utils.basic import AnsibleModule # noqa: F401
else:
from ansible.module_utils.basic import AnsibleModule # noqa: F401
+15
View File
@@ -144,3 +144,18 @@ DELETE_OPTS_ARG_SPEC = {
"options": {"resourceVersion": {"type": "str"}, "uid": {"type": "str"}},
},
}
METADATA_ONLY_HEADERS_SPEC = {
"accept": {"type": "str"},
}
METADATA_ONLY_HEADER_VALUES = {
"partial_object": (
"application/json;as=PartialObjectMetadata;g=meta.k8s.io;v=v1,"
"application/json;q=0.9"
),
"partial_object_list": (
"application/json;as=PartialObjectMetadataList;g=meta.k8s.io;v=v1,"
"application/json;q=0.9"
),
}
+178 -53
View File
@@ -18,7 +18,9 @@ from __future__ import absolute_import, division, print_function
__metaclass__ = type
import os
import shlex
import tarfile
import time
from abc import ABCMeta, abstractmethod
from select import select
from tempfile import NamedTemporaryFile, TemporaryFile
@@ -296,14 +298,156 @@ class K8SCopyToPod(K8SCopy):
Copy files/directory from local filesystem into remote Pod
"""
# Size of the chunks the tar archive is written to stdin with.
CHUNK_SIZE = 1024 * 1024
SHELL = "/bin/sh"
def __init__(self, module, client):
super(K8SCopyToPod, self).__init__(module, client)
self.files_to_copy = list()
self.named_temp_file = None
self.copy_timeout = module.params.get("copy_timeout")
def close_temp_file(self):
if self.named_temp_file:
self.named_temp_file.close()
def _fail(self, response, msg, stderr=(), **kwargs):
"""Fail, keeping whatever the remote already wrote to stderr.
An early close or a stall is usually the symptom, not the cause: tar
has typically said why on stderr first.
"""
if stderr:
msg = "{0}: {1}".format(msg, "".join(stderr))
response.close()
self.close_temp_file()
self.module.fail_json(
msg="Failed to copy local file/directory into Pod: {0}".format(msg),
**kwargs,
)
def tar_command(self, dest_file):
if self.no_preserve:
command = [
"tar",
"--no-same-permissions",
"--no-same-owner",
"-xmf",
"-",
]
else:
command = ["tar", "-xmf", "-"]
if dest_file.startswith("/"):
command.extend(["-C", "/"])
return command
def can_bound_stream(self):
"""Whether the container can run ``sh -c 'head -c ...'``."""
error, _out, _err = self._run_from_pod(
cmd=[self.SHELL, "-c", "command -v head"]
)
return (error or {}).get("status") == "Success"
def remote_command(self, dest_file, archive_size):
"""Build the command that extracts the archive inside the container.
The websocket exec protocol the kubernetes client speaks (v4) cannot
half-close stdin, so the only EOF the remote process can get is the
connection going away. Relying on that races with tar finishing and
truncates the file (issue #776), and tar cannot be relied on to stop at
the end-of-archive marker either -- busybox tar blocks for EOF.
Piping through ``head -c`` bounds the stream at the exact archive size,
so tar sees a clean EOF, exits on its own, and the API server reports
its exit status on the error channel. Only where the container has no
shell do we fall back to feeding tar directly, which cannot confirm
that the copy completed.
"""
command = self.tar_command(dest_file)
if not self.can_bound_stream():
self.module.warn(
"Container has no '{0}' with 'head', falling back to writing the"
" archive straight to tar. The module cannot confirm the copy"
" completed; verify the file after copying.".format(self.SHELL)
)
return command, False
return [
self.SHELL,
"-c",
"head -c {0} | {1}".format(
archive_size, " ".join(shlex.quote(arg) for arg in command)
),
], True
def _stream_tar_to_pod(self, response, tar_buffer, wait_for_exit):
"""Feed the archive to the remote extractor and wait for it to finish.
Writing the last chunk only hands the bytes to the local socket; they
still have to cross the API server and reach the container. Closing the
connection at that point kills the extractor mid-write and silently
truncates the file, which is what issue #776 reports.
"""
stdout, stderr = [], []
def drain(timeout):
response.update(timeout=timeout)
if response.peek_stdout():
stdout.append(response.read_stdout().rstrip("\n"))
if response.peek_stderr():
stderr.append(response.read_stderr().rstrip("\n"))
deadline = time.monotonic() + self.copy_timeout
chunk = tar_buffer.read(self.CHUNK_SIZE)
while chunk:
if not response.is_open():
self._fail(
response,
"connection to Pod {0}/{1} closed before the whole archive was"
" sent, the remote file is incomplete".format(
self.namespace, self.name
),
stderr,
)
if time.monotonic() > deadline:
self._fail(
response,
"timed out after {0}s sending the archive to Pod {1}/{2}".format(
self.copy_timeout, self.namespace, self.name
),
stderr,
)
# Non-blocking, keeps the receive buffer clear while we write.
drain(0)
response.write_stdin(chunk)
chunk = tar_buffer.read(self.CHUNK_SIZE)
if not wait_for_exit:
# No way to make the extractor exit, so no status to collect: read
# whatever it has already said and close, as the module always did.
drain(0)
response.close()
return None, stdout, stderr
# The archive is on the wire; wait for the extractor to consume it,
# exit, and have its status reported on the error channel.
while response.is_open():
if time.monotonic() > deadline:
self._fail(
response,
"timed out after {0}s waiting for tar to finish extracting on Pod"
" {1}/{2}".format(self.copy_timeout, self.namespace, self.name),
stderr,
)
drain(1)
error = yaml.safe_load(response.read_channel(ERROR_CHANNEL)) or {}
response.close()
return error, stdout, stderr
def run(self):
# remove trailing slash from destination path
dest_file = self.remote_path.rstrip("/")
@@ -336,65 +480,46 @@ class K8SCopyToPod(K8SCopy):
dest_file = os.path.join(dest_file, os.path.basename(src_file))
if not self.check_mode:
if self.no_preserve:
tar_command = [
"tar",
"--no-same-permissions",
"--no-same-owner",
"-xmf",
"-",
]
else:
tar_command = ["tar", "-xmf", "-"]
if dest_file.startswith("/"):
tar_command.extend(["-C", "/"])
response = stream(
self.api_instance.connect_get_namespaced_pod_exec,
self.name,
self.namespace,
command=tar_command,
stderr=True,
stdin=True,
stdout=True,
tty=False,
_preload_content=False,
**self.container_arg,
)
with TemporaryFile() as tar_buffer:
# Build the archive first: bounding the remote read needs its
# exact size.
with tarfile.open(fileobj=tar_buffer, mode="w") as tar:
tar.add(src_file, dest_file)
archive_size = tar_buffer.tell()
tar_buffer.seek(0)
commands = []
# push command in chunk mode
size = 1024 * 1024
while True:
data = tar_buffer.read(size)
if not data:
break
commands.append(data)
stderr, stdout = [], []
while response.is_open():
if response.peek_stdout():
stdout.append(response.read_stdout().rstrip("\n"))
if response.peek_stderr():
stderr.append(response.read_stderr().rstrip("\n"))
if commands:
cmd = commands.pop(0)
response.write_stdin(cmd)
else:
break
response.close()
if stderr:
self.close_temp_file()
self.module.fail_json(
command=tar_command,
msg="Failed to copy local file/directory into Pod due to: {0}".format(
"".join(stderr)
),
command, wait_for_exit = self.remote_command(dest_file, archive_size)
response = stream(
self.api_instance.connect_get_namespaced_pod_exec,
self.name,
self.namespace,
command=command,
stderr=True,
stdin=True,
stdout=True,
tty=False,
_preload_content=False,
**self.container_arg,
)
error, stdout, stderr = self._stream_tar_to_pod(
response, tar_buffer, wait_for_exit
)
if error is not None and error.get("status") != "Success":
self._fail(
response,
"".join(stderr) or error.get("message", "unknown error"),
command=command,
)
if error is None and stderr:
# Legacy path: no exit status to go on, so any output is fatal.
self._fail(response, "".join(stderr), command=command)
if stderr:
self.module.warn(
"tar wrote to stderr while extracting into Pod {0}/{1}: {2}".format(
self.namespace, self.name, "".join(stderr)
)
)
self.close_temp_file()
if self.content:
self.module.exit_json(
+47
View File
@@ -32,6 +32,53 @@ except ImportError:
HAS_YAML = False
# Helm's registry client writes these progress messages to stdout as of helm 4.2.1
# (helm/helm#32056); earlier versions sent them to stderr. Anything that parses the
# stdout of a helm command which may pull from an OCI registry has to skip them.
# Kept verbatim from helm's pkg/registry/client.go so that a reworded message fails
# to match, and therefore fails loudly, rather than silently eating real output.
REGISTRY_PROGRESS_RE = re.compile(r"^(?:Pulled|Pushed|Digest): \S")
REGISTRY_UNDERSCORE_SUFFIX = " contains an underscore.\n"
REGISTRY_UNDERSCORE_MESSAGE = """
OCI artifact references (e.g. tags) do not support the plus sign (+). To support
storing semantic versions, Helm adopts the convention of changing plus (+) to
an underscore (_) in chart version tags when pushing to a registry and back to
a plus (+) when pulling from a registry.
"""
REGISTRY_UNDERSCORE_MESSAGE_LINES = len(REGISTRY_UNDERSCORE_MESSAGE.splitlines())
def strip_registry_progress(out):
"""
Drop OCI registry progress messages from the front of a helm command's stdout.
Helm emits 'Pulled:'/'Pushed:' and 'Digest:' lines, followed - when the
reference contains an underscore - by a free-text explanation that is not
valid YAML. Since helm 4.2.1 these land on stdout ahead of the real output,
so 'helm show chart' can no longer be parsed and 'helm template' no longer
returns only rendered manifests.
"""
lines = out.splitlines(keepends=True)
index = 0
while index < len(lines):
if REGISTRY_PROGRESS_RE.match(lines[index]):
index += 1
elif lines[index].endswith(REGISTRY_UNDERSCORE_SUFFIX):
# helm prints its registryUnderscoreMessage right after this line, led
# by a blank line. Anything else is not a message we know about.
start = index + 1
end = start + REGISTRY_UNDERSCORE_MESSAGE_LINES
if "".join(lines[start:end]) != REGISTRY_UNDERSCORE_MESSAGE:
break
index = end
else:
break
return "".join(lines[index:])
def parse_helm_plugin_list(output=None):
"""
Parse `helm plugin list`, return list of plugins
+1
View File
@@ -181,6 +181,7 @@ def _create_headers(module=None, **kwargs):
header_map = {
"impersonate_user": "Impersonate-User",
"impersonate_groups": "Impersonate-Group",
"accept": "Accept",
}
headers = {}
+19 -6
View File
@@ -47,6 +47,18 @@ def pod_ready(pod: ResourceInstance) -> bool:
)
def job_complete(job: ResourceInstance) -> bool:
return bool(
job.status
and job.status.conditions
and any(
condition["type"] in ("Complete", "Failed")
and boolean(condition["status"], strict=False)
for condition in job.status.conditions
)
)
def daemonset_ready(daemonset: ResourceInstance) -> bool:
return bool(
daemonset.status
@@ -89,18 +101,18 @@ def custom_condition(condition: Dict, resource: ResourceInstance) -> bool:
if not matches:
return False
# There should never be more than one condition of a specific type
match: ResourceField = matches[0]
if match.status == "Unknown":
if match.status == condition["status"]:
match_condition: ResourceField = matches[0]
if match_condition.status == "Unknown":
if match_condition.status == condition["status"]:
if "reason" not in condition:
return True
if condition["reason"]:
return match.reason == condition["reason"]
return match_condition.reason == condition["reason"]
return False
status = True if match.status == "True" else False
status = True if match_condition.status == "True" else False
if status == boolean(condition["status"], strict=False):
if condition.get("reason"):
return match.reason == condition["reason"]
return match_condition.reason == condition["reason"]
return True
return False
@@ -143,6 +155,7 @@ RESOURCE_PREDICATES = {
"DaemonSet": daemonset_ready,
"Deployment": deployment_ready,
"Pod": pod_ready,
"Job": job_complete,
"StatefulSet": statefulset_ready,
"ClusterOperator": cluster_operator_ready,
}
+217 -61
View File
@@ -132,15 +132,19 @@ options:
version_added: 2.4.0
reuse_values:
description:
- When upgrading package, specifies wether to reuse the last release's values and merge in any overrides from parameters I(release_values),
I(values_files) or I(set_values).
