podman/tests/test_podman.yml
DistroBaker d55c8d8f5a Merged update from upstream sources
This is an automated DistroBaker update from upstream sources.
If you do not know what this is about or would like to opt out,
contact the OSCI team.

Source: https://src.fedoraproject.org/rpms/podman.git#3f2f50cda9c6dd6e86824f7b944944c20946fd6f
2020-11-05 22:53:01 +00:00

45 lines
1.5 KiB
YAML

---
- hosts: localhost
tags:
- classic
- container
vars:
- artifacts: ./artifacts
rootless_user: testuser
roles:
- role: rootless_user_ready
tasks:
# At the start of a run, clean up state. Useful for test reruns.
- name: local artifacts directory exists
local_action: file path="{{ artifacts }}" state=directory
- name: remove stale log files
local_action: shell rm -f {{ artifacts }}/test*.log
- name: clear test results (test.log)
local_action: command truncate --size=0 {{ artifacts }}/test.log
- name: clear test results (results.yml)
local_action: copy content="results:\n" dest={{ artifacts }}/results.yml
# These are the actual tests: set cgroups vN, then run root/rootless tests.
#
# FIXME FIXME FIXME: 2020-05-21: 'loop' should be '2, 1' but there's some
# nightmarish bug in CI wherein reboots hang forever. There's a bug open[1]
# but it seems dead. Without a working reboot, there's no way to test v1.
# [1] https://redhat.service-now.com/surl.do?n=PNT0808530
# I'm leaving this as a 'loop' in (foolish? vain?) hope that the bug will
# be fixed. Let's revisit this after, say, 2020-08. If the bug persists
# then let's just revert the entire cgroups v1 change, and go back to
# using standard-test-basic.
- name: set cgroups and run podman tests
include_tasks: test_podman_cgroups_vn.yml
loop: [ 2 ]
loop_control:
loop_var: want_cgroups
# Postprocessing: check for FAIL or ERROR in any test, exit 1 if so
- name: check results
include_tasks: check_results.yml