Package service integration tests fail on non-amd64 architectures
Metadata
Current evaluation
Package service integration tests hardcode amd64, causing failures on arm64 CI. Still open; QA workflow limits tests to amd64 as a workaround referencing this issue. No fix merged yet.
Suggested action: keep open
Reason: The QA workflow (.github/workflows/qa.yaml) explicitly references this issue as the reason for limiting tests to amd64: 'Limiting to amd64 is a workaround for https://github.com/canonical/charmcraft/issues/2018. Once that's resolved we should run fast and slow tests on ARM64 Ubuntu too.' This is a maintainer acknowledgement that the issue is valid and still relevant, and the failing tests in tests/integration/services/test_package.py still exist. It is a genuine test-infrastructure bug, not a support request, and remains unresolved.
Impact:
35
Quick Win:
21.0
Staleness:
55
Complexity:
40
Confidence:
78
Support Request:
5
Issue body
### Bug Description
We have 'amd64' hardcoded in some of the package service integration tests that cause them to fail on other architectures.
### To Reproduce
https://github.com/canonical/charmcraft/actions/runs/12205703353/job/34053604513?pr=2015
### Environment
Canonical hosted GH CI, arm64
### charmcraft.yaml
```yaml
n/a
```
### Relevant log output
```shell
=========================== short test summary info ============================
FAILED tests/integration/commands/test_pack.py::test_build_basic_charm[platforms-jammy-charm] - IndexError: list index out of range
FAILED tests/integration/services/test_package.py::test_write_metadata[complex-legacy] - RuntimeError: Could not determine run-on bases.
FAILED tests/integration/services/test_package.py::test_write_metadata[basic] - check charmcraft-version: 3.0-test-version
charmcraft-started-at: '2020-03-14T00:00:00+00:00'
bases:
- name: ubuntu
channel: '22.04'
architectures:
- arm64
analysis:
attributes:
- name: language
result: unknown
- name: framework
result: unknown
== charmcraft-version: 3.0-test-version
charmcraft-started-at: '2020-03-14T00:00:00+00:00'
bases:
- name: ubuntu
channel: '22.04'
architectures:
- amd64
analysis:
attributes:
- name: language
result: unknown
- name: framework
result: unknown
test_package.py:69 in test_write_metadata() -> pytest_check.equal((test_prime_dir / file.name).read_text(), file.read_text())
------------------------------------------------------------
Failed Checks: 1
FAILED tests/integration/services/test_package.py::test_write_metadata[complex-self-contained] - RuntimeError: Could not determine run-on bases.
FAILED tests/integration/services/test_package.py::test_write_metadata[basic-reactive] - check charmcraft-version: 3.0-test-version
charmcraft-started-at: '2020-03-14T00:00:00+00:00'
bases:
- name: ubuntu
channel: '22.04'
architectures:
- arm64
analysis:
attributes:
- name: language
result: unknown
- name: framework
result: unknown
== charmcraft-version: 3.0-test-version
charmcraft-started-at: '2020-03-14T00:00:00+00:00'
bases:
- name: ubuntu
channel: '22.04'
architectures:
- amd64
analysis:
attributes:
- name: language
result: unknown
- name: framework
result: unknown
test_package.py:69 in test_write_metadata() -> pytest_check.equal((test_prime_dir / file.name).read_text(), file.read_text())
------------------------------------------------------------
Failed Checks: 1
FAILED tests/integration/services/test_package.py::test_overwrite_metadata[complex-legacy] - RuntimeError: Could not determine run-on bases.
FAILED tests/integration/services/test_package.py::test_overwrite_metadata[basic] - check charmcraft-version: 3.0-test-version
charmcraft-started-at: '2020-03-14T00:00:00+00:00'
bases:
- name: ubuntu
channel: '22.04'
architectures:
- arm64
analysis:
attributes:
- name: language
result: unknown
- name: framework
result: unknown
== charmcraft-version: 3.0-test-version
charmcraft-started-at: '2020-03-14T00:00:00+00:00'
bases:
- name: ubuntu
channel: '22.04'
architectures:
- amd64
analysis:
attributes:
- name: language
result: unknown
- name: framework
result: unknown
test_package.py:103 in test_overwrite_metadata() -> pytest_check.equal((test_prime_dir / file.name).read_text(), file.read_text())
------------------------------------------------------------
Failed Checks: 1
FAILED tests/integration/services/test_package.py::test_overwrite_metadata[complex-self-contained] - RuntimeError: Could not determine run-on bases.