- If I(reset_values) is set to C(True), this is ignored.
- When upgrading package, specifies whether to reuse the last release's values and merge in any overrides from parameters O(release_values),
O(values_files) or O(set_values).
- Helm ignores this when C(--reset-values) is passed. Since O(reset_values) defaults to V(true), it has to be set to V(false)
for this option to take any effect.
- C(helm install) does not accept C(--reuse-values), so this option is ignored when O(replace) is set.
- The module warns whenever one of those combinations makes this option a no-op.
type: bool
required: false
version_added: 3.0.0
reset_values:
description:
- When upgrading package, reset the values to the ones built into the chart.
- Defaults to V(true), which makes helm ignore O(reuse_values) and O(reset_then_reuse_values). Set it to V(false) to use either of them.
type: bool
required: false
default: True
@@ -149,7 +153,10 @@ options:
description:
- When upgrading package, reset the values to the ones built into the chart, apply the last release's values and merge in any overrides from
parameters O(release_values), O(values_files) or O(set_values).
- If O(reset_values) or O(reuse_values) is set to V(True), this is ignored.
- Helm ignores this when C(--reset-values) or C(--reuse-values) is passed. Since O(reset_values) defaults to V(true), it has to be
set to V(false), and O(reuse_values) left unset or V(false), for this option to take any effect.
- C(helm install) does not accept C(--reset-then-reuse-values), so this option is ignored when O(replace) is set.
- The module warns whenever one of those combinations makes this option a no-op.
- This feature requires helm diff >= 3.9.12.
type: bool
required: false
@@ -187,10 +194,19 @@ options:
It will wait for as long as I(wait_timeout). This feature requires helm>=3.7.0. Added in version 2.3.0.
default: False
type: bool
wait_for_jobs:
description:
- When I(release_state) is set to C(present), and I(wait) is set to C(True), wait until all jobs are in a
completed state before marking the release as successful.
- Ignored when used without O(wait).
- Requires Helm >= 3.5.0
version_added: 6.6.0
default: False
type: bool
wait_timeout:
description:
- Timeout when wait option is enabled (helm2 is a number of seconds, helm3 is a duration).
- The use of I(wait_timeout) to wait for kubernetes commands to complete has been deprecated and will be removed after 2022-12-01.
- The use of I(wait_timeout) to wait for kubernetes commands to complete has been deprecated and will be removed in version 7.0.0. Use O(timeout) instead.
type: str
timeout:
description:
@@ -207,6 +223,18 @@ options:
O(atomic) option works the same way regardless of the installed Helm version.
type: bool
default: False
cleanup_on_fail:
description:
- Allow deletion of new resources created during a failed upgrade.
- Maps to the C(--cleanup-on-fail) flag, which is available since helm 3.0.0 and
was not renamed in Helm v4.
- Complements O(atomic) rather than replacing it, and the two can be combined so
that a rollback also drops resources orphaned by the failed upgrade.
- Helm only accepts this flag on C(helm upgrade), so it cannot be used together
with O(replace), which deploys through C(helm install).
type: bool
default: False
version_added: 6.6.0
server_side:
description:
- Control Helm v4 server-side apply when installing or upgrading a release.
@@ -385,6 +413,7 @@ EXAMPLES = r"""
release_namespace: default
force: True
wait: True
wait_for_jobs: True
replace: True
update_repo_cache: True
disable_hook: True
@@ -404,15 +433,30 @@ EXAMPLES = r"""
enabled: True
# Deploy latest version
# 'reset_values' defaults to true and helm ignores '--reuse-values' when it is set,
# so it has to be turned off explicitly.
- name: Deploy latest version of Grafana chart using reuse_values
kubernetes.core.helm:
name: test
chart_ref: stable/grafana
release_namespace: monitoring
reuse_values: true
reset_values: false
values:
replicas: 2
version: 3e8ec0b2dffa40fb97d5342e4af887de95faa8c61a62480dd7f8aa03dffcf533
# Same for 'reset_then_reuse_values', which helm ignores when either
# '--reset-values' or '--reuse-values' is set.
- name: Deploy latest version of Grafana chart using reset_then_reuse_values
kubernetes.core.helm:
name: test
chart_ref: stable/grafana
release_namespace: monitoring
reset_then_reuse_values: true
reset_values: false
values:
replicas: 2
"""
RETURN = r"""
@@ -454,8 +498,8 @@ status:
returned: always
description:
- Dict of Values used to deploy.
- This return value has been deprecated and will be removed in a release after
2027-01-08. Use RV(status.release_values) instead.
- This return value has been deprecated and will be removed in version 8.0.0.
Use RV(status.release_values) instead.
release_values:
type: dict
returned: always
@@ -501,6 +545,7 @@ from ansible.module_utils.basic import missing_required_lib
from ansible_collections.kubernetes.core.plugins.module_utils.helm import (
AnsibleHelmModule,
parse_helm_plugin_list,
strip_registry_progress,
)
from ansible_collections.kubernetes.core.plugins.module_utils.helm_args_common import (
HELM_AUTH_ARG_SPEC,
@@ -548,6 +593,13 @@ def get_release_status(module, release_name, all_status=False):
release["release_values"] = module.get_values(release_name)
release["values"] = release["release_values"]
# Add deprecation warning for status.values
module.deprecate(
"The 'status.values' return value is deprecated and will be removed in version 8.0.0. Use 'status.release_values' instead.",
version="8.0.0",
collection_name="kubernetes.core",
)
return release
@@ -591,7 +643,52 @@ def fetch_chart_info(
rc, out, err = module.run_helm_command(inspect_command)
return yaml.safe_load(out)
return yaml.safe_load(strip_registry_progress(out))
def validate_value_options(
module, replace, reuse_values, reset_values, reset_then_reuse_values
):
"""
Check the value-reuse options against each other and against 'replace'.
'helm install' accepts none of these flags, so 'replace' drops all three. The rest
are precedence rules helm applies silently: '--reset-values' wins over
'--reuse-values', and both win over '--reset-then-reuse-values'. Since 'reset_values'
defaults to true, requesting either of the other two on its own is a no-op.
Warn rather than fail throughout. Every one of these combinations is accepted by
playbooks today: 'reuse_values' has existed since 3.0.0, and a release that is
already converged never reaches deploy(), so the invalid command was never built.
Failing here would turn tasks that pass today into tasks that fail.
"""
if replace:
for name, requested in (
("reuse_values", reuse_values is not None),
("reset_then_reuse_values", reset_then_reuse_values),
):
if requested:
module.warn(
"{0} is ignored because replace=true deploys through 'helm install',"
" which does not accept the flag.".format(name)
)
return
if reset_values:
for name, requested in (
("reuse_values", reuse_values),
("reset_then_reuse_values", reset_then_reuse_values),
):
if requested:
module.warn(
"{0} is ignored because reset_values is true. "
"Set reset_values=false to make it take effect.".format(name)
)
elif reuse_values and reset_then_reuse_values:
module.warn(
"reset_then_reuse_values is ignored because reuse_values is true. "
"Set reuse_values=false to make it take effect."
)
def deploy(
@@ -601,12 +698,14 @@ def deploy(
release_values,
chart_name,
wait,
wait_for_jobs,
wait_timeout,
disable_hook,
force,
values_files,
history_max,
atomic=False,
cleanup_on_fail=False,
create_namespace=False,
replace=False,
post_renderer=None,
@@ -634,27 +733,41 @@ def deploy(
deploy_command += " --dependency-update"
else:
deploy_command = command + " upgrade -i" # install/upgrade
if reset_values:
deploy_command += " --reset-values"
if take_ownership:
deploy_command += " --take-ownership"
if reuse_values is not None:
deploy_command += " --reuse-values=" + str(reuse_values)
# '--reset-values', '--reuse-values' and '--reset-then-reuse-values' are all
# upgrade-only flags, so they must not leak into the 'replace' branch above.
# validate_value_options() has already warned about that combination.
if reset_values:
deploy_command += " --reset-values"
if reset_then_reuse_values:
helm_version = module.get_helm_version()
if LooseVersion(helm_version) < LooseVersion("3.14.0"):
module.fail_json(
msg="reset_then_reuse_values requires helm >= 3.14.0, current version is {0}".format(
helm_version
if reuse_values is not None:
deploy_command += " --reuse-values=" + str(reuse_values)
if reset_then_reuse_values:
helm_version = module.get_helm_version()
if LooseVersion(helm_version) < LooseVersion("3.14.0"):
module.fail_json(
msg="reset_then_reuse_values requires helm >= 3.14.0, current version is {0}".format(
helm_version
)
)
)
else:
deploy_command += " --reset-then-reuse-values"
else:
deploy_command += " --reset-then-reuse-values"
if wait:
deploy_command += " --wait"
if wait_for_jobs:
helm_version = module.get_helm_version()
if LooseVersion(helm_version) < LooseVersion("3.5.0"):
module.fail_json(
msg="wait_for_jobs requires helm >= 3.5.0, current version is {0}".format(
helm_version
)
)
else:
deploy_command += " --wait-for-jobs"
if wait_timeout is not None:
deploy_command += " --timeout " + wait_timeout
@@ -667,6 +780,17 @@ def deploy(
else:
deploy_command += " --atomic"
if cleanup_on_fail:
# '--cleanup-on-fail' is only accepted by 'helm upgrade', so it cannot be
# combined with 'replace', which deploys through 'helm install'. The flag
# exists since helm 3.0.0 and kept its name in v4, so no version gate is
# needed on top of the module-wide helm >= 3.0.0 requirement.
if replace:
module.fail_json(
msg="cleanup_on_fail is only supported by 'helm upgrade' and cannot be used with replace=true"
)
deploy_command += " --cleanup-on-fail"
if timeout:
deploy_command += " --timeout " + timeout
@@ -865,10 +989,10 @@ def helmdiff_check(
if set_value_args:
cmd += " " + set_value_args
if reuse_values:
if not replace and reuse_values:
cmd += " --reuse-values"
if reset_then_reuse_values:
if not replace and reset_then_reuse_values:
helm_diff_version = get_plugin_version("diff")
helm_version = module.get_helm_version()
fail_msg = ""
@@ -942,44 +1066,58 @@ def default_check(release_status, chart_info, values=None, values_files=None):
def argument_spec():
arg_spec = copy.deepcopy(HELM_AUTH_ARG_SPEC)
arg_spec.update(
dict(
chart_ref=dict(type="path"),
chart_repo_url=dict(type="str"),
chart_version=dict(type="str"),
dependency_update=dict(type="bool", default=False, aliases=["dep_up"]),
release_name=dict(type="str", required=True, aliases=["name"]),
release_namespace=dict(type="str", required=True, aliases=["namespace"]),
release_state=dict(
default="present", choices=["present", "absent"], aliases=["state"]
),
release_values=dict(type="dict", default={}, aliases=["values"]),
values_files=dict(type="list", default=[], elements="str"),
update_repo_cache=dict(type="bool", default=False),
disable_hook=dict(type="bool", default=False),
force=dict(type="bool", default=False),
purge=dict(type="bool", default=True),
wait=dict(type="bool", default=False),
wait_timeout=dict(type="str"),
timeout=dict(type="str"),
atomic=dict(type="bool", default=False),
server_side=dict(type="str", choices=["auto", "true", "false"]),
force_conflicts=dict(type="bool", default=False),
create_namespace=dict(type="bool", default=False),
post_renderer=dict(type="str"),
replace=dict(type="bool", default=False),
skip_crds=dict(type="bool", default=False),
history_max=dict(type="int"),
set_values=dict(type="list", elements="dict"),
reuse_values=dict(type="bool"),
reset_values=dict(type="bool", default=True),
reset_then_reuse_values=dict(type="bool", default=False),
insecure_skip_tls_verify=dict(
type="bool", default=False, aliases=["skip_tls_certs_check"]
),
plain_http=dict(type="bool", default=False),
take_ownership=dict(type="bool", default=False),
skip_schema_validation=dict(type="bool", default=False),
)
{
"chart_ref": {"type": "path"},
"chart_repo_url": {"type": "str"},
"chart_version": {"type": "str"},
"dependency_update": {
"type": "bool",
"default": False,
"aliases": ["dep_up"],
},
"release_name": {"type": "str", "required": True, "aliases": ["name"]},
"release_namespace": {
"type": "str",
"required": True,
"aliases": ["namespace"],
},
"release_state": {
"default": "present",
"choices": ["present", "absent"],
"aliases": ["state"],
},
"release_values": {"type": "dict", "default": {}, "aliases": ["values"]},
"values_files": {"type": "list", "default": [], "elements": "str"},
"update_repo_cache": {"type": "bool", "default": False},
"disable_hook": {"type": "bool", "default": False},
"force": {"type": "bool", "default": False},
"purge": {"type": "bool", "default": True},
"wait": {"type": "bool", "default": False},
"wait_for_jobs": {"type": "bool", "default": False},
"wait_timeout": {"type": "str"},
"timeout": {"type": "str"},
"atomic": {"type": "bool", "default": False},
"server_side": {"type": "str", "choices": ["auto", "true", "false"]},
"force_conflicts": {"type": "bool", "default": False},
"create_namespace": {"type": "bool", "default": False},
"post_renderer": {"type": "str"},
"replace": {"type": "bool", "default": False},
"skip_crds": {"type": "bool", "default": False},
"history_max": {"type": "int"},
"set_values": {"type": "list", "elements": "dict"},
"reuse_values": {"type": "bool"},
"reset_values": {"type": "bool", "default": True},
"reset_then_reuse_values": {"type": "bool", "default": False},
"insecure_skip_tls_verify": {
"type": "bool",
"default": False,
"aliases": ["skip_tls_certs_check"],
},
"plain_http": {"type": "bool", "default": False},
"take_ownership": {"type": "bool", "default": False},
"skip_schema_validation": {"type": "bool", "default": False},
"cleanup_on_fail": {"type": "bool", "default": False},
}
)
return arg_spec
@@ -1002,6 +1140,14 @@ def main():
supports_check_mode=True,
)
# Add deprecation warning for wait_timeout
if module.params.get("wait_timeout") is not None:
module.deprecate(
"The 'wait_timeout' parameter is deprecated and will be removed in version 7.0.0. Use 'timeout' instead.",
version="7.0.0",
collection_name="kubernetes.core",
)
if not IMP_YAML:
module.fail_json(msg=missing_required_lib("yaml"), exception=IMP_YAML_ERR)
@@ -1025,8 +1171,10 @@ def main():
force = module.params.get("force")
purge = module.params.get("purge")
wait = module.params.get("wait")
wait_for_jobs = module.params.get("wait_for_jobs")
wait_timeout = module.params.get("wait_timeout")
atomic = module.params.get("atomic")
cleanup_on_fail = module.params.get("cleanup_on_fail")
server_side = module.params.get("server_side")
force_conflicts = module.params.get("force_conflicts")
create_namespace = module.params.get("create_namespace")
@@ -1099,6 +1247,10 @@ def main():
)
changed = True
elif release_state == "present":
validate_value_options(
module, replace, reuse_values, reset_values, reset_then_reuse_values
)
if chart_version is not None:
helm_cmd += " --version=" + chart_version
@@ -1152,11 +1304,13 @@ def main():
release_values,
chart_ref,
wait,
wait_for_jobs,
wait_timeout,
disable_hook,
False,
values_files=values_files,
atomic=atomic,
cleanup_on_fail=cleanup_on_fail,
server_side=server_side,
force_conflicts=force_conflicts,
create_namespace=create_namespace,
@@ -1238,11 +1392,13 @@ def main():
release_values,
chart_ref,
wait,
wait_for_jobs,
wait_timeout,
disable_hook,
force,
values_files=values_files,
atomic=atomic,
cleanup_on_fail=cleanup_on_fail,
server_side=server_side,
force_conflicts=force_conflicts,
create_namespace=create_namespace,
+10 -2
View File
@@ -119,8 +119,8 @@ status:
returned: always
description:
- Dict of Values used to deploy
- This return value has been deprecated and will be removed in a release after
2027-01-08. Use RV(status.release_values) instead.
- This return value has been deprecated and will be removed in version 8.0.0.
Use RV(status.release_values) instead.
release_values:
type: dict
returned: always
@@ -213,6 +213,14 @@ def get_release_status(module, release_name, release_state, get_all_values=False
release["release_values"] = module.get_values(release_name, get_all_values)
release["values"] = release["release_values"]
# Add deprecation warning for status.values
module.deprecate(
"The 'status.values' return value is deprecated and will be removed in version 8.0.0. Use 'status.release_values' instead.",
version="8.0.0",
collection_name="kubernetes.core",
)
release["manifest"] = module.get_manifest(release_name)
release["notes"] = module.get_notes(release_name)
release["hooks"] = module.get_hooks(release_name)
+1 -1
View File
@@ -331,7 +331,7 @@ def main():
insecure_skip_tls_verify,
)
changed = True
elif repository_status["url"] != repo_url:
elif repository_status["url"].rstrip("/") != repo_url:
module.fail_json(
msg="Repository already have a repository named {0}".format(repo_name)
)
+4
View File
@@ -228,6 +228,7 @@ except ImportError:
from ansible.module_utils.basic import missing_required_lib
from ansible_collections.kubernetes.core.plugins.module_utils.helm import (
AnsibleHelmModule,
strip_registry_progress,
)
from ansible_collections.kubernetes.core.plugins.module_utils.version import (
LooseVersion,
@@ -394,6 +395,9 @@ def main():
if not check_mode:
rc, out, err = module.run_helm_command(tmpl_cmd)
# 'stdout' is documented to hold the rendered templates, so keep helm's
# OCI registry progress messages out of it.
out = strip_registry_progress(out)
else:
out = err = ""
rc = 0
+13
View File
@@ -76,9 +76,21 @@ options:
- This option is ignored when I(content) is set or when I(state) is set to C(from_pod).
type: bool
default: False
copy_timeout:
description:
- Maximum time in seconds to spend streaming the archive to the pod and waiting for the remote
C(tar) process to finish extracting it.
- Raise this when copying large files over a slow connection.
- This option is ignored when I(state) is set to C(from_pod).
type: int
default: 300
version_added: 6.6.0
notes:
- the tar binary is required on the container when copying from local filesystem to pod.
- a POSIX shell at C(/bin/sh) and C(head) are also used, when present, to bound the archive
sent to the container so that the module can confirm the copy finished. Without them the
copy still runs but completion cannot be verified, and a warning is emitted.
- the (init) container has to be started before you copy files or directories to it.
"""
@@ -180,6 +192,7 @@ def argspec():
"choices": ["to_pod", "from_pod"],
}
argument_spec["no_preserve"] = {"type": "bool", "default": False}
argument_spec["copy_timeout"] = {"type": "int", "default": 300}
return argument_spec
+8 -1
View File
@@ -118,7 +118,7 @@ result:
type: int
version_added: 2.2.0
return_code:
description: The command status code. This attribute is deprecated and will be removed in a future release. Please use rc instead.
description: The command status code. This attribute is deprecated and will be removed in version 7.0.0. Use RV(result.rc) instead.
type: int
"""
@@ -219,6 +219,13 @@ def execute_module(module, client):
else:
rc = int(err["details"]["causes"][0]["message"])
# Add deprecation warning for return_code return value
module.deprecate(
"The 'return_code' return value is deprecated and will be removed in version 7.0.0. Use 'rc' instead.",
version="7.0.0",
collection_name="kubernetes.core",
)
module.exit_json(
# Some command might change environment, but ultimately failing at end
changed=True,
+40 -10
View File
@@ -53,6 +53,15 @@ options:
type: list
elements: str
version_added: 3.0.0
metadata_only:
description:
- Request metadata only response from the Kubernetes API server.
- This feature is useful for clients that only need to check for the existence of an object,
- or that only need to read its metadata.
- It can significantly reduce the size of the response from the API server.
type: bool
default: false
version_added: 6.6.0
extends_documentation_fragment:
- kubernetes.core.k8s_auth_options
@@ -135,6 +144,12 @@ EXAMPLES = r"""
(ocp_bootstrap_status.resources[0].data.status == 'complete')
retries: 60
delay: 15
- name: Get a list of all pods metadata only
kubernetes.core.k8s_info:
kind: Pod
metadata_only: true
register: pod_metadata_list
"""
RETURN = r"""
@@ -179,6 +194,8 @@ from ansible_collections.kubernetes.core.plugins.module_utils.ansiblemodule impo
)
from ansible_collections.kubernetes.core.plugins.module_utils.args_common import (
AUTH_ARG_SPEC,
METADATA_ONLY_HEADER_VALUES,
METADATA_ONLY_HEADERS_SPEC,
WAIT_ARG_SPEC,
)
from ansible_collections.kubernetes.core.plugins.module_utils.k8s.client import (
@@ -216,15 +233,16 @@ def argspec():
args = copy.deepcopy(AUTH_ARG_SPEC)
args.update(WAIT_ARG_SPEC)
args.update(
dict(
kind=dict(required=True),
api_version=dict(default="v1", aliases=["api", "version"]),
name=dict(),
namespace=dict(),
label_selectors=dict(type="list", elements="str", default=[]),
field_selectors=dict(type="list", elements="str", default=[]),
hidden_fields=dict(type="list", elements="str"),
)
{
"kind": {"required": True},
"api_version": {"default": "v1", "aliases": ["api", "version"]},
"name": {},
"namespace": {},
"label_selectors": {"type": "list", "elements": "str", "default": []},
"field_selectors": {"type": "list", "elements": "str", "default": []},
"hidden_fields": {"type": "list", "elements": "str"},
"metadata_only": {"type": "bool", "default": False},
}
)
return args
@@ -233,8 +251,20 @@ def main():
module = AnsibleK8SModule(
module_class=AnsibleModule, argument_spec=argspec(), supports_check_mode=True
)
metadata_only_headers = {}
if module.params["metadata_only"]:
metadata_only_headers = copy.deepcopy(METADATA_ONLY_HEADERS_SPEC)
metadata_only_headers.update(
{"accept": METADATA_ONLY_HEADER_VALUES["partial_object_list"]}
)
if module.params["name"]:
metadata_only_headers.update(
{"accept": METADATA_ONLY_HEADER_VALUES["partial_object"]}
)
try:
client = get_api_client(module)
client = get_api_client(module, **metadata_only_headers)
svc = K8sService(client, module)
execute_module(module, svc)
except CoreException as e:
+10
View File
@@ -36,6 +36,7 @@ options:
- If more than one C(merge_type) is given, the merge_types will be tried in order
- This defaults to C(['strategic-merge', 'merge']), which is ideal for using the same parameters
on resource kinds that combine Custom Resources and built-in resources.
- I(merge_type=json) has been deprecated and will be removed in version 7.0.0. Use M(kubernetes.core.k8s_json_patch) instead.
choices:
- json
- merge
@@ -251,6 +252,15 @@ def main():
supports_check_mode=True,
)
# Add deprecation warning for merge_type=json
merge_type = module.params.get("merge_type")
if merge_type and "json" in merge_type:
module.deprecate(
"The 'merge_type=json' option is deprecated and will be removed in version 7.0.0. Use the 'kubernetes.core.k8s_json_patch' module instead.",
version="7.0.0",
collection_name="kubernetes.core",
)
try:
client = get_api_client(module=module)
svc = K8sService(client, module)
+11
View File
@@ -13,6 +13,17 @@ sonar.newCode.referenceBranch=stable-6
sonar.exclusions=tests/**,.tox/**
# Exclude test code from copy/paste detection. 'sonar.exclusions' above only
# filters the *source* set (it disambiguates 'sonar.sources=.' from
# 'sonar.tests'), so test files are still indexed as tests and still scored for
# duplication. Unit tests here are unittest.TestCase classes: every class needs
# its own setUp, and per-case arrange/assert blocks are deliberately repeated so
# a failing test reads on its own. Because SonarCloud scores only *new code* on
# a pull request, any PR adding cases to an already-repetitive test module
# inherits duplication accrued across earlier PRs. Production code under
# plugins/ stays fully covered by duplication analysis.
sonar.cpd.exclusions=tests/**
# Ansible module argument_spec is conventionally defined with the dict(...)
# constructor (not {...} literals) for readability and consistency across the
# collection. Suppress python:S7498 ("prefer literal syntax"), which conflicts
+1 -1
View File
@@ -1,2 +1,2 @@
modules:
python_requires: ">=3.6"
python_requires: ">=3.9"
@@ -31,3 +31,4 @@ test_namespace:
- "helm-insecure-{{ helm_version | replace('.', '-') }}"
- "helm-test-take-ownership-{{ helm_version | replace('.', '-') }}"
- "helm-skip-schema-validation-{{ helm_version | replace('.', '-') }}"
- "helm-wait-for-jobs-{{ helm_version | replace('.', '-') }}"
@@ -0,0 +1,6 @@
apiVersion: v2
name: test-chart-job
description: A chart with a Job for testing wait_for_jobs
type: application
version: 0.1.0
appVersion: "default"
@@ -0,0 +1,21 @@
apiVersion: batch/v1
kind: Job
metadata:
name: "{{ .Release.Name }}-job"
spec:
backoffLimit: 0
template:
spec:
automountServiceAccountToken: false
restartPolicy: Never
containers:
- name: test
image: "busybox:1.37"
command: ["sh", "-c", "echo done"]
resources:
requests:
cpu: "50m"
memory: "32Mi"
ephemeral-storage: "16Mi"
limits:
memory: "64Mi"
@@ -49,6 +49,9 @@
- name: Test helm skip_schema_validation
include_tasks: test_skip_schema_validation.yml
- name: Test helm wait_for_jobs
include_tasks: test_helm_wait_for_jobs.yml
always:
- name: Remove helm-diff plugin
helm_plugin:
@@ -0,0 +1,59 @@
---
- name: Test helm wait_for_jobs
vars:
helm_namespace: "{{ test_namespace[15] }}"
chart_local: "{{ role_path }}/files/test-chart-job"
block:
- name: Install chart with wait and wait_for_jobs
helm:
binary_path: "{{ helm_binary }}"
chart_ref: "{{ chart_local }}"
release_name: test-wait-for-jobs
release_namespace: "{{ helm_namespace }}"
create_namespace: true
wait: true
wait_for_jobs: true
wait_timeout: "120s"
register: install
ignore_errors: true
- name: Validate wait_for_jobs with helm >= 3.5.0 works
assert:
that:
- install is changed
- "'--wait' in install.command"
- "'--wait-for-jobs' in install.command"
when: "helm_version is ansible.builtin.version('v3.5.0', '>=')"
- name: Validate wait_for_jobs with helm < 3.5.0 fails
assert:
that:
- install is failed
- "'wait_for_jobs requires helm >= 3.5.0' in install.msg"
when: "helm_version is ansible.builtin.version('v3.5.0', '<')"
- name: Install chart with wait_for_jobs but without wait
helm:
binary_path: "{{ helm_binary }}"
chart_ref: "{{ chart_local }}"
release_name: test-wait-for-jobs-no-wait
release_namespace: "{{ helm_namespace }}"
create_namespace: true
wait_for_jobs: true
register: install_no_wait
- name: Validate wait_for_jobs is ignored without wait
assert:
that:
- install_no_wait is changed
- "'--wait-for-jobs' not in install_no_wait.command"
- "'--wait' not in install_no_wait.command"
always:
- name: Remove helm namespace
k8s:
api_version: v1
kind: Namespace
name: "{{ helm_namespace }}"
state: absent
@@ -0,0 +1,3 @@
time=10
helm
helm_template
@@ -0,0 +1,20 @@
---
# Helm >= 4.2.1 is required for this test to be meaningful: that is the release which
# moved the OCI registry progress messages from stderr to stdout (helm/helm#32056).
helm_oci_version: v4.2.4
helm_oci_namespace: helm-oci-registry-progress
test_namespace: "{{ helm_oci_namespace }}"
# Matches the registry stood up by the setup_helm_registry role.
helm_oci_registry: "localhost:6035"
helm_oci_registry_username: testuser
helm_oci_registry_password: 'pass123!'
helm_oci_repo_path: testing
helm_oci_chart_name: test-chart-plus-version
# '+' is not allowed in an OCI tag, so Helm stores this as '1.0.0_build.1'.
helm_oci_chart_version: 1.0.0+build.1
# A version without build metadata, to cover the case where the progress messages
# happen to be valid YAML and so corrupt the parsed output silently.
helm_oci_chart_version_plain: 1.0.1
@@ -0,0 +1,6 @@
apiVersion: v2
name: test-chart-plus-version
description: A chart whose version carries semver build metadata
type: application
version: 1.0.0+build.1
appVersion: "default"
@@ -0,0 +1,6 @@
apiVersion: v1
kind: ConfigMap
metadata:
name: test-chart-plus-version-configmap
data:
myValue: test
@@ -0,0 +1,4 @@
---
dependencies:
- remove_namespace
- setup_helm_registry
@@ -0,0 +1,158 @@
---
# Regression test for https://github.com/ansible-collections/kubernetes.core/issues/1222
#
# Helm >= 4.2.1 writes OCI registry progress messages to stdout rather than stderr
# (helm/helm#32056): 'Pulled:', 'Digest:' and, when the reference contains an
# underscore, a free-text explanation. Helm rewrites '+' to '_' in OCI tags, so a
# chart version carrying semver build metadata is what triggers that explanation,
# which is not valid YAML.
#
# The helm module parses the stdout of 'helm show chart', and helm_template returns
# the stdout of 'helm template' as its documented result, so both have to skip those
# messages. Without the fix, the helm task below fails with
# "could not find expected ':'".
- name: Test OCI charts whose version contains build metadata
vars:
chart_ref: "oci://{{ helm_oci_registry }}/{{ helm_oci_repo_path }}/{{ helm_oci_chart_name }}"
block:
- name: "Install Helm {{ helm_oci_version }}"
ansible.builtin.include_role:
name: install_helm
vars:
helm_version: "{{ helm_oci_version }}"
- name: Create temporary directory to hold the chart
ansible.builtin.tempfile:
state: directory
suffix: .chart
register: _tmpd
- name: Copy the test chart
ansible.builtin.copy:
src: "{{ helm_oci_chart_name }}"
dest: "{{ _tmpd.path }}/"
- name: Log into the helm registry
ansible.builtin.command: >-
{{ helm_binary }} registry login --plain-http
-u {{ helm_oci_registry_username }} -p {{ helm_oci_registry_password }}
{{ helm_oci_registry }}
- name: Package and push both chart versions
ansible.builtin.command: "{{ item }}"
loop:
- >-
{{ helm_binary }} package {{ _tmpd.path }}/{{ helm_oci_chart_name }}
--version {{ helm_oci_chart_version }} --destination {{ _tmpd.path }}
- >-
{{ helm_binary }} package {{ _tmpd.path }}/{{ helm_oci_chart_name }}
--version {{ helm_oci_chart_version_plain }} --destination {{ _tmpd.path }}
- >-
{{ helm_binary }} push --plain-http
{{ _tmpd.path }}/{{ helm_oci_chart_name }}-{{ helm_oci_chart_version }}.tgz
oci://{{ helm_oci_registry }}/{{ helm_oci_repo_path }}
- >-
{{ helm_binary }} push --plain-http
{{ _tmpd.path }}/{{ helm_oci_chart_name }}-{{ helm_oci_chart_version_plain }}.tgz
oci://{{ helm_oci_registry }}/{{ helm_oci_repo_path }}
register: _push
- name: Ensure Helm stored the build metadata version with an underscore
ansible.builtin.assert:
that:
- "'1.0.0_build.1' in (_push.results[2].stdout + _push.results[2].stderr)"
fail_msg: >-
Expected Helm to rewrite '+' to '_' when pushing the chart tag. Without
that rewrite this target no longer covers the case it was written for.
# helm - the version with build metadata is the case that failed outright.
- name: Install the chart from the OCI registry
kubernetes.core.helm:
binary_path: "{{ helm_binary }}"
name: "{{ helm_oci_chart_name }}"
chart_ref: "{{ chart_ref }}"
chart_version: "{{ helm_oci_chart_version }}"
release_namespace: "{{ helm_oci_namespace }}"
create_namespace: true
plain_http: true
register: _install
- name: Ensure the release reports the chart version verbatim
ansible.builtin.assert:
that:
- _install is changed
- _install.status.chart == helm_oci_chart_name ~ '-' ~ helm_oci_chart_version
- name: Install the same chart again
kubernetes.core.helm:
binary_path: "{{ helm_binary }}"
name: "{{ helm_oci_chart_name }}"
chart_ref: "{{ chart_ref }}"
chart_version: "{{ helm_oci_chart_version }}"
release_namespace: "{{ helm_oci_namespace }}"
plain_http: true
register: _reinstall
- name: Ensure the module is idempotent
ansible.builtin.assert:
that:
- _reinstall is not changed
# helm_template - stdout is documented to hold only the rendered templates. Both
# chart versions are checked: the plain one covers the case where the progress
# messages are valid YAML and would otherwise be parsed as a manifest.
- name: Render the chart from the OCI registry
kubernetes.core.helm_template:
binary_path: "{{ helm_binary }}"
chart_ref: "{{ chart_ref }}"
chart_version: "{{ item }}"
release_name: "{{ helm_oci_chart_name }}"
plain_http: true
loop:
- "{{ helm_oci_chart_version }}"
- "{{ helm_oci_chart_version_plain }}"
register: _template
- name: Ensure stdout holds the rendered manifests and nothing else
ansible.builtin.assert:
that:
- "'Pulled:' not in item.stdout"
- "'Digest:' not in item.stdout"
- "'contains an underscore' not in item.stdout"
- item.stdout.startswith('---')
- (item.stdout | from_yaml_all | list | map(attribute='kind') | list) == ['ConfigMap']
fail_msg: "helm_template stdout was not only rendered manifests: {{ item.stdout[:200] }}"
loop: "{{ _template.results }}"
loop_control:
label: "{{ item.item }}"
always:
- name: Uninstall the release
kubernetes.core.helm:
binary_path: "{{ helm_binary }}"
name: "{{ helm_oci_chart_name }}"
release_namespace: "{{ helm_oci_namespace }}"
state: absent
when: helm_binary is defined
ignore_errors: true
- name: Remove the test namespace
kubernetes.core.k8s:
api_version: v1
kind: Namespace
name: "{{ helm_oci_namespace }}"
state: absent
wait: true
ignore_errors: true
- name: Log out of the helm registry
ansible.builtin.command: "{{ helm_binary }} registry logout {{ helm_oci_registry }}"
when: helm_binary is defined
ignore_errors: true
- name: Remove the temporary directory
ansible.builtin.file:
path: "{{ _tmpd.path }}"
state: absent
when: _tmpd is defined and _tmpd.path is defined
ignore_errors: true
@@ -103,3 +103,81 @@
binary_path: "{{ helm_binary }}"
name: test_helm_repo
state: absent
# https://github.com/ansible-collections/kubernetes.core/issues/1235
- name: Ensure the trailing slash test repositories don't exist
helm_repository:
binary_path: "{{ helm_binary }}"
name: "{{ item }}"
state: absent
loop:
- test_helm_repo_slash
- test_helm_repo_noslash
- name: Register test_helm_repo_slash directly through helm, keeping the trailing slash
ansible.builtin.command: "{{ helm_binary }} repo add test_helm_repo_slash {{ chart_test_repo }}/"
changed_when: true
- name: Add test_helm_repo_noslash chart repository without a trailing slash
helm_repository:
binary_path: "{{ helm_binary }}"
name: test_helm_repo_noslash
repo_url: "{{ chart_test_repo }}"
- name: Check idempotency when the stored and the requested URL both have a trailing slash
helm_repository:
binary_path: "{{ helm_binary }}"
name: test_helm_repo_slash
repo_url: "{{ chart_test_repo }}/"
register: repository
- name: Assert idempotency when the stored and the requested URL both have a trailing slash
assert:
that:
- repository is not changed
- name: Check idempotency when only the stored URL has a trailing slash
helm_repository:
binary_path: "{{ helm_binary }}"
name: test_helm_repo_slash
repo_url: "{{ chart_test_repo }}"
register: repository
- name: Assert idempotency when only the stored URL has a trailing slash
assert:
that:
- repository is not changed
- name: Check idempotency when only the requested URL has a trailing slash
helm_repository:
binary_path: "{{ helm_binary }}"
name: test_helm_repo_noslash
repo_url: "{{ chart_test_repo }}/"
register: repository
- name: Assert idempotency when only the requested URL has a trailing slash
assert:
that:
- repository is not changed
- name: Failed to add repository with the same name and a different URL, trailing slash aside
helm_repository:
binary_path: "{{ helm_binary }}"
name: test_helm_repo_slash
repo_url: "https://other-charts.url/"
register: repository_errors
ignore_errors: true
- name: Assert that adding repository with the same name and a different URL failed
assert:
that:
- repository_errors is failed
- name: Clean the trailing slash test repositories
helm_repository:
binary_path: "{{ helm_binary }}"
name: "{{ item }}"
state: absent
loop:
- test_helm_repo_slash
- test_helm_repo_noslash
@@ -143,9 +143,7 @@ def compare_directories(dir1, dir2):
[len(test.left_only) > 0, len(test.right_only) > 0, len(test.funny_files) > 0]
):
return False
(t, mismatch, errors) = filecmp.cmpfiles(
dir1, dir2, test.common_files, shallow=False
)
t, mismatch, errors = filecmp.cmpfiles(dir1, dir2, test.common_files, shallow=False)
if len(mismatch) > 0 or len(errors) > 0:
return False
for common_dir in test.common_dirs:
@@ -41,6 +41,7 @@
- include_tasks: test_multi_container_pod.yml
- include_tasks: test_copy_directory.yml
- include_tasks: test_copy_large_file.yml
- include_tasks: test_copy_repeated.yml
- include_tasks: test_copy_item_with_space_in_its_name.yml
- include_tasks: test_init_container_pod.yml
@@ -0,0 +1,46 @@
---
# Regression coverage for
# https://github.com/ansible-collections/kubernetes.core/issues/776
#
# The module used to close the exec connection as soon as the last chunk had
# been written, which killed the remote tar mid-extract and truncated the file.
# The failure was timing dependent, so a single copy rarely caught it: repeat
# the copy and compare every time.
- name: create temporary directory for local files
ansible.builtin.tempfile:
state: directory
suffix: k8scp_repeated
register: test_dir
- name: test repeated copies are never truncated
block:
- set_fact:
test_directory: "{{ test_dir.path }}"
copy_iterations: 10
no_log: true
# Large enough to span several 1MiB write chunks, so the last chunk is not
# also the first and the drain actually has something to wait for.
- name: create a multi-chunk file
ansible.builtin.command:
cmd: dd if=/dev/urandom of={{ test_directory }}/repeated.bin bs=1M count=12
- name: copy the same file repeatedly and verify each time
ansible.builtin.include_tasks: verify_one_copy.yml
loop: "{{ range(0, copy_iterations) | list }}"
loop_control:
loop_var: iteration
always:
- name: delete temporary directory created for the test
ansible.builtin.file:
path: "{{ test_dir.path }}"
state: absent
ignore_errors: true
- name: delete file created on Pod
k8s_exec:
namespace: '{{ copy_namespace }}'
pod: '{{ pod_with_one_container.name }}'
command: 'rm -f /repeated.bin'
ignore_errors: true
@@ -0,0 +1,23 @@
---
# One iteration of the repeated-copy regression check, see test_copy_repeated.yml
- name: remove any file left by the previous iteration
k8s_exec:
namespace: '{{ copy_namespace }}'
pod: '{{ pod_with_one_container.name }}'
command: 'rm -f /repeated.bin'
ignore_errors: true
- name: "copy file into remote Pod (iteration {{ iteration }})"
k8s_cp:
namespace: '{{ copy_namespace }}'
pod: '{{ pod_with_one_container.name }}'
remote_path: /repeated.bin
local_path: "{{ test_directory }}/repeated.bin"
state: to_pod
- name: "compare local and remote file (iteration {{ iteration }})"
kubectl_file_compare:
namespace: '{{ copy_namespace }}'
pod: '{{ pod_with_one_container.name }}'
remote_path: /repeated.bin
local_path: "{{ test_directory }}/repeated.bin"
@@ -4,3 +4,4 @@
- wait
- api-server-caching
- discovery
- metadata-only-fetches
@@ -0,0 +1,22 @@
---
- name: Retrieve Node list with metadata_only
kubernetes.core.k8s_info:
kind: Node
metadata_only: true
register: node_metadata_only_list
- name: Assert response has kind PartialObjectMetadata
assert:
that:
- node_metadata_only_list.resources[0].kind == "PartialObjectMetadata"
- name: Retrieve full Node list
kubernetes.core.k8s_info:
kind: Node
metadata_only: false
register: node_full_list
- name: Assert response has kind Node
assert:
that:
- node_full_list.resources[0].kind == "Node"
@@ -238,7 +238,55 @@
- resource.result.results | selectattr('changed') | list | length == 1
- resource.result.results | selectattr('error', 'defined') | list | length == 1
- name: Apply template where all items are disabled (empty output)
kubernetes.core.k8s:
state: present
template:
path: "conditional_empty.j2"
vars:
items:
- name: test1
enabled: false
- name: test2
enabled: false
k8s_pod_namespace: "{{ template_namespace }}"
register: r
- name: Check that empty template succeeds with no changes
assert:
that:
- r is successful
- r is not changed
- name: Apply template where some items are enabled
kubernetes.core.k8s:
state: present
template:
path: "conditional_empty.j2"
vars:
items:
- name: conditional-configmap-1
enabled: true
- name: conditional-configmap-2
enabled: false
k8s_pod_namespace: "{{ template_namespace }}"
register: r
- name: Check that partial template creates resources
assert:
that:
- r is successful
- r is changed
always:
- name: Cleanup conditional configmap
kubernetes.core.k8s:
state: absent
kind: ConfigMap
name: conditional-configmap-1
namespace: "{{ template_namespace }}"
ignore_errors: true
- name: Remove namespace (Cleanup)
kubernetes.core.k8s:
kind: Namespace
@@ -0,0 +1,12 @@
{% for item in items %}
{% if item.enabled %}
---
apiVersion: v1
kind: ConfigMap
metadata:
name: "{{ item.name }}"
namespace: "{{ k8s_pod_namespace }}"
data:
key: value
{% endif %}
{% endfor %}
+1
View File
@@ -21,6 +21,7 @@ tests/integration/targets/k8s_delete/files/deployments.yaml yamllint!skip
tests/unit/module_utils/fixtures/pods.yml yamllint!skip
tests/integration/targets/helm/files/appversionless-chart-v2/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/appversionless-chart/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart-job/Chart.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart-v2/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm_diff/files/test-chart/templates/configmap.yaml yamllint!skip
+1
View File
@@ -21,6 +21,7 @@ tests/integration/targets/k8s_delete/files/deployments.yaml yamllint!skip
tests/unit/module_utils/fixtures/pods.yml yamllint!skip
tests/integration/targets/helm/files/appversionless-chart-v2/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/appversionless-chart/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart-job/Chart.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart-v2/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm_diff/files/test-chart/templates/configmap.yaml yamllint!skip
+7
View File
@@ -1,9 +1,15 @@
plugins/module_utils/client/discovery.py import-3.9!skip
plugins/module_utils/client/discovery.py import-3.10!skip
plugins/module_utils/client/discovery.py import-3.11!skip
plugins/module_utils/client/discovery.py import-3.12!skip
plugins/module_utils/client/discovery.py import-3.13!skip
plugins/module_utils/client/resource.py import-3.9!skip
plugins/module_utils/client/resource.py import-3.10!skip
plugins/module_utils/client/resource.py import-3.11!skip
plugins/module_utils/client/resource.py import-3.12!skip
plugins/module_utils/client/resource.py import-3.13!skip
plugins/module_utils/k8sdynamicclient.py import-3.9!skip
plugins/module_utils/k8sdynamicclient.py import-3.10!skip
plugins/module_utils/k8sdynamicclient.py import-3.11!skip
plugins/module_utils/k8sdynamicclient.py import-3.12!skip
plugins/module_utils/k8sdynamicclient.py import-3.13!skip
@@ -18,6 +24,7 @@ tests/integration/targets/k8s_delete/files/deployments.yaml yamllint!skip
tests/unit/module_utils/fixtures/pods.yml yamllint!skip
tests/integration/targets/helm/files/appversionless-chart-v2/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/appversionless-chart/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart-job/Chart.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart-v2/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm_diff/files/test-chart/templates/configmap.yaml yamllint!skip
+7
View File
@@ -1,9 +1,15 @@
plugins/module_utils/client/discovery.py import-3.9!skip
plugins/module_utils/client/discovery.py import-3.10!skip
plugins/module_utils/client/discovery.py import-3.11!skip
plugins/module_utils/client/discovery.py import-3.12!skip
plugins/module_utils/client/discovery.py import-3.13!skip
plugins/module_utils/client/resource.py import-3.9!skip
plugins/module_utils/client/resource.py import-3.10!skip
plugins/module_utils/client/resource.py import-3.11!skip
plugins/module_utils/client/resource.py import-3.12!skip
plugins/module_utils/client/resource.py import-3.13!skip
plugins/module_utils/k8sdynamicclient.py import-3.9!skip
plugins/module_utils/k8sdynamicclient.py import-3.10!skip
plugins/module_utils/k8sdynamicclient.py import-3.11!skip
plugins/module_utils/k8sdynamicclient.py import-3.12!skip
plugins/module_utils/k8sdynamicclient.py import-3.13!skip
@@ -18,6 +24,7 @@ tests/integration/targets/k8s_delete/files/deployments.yaml yamllint!skip
tests/unit/module_utils/fixtures/pods.yml yamllint!skip
tests/integration/targets/helm/files/appversionless-chart-v2/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/appversionless-chart/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart-job/Chart.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart-v2/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm_diff/files/test-chart/templates/configmap.yaml yamllint!skip
+7
View File
@@ -1,11 +1,17 @@
plugins/module_utils/client/discovery.py import-3.9!skip
plugins/module_utils/client/discovery.py import-3.10!skip
plugins/module_utils/client/discovery.py import-3.11!skip
plugins/module_utils/client/discovery.py import-3.12!skip
plugins/module_utils/client/discovery.py import-3.13!skip
plugins/module_utils/client/discovery.py import-3.14!skip
plugins/module_utils/client/resource.py import-3.9!skip
plugins/module_utils/client/resource.py import-3.10!skip
plugins/module_utils/client/resource.py import-3.11!skip
plugins/module_utils/client/resource.py import-3.12!skip
plugins/module_utils/client/resource.py import-3.13!skip
plugins/module_utils/client/resource.py import-3.14!skip
plugins/module_utils/k8sdynamicclient.py import-3.9!skip
plugins/module_utils/k8sdynamicclient.py import-3.10!skip
plugins/module_utils/k8sdynamicclient.py import-3.11!skip
plugins/module_utils/k8sdynamicclient.py import-3.12!skip
plugins/module_utils/k8sdynamicclient.py import-3.13!skip
@@ -21,6 +27,7 @@ tests/integration/targets/k8s_delete/files/deployments.yaml yamllint!skip
tests/unit/module_utils/fixtures/pods.yml yamllint!skip
tests/integration/targets/helm/files/appversionless-chart-v2/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/appversionless-chart/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart-job/Chart.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart-v2/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm_diff/files/test-chart/templates/configmap.yaml yamllint!skip
+7
View File
@@ -1,11 +1,17 @@
plugins/module_utils/client/discovery.py import-3.9!skip
plugins/module_utils/client/discovery.py import-3.10!skip
plugins/module_utils/client/discovery.py import-3.11!skip
plugins/module_utils/client/discovery.py import-3.12!skip
plugins/module_utils/client/discovery.py import-3.13!skip
plugins/module_utils/client/discovery.py import-3.14!skip
plugins/module_utils/client/resource.py import-3.9!skip
plugins/module_utils/client/resource.py import-3.10!skip
plugins/module_utils/client/resource.py import-3.11!skip
plugins/module_utils/client/resource.py import-3.12!skip
plugins/module_utils/client/resource.py import-3.13!skip
plugins/module_utils/client/resource.py import-3.14!skip
plugins/module_utils/k8sdynamicclient.py import-3.9!skip
plugins/module_utils/k8sdynamicclient.py import-3.10!skip
plugins/module_utils/k8sdynamicclient.py import-3.11!skip
plugins/module_utils/k8sdynamicclient.py import-3.12!skip
plugins/module_utils/k8sdynamicclient.py import-3.13!skip
@@ -21,6 +27,7 @@ tests/integration/targets/k8s_delete/files/deployments.yaml yamllint!skip
tests/unit/module_utils/fixtures/pods.yml yamllint!skip
tests/integration/targets/helm/files/appversionless-chart-v2/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/appversionless-chart/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart-job/Chart.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart-v2/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm_diff/files/test-chart/templates/configmap.yaml yamllint!skip
+10
View File
@@ -1,15 +1,24 @@
plugins/module_utils/client/discovery.py import-3.9!skip
plugins/module_utils/client/discovery.py import-3.10!skip
plugins/module_utils/client/discovery.py import-3.11!skip
plugins/module_utils/client/discovery.py import-3.12!skip
plugins/module_utils/client/discovery.py import-3.13!skip
plugins/module_utils/client/discovery.py import-3.14!skip
plugins/module_utils/client/discovery.py import-3.15!skip
plugins/module_utils/client/resource.py import-3.9!skip
plugins/module_utils/client/resource.py import-3.10!skip
plugins/module_utils/client/resource.py import-3.11!skip
plugins/module_utils/client/resource.py import-3.12!skip
plugins/module_utils/client/resource.py import-3.13!skip
plugins/module_utils/client/resource.py import-3.14!skip
plugins/module_utils/client/resource.py import-3.15!skip
plugins/module_utils/k8sdynamicclient.py import-3.9!skip
plugins/module_utils/k8sdynamicclient.py import-3.10!skip
plugins/module_utils/k8sdynamicclient.py import-3.11!skip
plugins/module_utils/k8sdynamicclient.py import-3.12!skip
plugins/module_utils/k8sdynamicclient.py import-3.13!skip
plugins/module_utils/k8sdynamicclient.py import-3.14!skip
plugins/module_utils/k8sdynamicclient.py import-3.15!skip
plugins/module_utils/version.py pylint!skip
plugins/modules/k8s.py validate-modules:parameter-type-not-in-doc
plugins/modules/k8s_scale.py validate-modules:parameter-type-not-in-doc
@@ -21,6 +30,7 @@ tests/integration/targets/k8s_delete/files/deployments.yaml yamllint!skip
tests/unit/module_utils/fixtures/pods.yml yamllint!skip
tests/integration/targets/helm/files/appversionless-chart-v2/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/appversionless-chart/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart-job/Chart.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart-v2/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm/files/test-chart/templates/configmap.yaml yamllint!skip
tests/integration/targets/helm_diff/files/test-chart/templates/configmap.yaml yamllint!skip
+1 -1
View File
@@ -4,11 +4,11 @@ __metaclass__ = type
import json
import sys
from collections.abc import MutableMapping
from io import BytesIO
import ansible.module_utils.basic
import pytest
from ansible.module_utils.common._collections_compat import MutableMapping
from ansible.module_utils.common.text.converters import to_bytes
@@ -0,0 +1,111 @@
# -*- coding: utf-8 -*-
# Copyright: (c) 2026, Ansible Project
# GNU General Public License v3.0+ (see COPYING or https://www.gnu.org/licenses/gpl-3.0.txt)
from __future__ import absolute_import, division, print_function
__metaclass__ = type
import pytest
import yaml
from ansible_collections.kubernetes.core.plugins.module_utils.helm import (
strip_registry_progress,
)