FAILED tests/integration/services/test_package.py::test_overwrite_metadata[basic-reactive] - check charmcraft-version: 3.0-test-version
charmcraft-started-at: '2020-03-14T00:00:00+00:00'
bases:
- name: ubuntu
channel: '22.04'
architectures:
- arm64
analysis:
attributes:
- name: language
result: unknown
- name: framework
result: unknown
== charmcraft-version: 3.0-test-version
charmcraft-started-at: '2020-03-14T00:00:00+00:00'
bases:
- name: ubuntu
channel: '22.04'
architectures:
- amd64
analysis:
attributes:
- name: language
result: unknown
- name: framework
result: unknown
test_package.py:103 in test_overwrite_metadata() -> pytest_check.equal((test_prime_dir / file.name).read_text(), file.read_text())
------------------------------------------------------------
Failed Checks: 1
= 9 failed, 2398 passed, 1 skipped, 20 deselected, 1 xfailed, 2078 warnings in 54.37s =
```
Evaluation history
| Date | Model | Scores | Action | Summary |
|---|---|---|---|---|
| qwen/qwen3.8-27b |
Impact:
35
Quick Win:
21.0
Staleness:
55
Complexity:
40
Confidence:
78
Support Request:
5
|
keep open | Package service integration tests hardcode amd64, causing failures on arm64 CI. Still open; QA workflow limits tests to amd64 as a workaround referencing this issue. No fix merged yet. | |
| qwen/qwen3.6-35b-a3b |
Staleness:
90
Complexity:
30
Confidence:
85
Support Request:
5
|
needs triage | Integration tests fail on non-amd64 architectures due to hardcoded amd64 values. Open for over a year with no maintainer response, labels, or assignee. | |
| qwen3.6-35b-a3b-mtp-q6 |
Staleness:
85
Complexity:
25
Confidence:
80
Support Request:
5
|
needs triage | Integration tests for the package service fail on non-amd64 architectures due to hardcoded amd64 references. Open for over a year with no maintainer triage or follow-up beyond an automated Jira sync. | |
| qwen3.6-35b-a3b-mtp-q6 |
Staleness:
90
Complexity:
30
Confidence:
80
Support Request:
5
|
needs triage | Package service integration tests fail on non-amd64 architectures due to hardcoded amd64 references. Open for over a year with no maintainer labels or direct responses, only a Jira sync bot comment. | |
| qwen3.6-35b-a3b-mtp-q6 |
Staleness:
85
Complexity:
20
Confidence:
85
Support Request:
5
|
needs triage | Package service integration tests fail on non-amd64 architectures due to hardcoded amd64 expectations. Open, unassigned, unlabelled, 304 days inactive. Requires updating test fixtures to dynamically detect architectures for multi-arch CI. |
Update history
No update history recorded yet.
Related work
-
Related To:
canonical/charmcraft#2151
(confidence 55%)
A related non-amd64 fix (bash completion path) was merged, showing maintainers address arch-specific issues, but it does not fix this test hardcoding.
Related issues
| Issue | Project | State | Summary | Similarity |
|---|---|---|---|---|
| #144 Some tests fail on arm64 | craft-archives | closed | Closed without resolution. Tests fail on arm64 due to hard-coded amd64 architecture in expected values. No fix or merge was provided, leaving the issue abandoned. | |
| #1255 Make tests platform-agnostic | craft-parts | open | Tests have hardcoded amd64 strings preventing execution on other architectures. CI runners for arm/s390x/ppc64el have been added, but 45 hardcoded amd64 references remain in test files. Triaged, no recent activity in ~12 months. |