# What helm's registry client prints, verbatim. Since helm 4.2.1 this goes to
# stdout ahead of the command's real output (helm/helm#32056).
PULLED = "Pulled: registry.example.com/charts/demochart:109.0.1_up1.0.2\n"
DIGEST = (
"Digest: sha256:41e9b37f85fa7f5df5fc7f2b7b068eb56ae1592784eee28f535563b599850249\n"
)
PUSHED = "Pushed: registry.example.com/charts/demochart:109.0.1_up1.0.2\n"
UNDERSCORE_WARNING = (
"registry.example.com/charts/demochart:109.0.1_up1.0.2 contains an underscore.\n"
"\n"
"OCI artifact references (e.g. tags) do not support the plus sign (+). To support\n"
"storing semantic versions, Helm adopts the convention of changing plus (+) to\n"
"an underscore (_) in chart version tags when pushing to a registry and back to\n"
"a plus (+) when pulling from a registry.\n"
)
# 'helm show chart' output. Note that helm marshals Chart.yaml with sorted keys, so
# 'annotations' precedes 'apiVersion' and the metadata does not reliably start with
# a fixed key.
CHART_YAML = (
"annotations:\n"
" category: Demo\n"
"apiVersion: v2\n"
"appVersion: 1.16.0\n"
"description: A Helm chart for Kubernetes\n"
"name: demochart\n"
"type: application\n"
"version: 109.0.1+up1.0.2\n"
)
# 'helm template' output.
MANIFESTS = (
"---\n"
"# Source: demochart/templates/serviceaccount.yaml\n"
"apiVersion: v1\n"
"kind: ServiceAccount\n"
"metadata:\n"
" name: rel-demochart\n"
)
# 'helm template --output-dir' output.
WROTE = (
"wrote outdir/demochart/templates/serviceaccount.yaml\n"
"wrote outdir/demochart/templates/service.yaml\n"
)
@pytest.mark.parametrize(
"payload",
[CHART_YAML, MANIFESTS, WROTE, ""],
ids=["show_chart", "template", "template_output_dir", "empty"],
)
@pytest.mark.parametrize(
"progress",
[
"",
PULLED + DIGEST,
PULLED + DIGEST + UNDERSCORE_WARNING,
PUSHED + DIGEST + UNDERSCORE_WARNING,
],
ids=["none", "pulled_digest", "with_warning", "pushed_with_warning"],
)
def test_progress_is_stripped_and_payload_is_untouched(progress, payload):
assert strip_registry_progress(progress + payload) == payload
def test_show_chart_output_is_parseable_once_stripped():
# Unstripped, the free-text warning makes this invalid YAML.
out = PULLED + DIGEST + UNDERSCORE_WARNING + CHART_YAML
with pytest.raises(yaml.YAMLError):
yaml.safe_load(out)
chart_info = yaml.safe_load(strip_registry_progress(out))
assert chart_info["name"] == "demochart"
assert chart_info["version"] == "109.0.1+up1.0.2"
assert chart_info["annotations"] == {"category": "Demo"}
def test_pulled_and_digest_alone_parse_as_yaml_but_pollute_the_result():
# Without the warning the output is still valid YAML, so the pollution is silent.
assert yaml.safe_load(PULLED + DIGEST + CHART_YAML)["Pulled"]
assert "Pulled" not in yaml.safe_load(
strip_registry_progress(PULLED + DIGEST + CHART_YAML)
)
def test_reworded_warning_is_left_alone():
# A message helm no longer emits verbatim must not be partially consumed: better
# to fail loudly than to silently drop a line of real output.
reworded = UNDERSCORE_WARNING.replace("plus sign (+)", "plus character (+)")
out = PULLED + DIGEST + reworded + CHART_YAML
assert strip_registry_progress(out) == reworded + CHART_YAML
def test_progress_like_lines_inside_the_payload_are_kept():
payload = "description: |\n Digest: not a helm message\n" + CHART_YAML
assert strip_registry_progress(PULLED + payload) == payload
+298
View File
@@ -0,0 +1,298 @@
# Copyright [2026] [Red Hat, Inc.]
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# Tests for the to_pod streaming path of K8SCopyToPod. The v4 exec protocol
# cannot half-close stdin, so the archive is bounded with `head -c` to give the
# remote tar a clean EOF; the module must then keep reading until the API server
# closes the connection and check the status reported on the error channel. See
# https://github.com/ansible-collections/kubernetes.core/issues/776
from __future__ import absolute_import, division, print_function
__metaclass__ = type
import io
import pytest
from ansible_collections.kubernetes.core.plugins.module_utils.copy import K8SCopyToPod
ERROR_CHANNEL = 3
SUCCESS = '{"metadata":{},"status":"Success"}'
FAILURE = (
'{"metadata":{},"status":"Failure","message":"command terminated with'
' non-zero exit code","reason":"NonZeroExitCode"}'
)
class FakeModuleFailure(Exception):
"""Stands in for the SystemExit that AnsibleModule.fail_json raises."""
def __init__(self, **kwargs):
self.kwargs = kwargs
super(FakeModuleFailure, self).__init__(kwargs.get("msg"))
class FakeModule:
def __init__(self, **params):
params.setdefault("copy_timeout", 300)
self.params = params
self.check_mode = False
self.warnings = []
def fail_json(self, **kwargs):
raise FakeModuleFailure(**kwargs)
def warn(self, warning):
self.warnings.append(warning)
class FakeWSClient:
"""Minimal stand-in for kubernetes.stream.ws_client.WSClient.
``close_after`` is the number of ``update()`` calls the connection stays
open for once the last chunk has been written, modelling the remote tar
consuming the archive and exiting.
"""
def __init__(
self,
close_after=2,
error=SUCCESS,
stderr=None,
drop_after=None,
stderr_early=False,
):
self.written = []
self.closed = False
self._open = True
self._updates = 0
self._close_after = close_after
self._error = error
self._pending_stderr = stderr
self._drop_after = drop_after
self._stderr_early = stderr_early
self.writes_done = False
def is_open(self):
return self._open
def update(self, timeout=0):
self._updates += 1
if self.writes_done and self._updates > self._close_after:
self._open = False
def write_stdin(self, data):
if self._drop_after is not None and len(self.written) >= self._drop_after:
self._open = False
return
self.written.append(data)
def peek_stdout(self):
return ""
def read_stdout(self):
return ""
def peek_stderr(self):
# tar only complains once it has started extracting, i.e. after the
# whole archive has been handed over, unless it failed outright.
if not (self.writes_done or self._stderr_early):
return ""
return self._pending_stderr or ""
def read_stderr(self):
out, self._pending_stderr = self._pending_stderr, None
return out
def read_channel(self, channel):
assert channel == ERROR_CHANNEL
return self._error
def close(self):
self.closed = True
self._open = False
def make_copier(has_shell=True, no_preserve=False, **params):
module = FakeModule(**params)
copier = K8SCopyToPod.__new__(K8SCopyToPod)
copier.module = module
copier.namespace = "testing"
copier.name = "some-pod"
copier.named_temp_file = None
copier.no_preserve = no_preserve
copier.copy_timeout = module.params["copy_timeout"]
copier.can_bound_stream = lambda: has_shell
return copier
def stream_archive(copier, response, size, wait_for_exit=True):
"""Drive _stream_tar_to_pod, marking the writes finished for the fake."""
payload = io.BytesIO(b"x" * size)
original_read = payload.read
def read(n):
data = original_read(n)
if not data:
response.writes_done = True
return data
payload.read = read
return copier._stream_tar_to_pod(response, payload, wait_for_exit)
def test_waits_for_tar_to_exit_before_closing():
"""The connection must not be closed until the remote process is done."""
copier = make_copier()
response = FakeWSClient(close_after=5)
error, stdout, stderr = stream_archive(
copier, response, 3 * K8SCopyToPod.CHUNK_SIZE
)
assert error["status"] == "Success"
assert stdout == []
assert stderr == []
# Whole archive sent, in CHUNK_SIZE pieces.
assert len(response.written) == 3
assert sum(len(c) for c in response.written) == 3 * K8SCopyToPod.CHUNK_SIZE
# Closed only after the server hung up, not right after the last write.
assert response.closed
assert response._updates > 3
def test_partial_write_is_not_reported_as_success():
"""A connection dropped mid-archive must fail, not silently truncate."""
copier = make_copier()
response = FakeWSClient(drop_after=2)
with pytest.raises(FakeModuleFailure) as exc:
stream_archive(copier, response, 5 * K8SCopyToPod.CHUNK_SIZE)
assert "the remote file is incomplete" in exc.value.kwargs["msg"]
assert len(response.written) == 2
assert response.closed
def test_early_close_reports_what_tar_said():
"""The remote error must not be masked by the generic truncation message."""
copier = make_copier()
response = FakeWSClient(
drop_after=2,
stderr="tar: /etc/foo: Cannot open: Read-only file system",
stderr_early=True,
)
with pytest.raises(FakeModuleFailure) as exc:
stream_archive(copier, response, 5 * K8SCopyToPod.CHUNK_SIZE)
assert "the remote file is incomplete" in exc.value.kwargs["msg"]
assert "Read-only file system" in exc.value.kwargs["msg"]
def test_non_zero_tar_exit_is_surfaced():
"""A Failure status on the error channel must reach the caller."""
copier = make_copier()
response = FakeWSClient(error=FAILURE)
error, _stdout, _stderr = stream_archive(copier, response, K8SCopyToPod.CHUNK_SIZE)
assert error["status"] == "Failure"
assert error["reason"] == "NonZeroExitCode"
def test_stderr_is_collected_while_draining():
"""Output tar writes after the last chunk must still be captured."""
copier = make_copier()
response = FakeWSClient(stderr="tar: /foo: Cannot utime: Operation not permitted")
_error, _stdout, stderr = stream_archive(copier, response, K8SCopyToPod.CHUNK_SIZE)
assert stderr == ["tar: /foo: Cannot utime: Operation not permitted"]
def test_drain_honours_copy_timeout(monkeypatch):
"""A remote tar that never exits must time out instead of hanging."""
copier = make_copier(copy_timeout=30)
response = FakeWSClient(close_after=10**6)
# Time only moves past the deadline once the archive has been sent, so the
# timeout is exercised in the drain loop rather than the write loop.
monkeypatch.setattr(
"time.monotonic", lambda: 10**6 if response.writes_done else 0
)
with pytest.raises(FakeModuleFailure) as exc:
stream_archive(copier, response, K8SCopyToPod.CHUNK_SIZE)
assert "waiting for tar to finish extracting" in exc.value.kwargs["msg"]
assert response.closed
def test_archive_is_bounded_with_head():
"""head -c gives the remote tar a clean EOF the v4 protocol cannot."""
copier = make_copier()
command, wait_for_exit = copier.remote_command("/tmp/foo", 1003520)
assert wait_for_exit is True
assert command == ["/bin/sh", "-c", "head -c 1003520 | tar -xmf - -C /"]
assert copier.module.warnings == []
def test_bounded_command_carries_no_preserve():
copier = make_copier(no_preserve=True)
command, _wait = copier.remote_command("/tmp/foo", 512)
assert command[2] == (
"head -c 512 | tar --no-same-permissions --no-same-owner -xmf - -C /"
)
def test_relative_destination_is_not_anchored():
"""Only absolute destinations get -C /, as before."""
copier = make_copier()
command, _wait = copier.remote_command("tmp/foo", 512)
assert command[2] == "head -c 512 | tar -xmf -"
def test_shell_less_container_falls_back_and_warns():
"""Without a shell the copy still runs, but completion is unverifiable."""
copier = make_copier(has_shell=False)
command, wait_for_exit = copier.remote_command("/tmp/foo", 512)
assert wait_for_exit is False
assert command == ["tar", "-xmf", "-", "-C", "/"]
assert len(copier.module.warnings) == 1
assert "cannot confirm the copy" in copier.module.warnings[0]
def test_fallback_path_does_not_wait_for_exit():
"""The legacy path must not block on a tar that will never exit."""
copier = make_copier(has_shell=False)
response = FakeWSClient(close_after=10**6)
error, _stdout, _stderr = stream_archive(
copier, response, K8SCopyToPod.CHUNK_SIZE, wait_for_exit=False
)
# No exit status is available on this path.
assert error is None
assert response.closed
assert sum(len(c) for c in response.written) == K8SCopyToPod.CHUNK_SIZE
+470 -33
View File
@@ -23,6 +23,79 @@ from ansible_collections.kubernetes.core.tests.unit.utils.ansible_module_mock im
set_module_args,
)
CHART_INFO = {
"apiVersion": "v2",
"appVersion": "default",
"description": "A chart used in molecule tests",
"name": "test-chart",
"type": "application",
"version": "0.1.0",
}
def patch_ansible_module(test_case):
"""Patch AnsibleModule's exit/fail/bin-path hooks for the duration of test_case."""
mock_module_helper = patch.multiple(
basic.AnsibleModule,
exit_json=exit_json,
fail_json=fail_json,
get_bin_path=get_bin_path,
)
mock_module_helper.start()
test_case.addCleanup(mock_module_helper.stop)
def deploy_command_for_version(args, helm_version):
"""Run helm.main() against a mocked helm of helm_version and return its command.
Only the fresh-install path is mocked, so the caller sees the command built by
deploy(). Failures raised by the module (AnsibleFailJson) propagate to the test.
"""
set_module_args(args)
helm.get_release_status = MagicMock(return_value=None)
helm.fetch_chart_info = MagicMock(return_value=CHART_INFO)
with patch.object(
helm.AnsibleHelmModule, "get_helm_version", return_value=helm_version
):
with patch.object(basic.AnsibleModule, "run_command") as mock_run_command:
mock_run_command.return_value = (0, "configuration updated", "")
try:
helm.main()
except AnsibleExitJson as exc:
return exc.args[0]["command"]
raise AssertionError("helm.main() did not exit successfully")
def diff_command_for_version(args, helm_version="3.17.0", helm_diff_version="3.14.0"):
"""Run helm.main() against an existing release and return the 'helm diff' command.
The release is reported as deployed and helm diff as installed, so main() takes the
idempotency branch. The diff output is empty, so nothing is deployed afterwards and
the first command run is the one helmdiff_check() built.
"""
set_module_args(args)
helm.get_release_status = MagicMock(
return_value={
"status": "deployed",
"values": {},
"release_values": {},
"chart": "{0}-{1}".format(CHART_INFO["name"], CHART_INFO["version"]),
"app_version": CHART_INFO["appVersion"],
}
)
helm.fetch_chart_info = MagicMock(return_value=CHART_INFO)
helm.get_plugin_version = MagicMock(return_value=helm_diff_version)
with patch.object(
helm.AnsibleHelmModule, "get_helm_version", return_value=helm_version
):
with patch.object(basic.AnsibleModule, "run_command") as mock_run_command:
mock_run_command.return_value = (0, "", "")
try:
helm.main()
except AnsibleExitJson:
return mock_run_command.call_args_list[0][0][0]
raise AssertionError("helm.main() did not exit successfully")
class TestDependencyUpdateWithoutChartRepoUrlOption(unittest.TestCase):
def setUp(self):
@@ -716,39 +789,10 @@ class TestAtomicFlagHelmVersion(unittest.TestCase):
"""Helm v4 renamed '--atomic' to '--rollback-on-failure' (issue #1143)."""
def setUp(self):
self.mock_module_helper = patch.multiple(
basic.AnsibleModule,
exit_json=exit_json,
fail_json=fail_json,
get_bin_path=get_bin_path,
)
self.mock_module_helper.start()
self.addCleanup(self.mock_module_helper.stop)
self.chart_info = {
"apiVersion": "v2",
"appVersion": "default",
"description": "A chart used in molecule tests",
"name": "test-chart",
"type": "application",
"version": "0.1.0",
}
def _run_with_version(self, args, helm_version):
set_module_args(args)
helm.get_release_status = MagicMock(return_value=None)
helm.fetch_chart_info = MagicMock(return_value=self.chart_info)
with patch.object(
helm.AnsibleHelmModule, "get_helm_version", return_value=helm_version
):
with patch.object(basic.AnsibleModule, "run_command") as mock_run_command:
mock_run_command.return_value = (0, "configuration updated", "")
with self.assertRaises(AnsibleExitJson) as result:
helm.main()
return result.exception.args[0]["command"]
patch_ansible_module(self)
def test_atomic_uses_atomic_flag_on_helm_v3(self):
command = self._run_with_version(
command = deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
@@ -761,7 +805,7 @@ class TestAtomicFlagHelmVersion(unittest.TestCase):
assert "--rollback-on-failure" not in command
def test_atomic_uses_rollback_on_failure_on_helm_v4(self):
command = self._run_with_version(
command = deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
@@ -775,7 +819,7 @@ class TestAtomicFlagHelmVersion(unittest.TestCase):
def test_atomic_with_replace_uses_rollback_on_failure_on_helm_v4(self):
# 'replace' uses 'helm install', where '--atomic' is removed in v4.
command = self._run_with_version(
command = deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
@@ -788,3 +832,396 @@ class TestAtomicFlagHelmVersion(unittest.TestCase):
assert " install " in command
assert "--rollback-on-failure" in command
assert "--atomic" not in command
class TestCleanupOnFail(unittest.TestCase):
"""'--cleanup-on-fail' is upgrade-only and unchanged in Helm v4 (issue #1198)."""
def setUp(self):
patch_ansible_module(self)
def test_cleanup_on_fail_not_set_by_default(self):
command = deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
},
"3.17.0",
)
assert "--cleanup-on-fail" not in command
def test_cleanup_on_fail_on_helm_v3(self):
# '--cleanup-on-fail' exists since helm 3.0.0, the module-wide minimum.
command = deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"cleanup_on_fail": True,
},
"3.0.0",
)
assert " upgrade -i" in command
assert "--cleanup-on-fail" in command
def test_cleanup_on_fail_keeps_same_flag_on_helm_v4(self):
# Unlike '--atomic', the flag was not renamed in Helm v4.
command = deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"cleanup_on_fail": True,
},
"4.0.0",
)
assert "--cleanup-on-fail" in command
def test_cleanup_on_fail_combines_with_atomic_on_helm_v3(self):
command = deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"atomic": True,
"cleanup_on_fail": True,
},
"3.17.0",
)
assert "--atomic" in command
assert "--rollback-on-failure" not in command
assert "--cleanup-on-fail" in command
def test_cleanup_on_fail_combines_with_atomic_on_helm_v4(self):
command = deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"atomic": True,
"cleanup_on_fail": True,
},
"4.0.0",
)
assert "--rollback-on-failure" in command
assert "--atomic" not in command
assert "--cleanup-on-fail" in command
def test_cleanup_on_fail_with_replace_fails(self):
# 'replace' deploys via 'helm install', which has no '--cleanup-on-fail'.
with self.assertRaises(AnsibleFailJson):
deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"cleanup_on_fail": True,
"replace": True,
},
"3.17.0",
)
class TestValueReuseOptions(unittest.TestCase):
"""reset_values wins over reuse_values, and both are upgrade-only (issue #1229)."""
def setUp(self):
patch_ansible_module(self)
def test_reuse_values_alone_is_overridden_by_the_reset_values_default(self):
# What the pre-fix example in the docs produced: helm ignores '--reuse-values'.
command = deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"reuse_values": True,
},
"3.17.0",
)
assert "--reset-values" in command
assert "--reuse-values=True" in command
def test_reuse_values_with_reset_values_disabled(self):
command = deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"reuse_values": True,
"reset_values": False,
},
"3.17.0",
)
assert "--reset-values" not in command
assert "--reuse-values=True" in command
def test_reset_then_reuse_values_with_reset_values_disabled(self):
command = deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"reset_then_reuse_values": True,
"reset_values": False,
},
"3.17.0",
)
assert "--reset-values" not in command
assert "--reset-then-reuse-values" in command
def test_reuse_values_with_replace_is_dropped_and_warns(self):
# 'replace' deploys via 'helm install', which has no '--reuse-values'.
with patch.object(basic.AnsibleModule, "warn") as mock_warn:
command = deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"reuse_values": True,
"replace": True,
},
"3.17.0",
)
assert " install" in command
assert "--reuse-values" not in command
warnings = [call.args[0] for call in mock_warn.call_args_list]
assert any(
"reuse_values is ignored because replace=true" in warning
for warning in warnings
)
def test_reuse_values_false_with_replace_is_dropped_and_warns(self):
# Even 'reuse_values: false' used to emit '--reuse-values=False' on install.
with patch.object(basic.AnsibleModule, "warn") as mock_warn:
command = deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"reuse_values": False,
"replace": True,
},
"3.17.0",
)
assert "--reuse-values" not in command
warnings = [call.args[0] for call in mock_warn.call_args_list]
assert any(
"reuse_values is ignored because replace=true" in warning
for warning in warnings
)
def test_reset_then_reuse_values_with_replace_is_dropped_and_warns(self):
with patch.object(basic.AnsibleModule, "warn") as mock_warn:
command = deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"reset_then_reuse_values": True,
"replace": True,
},
"3.17.0",
)
assert "--reset-then-reuse-values" not in command
warnings = [call.args[0] for call in mock_warn.call_args_list]
assert any(
"reset_then_reuse_values is ignored because replace=true" in warning
for warning in warnings
)
def test_replace_alone_keeps_working(self):
command = deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"replace": True,
},
"3.17.0",
)
assert " install" in command
assert "--reuse-values" not in command
assert "--reset-values" not in command
def test_reuse_values_alone_warns(self):
with patch.object(basic.AnsibleModule, "warn") as mock_warn:
deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"reuse_values": True,
},
"3.17.0",
)
warnings = [call.args[0] for call in mock_warn.call_args_list]
assert any("reuse_values is ignored" in warning for warning in warnings)
def test_reset_then_reuse_values_alone_warns(self):
with patch.object(basic.AnsibleModule, "warn") as mock_warn:
deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"reset_then_reuse_values": True,
},
"3.17.0",
)
warnings = [call.args[0] for call in mock_warn.call_args_list]
assert any(
"reset_then_reuse_values is ignored" in warning for warning in warnings
)
def test_reset_then_reuse_values_warns_when_reuse_values_wins(self):
with patch.object(basic.AnsibleModule, "warn") as mock_warn:
deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"reuse_values": True,
"reset_then_reuse_values": True,
"reset_values": False,
},
"3.17.0",
)
warnings = [call.args[0] for call in mock_warn.call_args_list]
assert any(
"reset_then_reuse_values is ignored because reuse_values is true" in warning
for warning in warnings
)
def test_no_warning_when_options_are_combined_correctly(self):
with patch.object(basic.AnsibleModule, "warn") as mock_warn:
deploy_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"reuse_values": True,
"reset_values": False,
},
"3.17.0",
)
assert mock_warn.call_args_list == []
class TestValueReuseOptionsWithHelmDiff(unittest.TestCase):
"""The same options on the idempotency path, where 'helm diff' builds the command."""
def setUp(self):
patch_ansible_module(self)
def test_reuse_values_alone_is_overridden_by_the_reset_values_default(self):
# helm diff is handed both flags and applies helm's precedence, which is why
# 'reuse_values' on its own still reports a change (issue #1229).
command = diff_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"reuse_values": True,
}
)
assert "--reset-values=True" in command
assert "--reuse-values" in command
def test_reuse_values_with_reset_values_disabled(self):
command = diff_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"reuse_values": True,
"reset_values": False,
}
)
assert "--reset-values=False" in command
assert "--reuse-values" in command
def test_reset_then_reuse_values_with_reset_values_disabled(self):
command = diff_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"reset_then_reuse_values": True,
"reset_values": False,
}
)
assert "--reset-values=False" in command
assert "--reset-then-reuse-values" in command
def test_reuse_values_with_replace_is_dropped_and_warns(self):
with patch.object(basic.AnsibleModule, "warn") as mock_warn:
command = diff_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"reuse_values": True,
"replace": True,
}
)
assert "--reuse-values" not in command
assert "--reset-values" not in command
warnings = [call.args[0] for call in mock_warn.call_args_list]
assert any(
"reuse_values is ignored because replace=true" in warning
for warning in warnings
)
def test_reset_then_reuse_values_with_replace_is_dropped_and_warns(self):
with patch.object(basic.AnsibleModule, "warn") as mock_warn:
command = diff_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"reset_then_reuse_values": True,
"replace": True,
}
)
assert "--reset-then-reuse-values" not in command
warnings = [call.args[0] for call in mock_warn.call_args_list]
assert any(
"reset_then_reuse_values is ignored because replace=true" in warning
for warning in warnings
)
def test_reset_then_reuse_values_requires_helm_3_14(self):
with self.assertRaises(AnsibleFailJson) as result:
diff_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"reset_then_reuse_values": True,
"reset_values": False,
},
helm_version="3.13.0",
)
assert (
"reset_then_reuse_values requires helm >= 3.14.0"
in result.exception.args[0]["msg"]
)
def test_reset_then_reuse_values_requires_helm_diff_3_9_12(self):
with self.assertRaises(AnsibleFailJson) as result:
diff_command_for_version(
{
"release_name": "test",
"release_namespace": "test",
"chart_ref": "chart1",
"reset_then_reuse_values": True,
"reset_values": False,
},
helm_diff_version="3.9.0",
)
assert (
"reset_then_reuse_values requires helm diff >= 3.9.12"
in result.exception.args[0]["msg"]
)
@@ -0,0 +1,89 @@
from __future__ import absolute_import, division, print_function
__metaclass__ = type
import sys
from unittest.mock import MagicMock, patch
import pytest
from ansible.plugins.lookup import LookupBase as StandardLookupBase
class TestK8sTurboMode:
"""Test turbo mode version checking in k8s lookup plugin."""
@pytest.fixture(autouse=True)
def cleanup_k8s_module(self):
"""Ensure clean module state before each test."""
sys.modules.pop("ansible_collections.kubernetes.core.plugins.lookup.k8s", None)
@pytest.mark.parametrize("ansible_version", ["2.19.0", "2.21.4"])
def test_turbo_mode_disabled_on_newer_ansible(self, monkeypatch, ansible_version):
"""Test that turbo mode is disabled on ansible-core >= 2.19.0."""
monkeypatch.setenv("ENABLE_TURBO_MODE", "true")
mock_display = MagicMock()
with patch.dict(
"sys.modules",
{
"ansible.module_utils.ansible_release": MagicMock(
__version__=ansible_version
),
"ansible.utils.display": MagicMock(Display=lambda: mock_display),
},
):
from ansible_collections.kubernetes.core.plugins.lookup import k8s
mock_display.deprecated.assert_called()
assert (
"ENABLE_TURBO_MODE is deprecated"
in mock_display.deprecated.call_args[0][0]
)
mock_display.warning.assert_called()
assert (
f"ENABLE_TURBO_MODE is ignored on ansible-core {ansible_version}"
in mock_display.warning.call_args[0][0]
)
assert "ansible-core < 2.19.0 only" in mock_display.warning.call_args[0][0]
assert k8s.LookupBase == StandardLookupBase
def test_turbo_mode_enabled_on_ansible_core_2_18(self, monkeypatch):
"""Test that turbo mode is still enabled on ansible-core < 2.19.0."""
monkeypatch.setenv("ENABLE_TURBO_MODE", "true")
mock_display = MagicMock()
mock_turbo_lookup = MagicMock()
with patch.dict(
"sys.modules",
{
"ansible.module_utils.ansible_release": MagicMock(__version__="2.18.0"),
"ansible.utils.display": MagicMock(Display=lambda: mock_display),
"ansible_collections.cloud.common.plugins.plugin_utils.turbo.lookup": mock_turbo_lookup,
},
):
from ansible_collections.kubernetes.core.plugins.lookup import ( # noqa: F401
k8s,
)
mock_display.deprecated.assert_called()
assert not mock_display.warning.called
def test_turbo_mode_disabled_when_env_not_set(self, monkeypatch):
"""Test that turbo mode is disabled when ENABLE_TURBO_MODE is not set."""
monkeypatch.delenv("ENABLE_TURBO_MODE", raising=False)
mock_display = MagicMock()
with patch.dict(
"sys.modules",
{
"ansible.module_utils.ansible_release": MagicMock(__version__="2.18.0"),
"ansible.utils.display": MagicMock(Display=lambda: mock_display),
},
):
from ansible_collections.kubernetes.core.plugins.lookup import k8s
mock_display.deprecated.assert_not_called()
mock_display.warning.assert_not_called()
assert k8s.LookupBase == StandardLookupBase
+156
View File
@@ -0,0 +1,156 @@
# Copyright (c) 2026 Yuriy Novostavskiy
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
# implied. See the License for the specific language governing
# permissions and limitations under the License.
#
# This file is licensed permissively rather than under the collection's
# GPL-3.0-or-later because it lives outside plugins/ and imports no
# GPL-licensed code: membership is read statically with ast, so nothing from
# ansible-core or plugins/ is imported. See the collection requirements,
# "Licensing" section.
# Guards the action groups declared in meta/runtime.yml.
#
# A module belongs in an action group if and only if it accepts that group's
# shared authentication options, because module_defaults set on a group is
# applied to every member. A member that does not accept the options fails with
# "Unsupported parameters"; a non-member that does accept them silently ignores
# the defaults the user set. Both directions are bugs, so both are asserted.
#
# Membership is derived from whether a module builds its argument spec from the
# shared constant, which is how every module in the collection does it today.
import ast
import os
import pytest
import yaml
COLLECTION_ROOT = os.path.dirname(
os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
)
MODULES_DIR = os.path.join(COLLECTION_ROOT, "plugins", "modules")
RUNTIME_YML = os.path.join(COLLECTION_ROOT, "meta", "runtime.yml")
# action group -> the module_utils file defining its shared spec, and the constant name
SHARED_ARG_SPECS = {
"k8s": {"module_utils": "args_common", "constant": "AUTH_ARG_SPEC"},
"helm": {"module_utils": "helm_args_common", "constant": "HELM_AUTH_ARG_SPEC"},
}
# Modules that intentionally stay out of the groups, with the reason. Listed
# explicitly so that adding one back is a deliberate decision rather than an
# oversight. See https://github.com/ansible-collections/kubernetes.core/issues/577
EXCLUDED_FROM_GROUPS = {
"helm_pull": "accepts only binary_path, not the shared helm auth options",
"helm_template": "accepts only binary_path, not the shared helm auth options",
"helm_registry_auth": "its 'host' is an OCI registry URL, not a Kubernetes API server",
"kubeconfig": "writes a kubeconfig file locally, takes no cluster connection options",
}
def _module_names():
return sorted(
name[:-3]
for name in os.listdir(MODULES_DIR)
if name.endswith(".py") and name != "__init__.py"
)
def _shared_specs_used_by(module_name):
"""Return the shared arg-spec constants a module imports."""
path = os.path.join(MODULES_DIR, module_name + ".py")
with open(path, "rb") as handle:
tree = ast.parse(handle.read(), filename=path)
group_of_constant = {
spec["constant"]: group for group, spec in SHARED_ARG_SPECS.items()
}
modules_of_interest = {spec["module_utils"] for spec in SHARED_ARG_SPECS.values()}
used = set()
for node in ast.walk(tree):
if not isinstance(node, ast.ImportFrom) or not node.module:
continue
if node.module.rsplit(".", 1)[-1] not in modules_of_interest:
continue
for alias in node.names:
if alias.name in group_of_constant:
used.add(group_of_constant[alias.name])
return used
@pytest.fixture(scope="module")
def action_groups():
with open(RUNTIME_YML) as handle:
return yaml.safe_load(handle)["action_groups"]
def test_runtime_declares_expected_action_groups(action_groups):
assert set(action_groups) == set(SHARED_ARG_SPECS), (
"meta/runtime.yml declares action groups this test does not know about. "
"Add the new group and its shared arg-spec constant to SHARED_ARG_SPECS."
)
@pytest.mark.parametrize("group", sorted(SHARED_ARG_SPECS))
def test_action_group_members_exist(action_groups, group):
known = set(_module_names())
missing = sorted(set(action_groups[group]) - known)
assert not missing, (
"meta/runtime.yml lists modules in the '%s' action group that do not exist "
"under plugins/modules/: %s" % (group, ", ".join(missing))
)
@pytest.mark.parametrize("group", sorted(SHARED_ARG_SPECS))
def test_action_group_matches_shared_arg_spec_users(action_groups, group):
constant = SHARED_ARG_SPECS[group]["constant"]
declared = set(action_groups[group])
actual = {name for name in _module_names() if group in _shared_specs_used_by(name)}
missing_from_group = sorted(actual - declared)
assert not missing_from_group, (
"%s build their argument spec from %s but are not listed in the '%s' action "
"group, so module_defaults set on that group silently does not reach them. "
"Add them to meta/runtime.yml."
% (", ".join(missing_from_group), constant, group)
)
not_shared = sorted(declared - actual)
assert not not_shared, (
"%s are listed in the '%s' action group but do not build their argument spec "
"from %s, so module_defaults set on that group will fail for them with "
"'Unsupported parameters'. Remove them from meta/runtime.yml."
% (", ".join(not_shared), group, constant)
)
@pytest.mark.parametrize("module_name,reason", sorted(EXCLUDED_FROM_GROUPS.items()))
def test_excluded_modules_stay_out_of_action_groups(action_groups, module_name, reason):
assert module_name in _module_names(), (
"%s is listed in EXCLUDED_FROM_GROUPS but no longer exists; drop the entry."
% module_name
)
for group, members in action_groups.items():
assert module_name not in members, (
"%s was added to the '%s' action group, but it %s. If it now accepts the "
"group's shared authentication options, remove it from EXCLUDED_FROM_GROUPS "
"in this test as well." % (module_name, group, reason)
)
assert not _shared_specs_used_by(module_name), (
"%s now builds its argument spec from a shared auth spec, so the reason it was "
"excluded (%s) no longer holds. Add it to the matching action group and remove "
"it from EXCLUDED_FROM_GROUPS." % (module_name, reason)
)