summaryrefslogtreecommitdiff
path: root/.cirrus.yml
diff options
context:
space:
mode:
authorChris Evich <cevich@redhat.com>2022-04-11 15:30:15 -0400
committerChris Evich <cevich@redhat.com>2022-04-11 15:30:15 -0400
commit61cf2228826aab06c3179c83ebc96b58996b133f (patch)
tree552d64d56149ed2b4467796a49cd0b50ca0ee8a9 /.cirrus.yml
parent1d01815c107c91f6cfe98446d334c94a97d11080 (diff)
downloadpodman-61cf2228826aab06c3179c83ebc96b58996b133f.tar.gz
podman-61cf2228826aab06c3179c83ebc96b58996b133f.tar.bz2
podman-61cf2228826aab06c3179c83ebc96b58996b133f.zip
Cirrus: Fix unsupported cirrus-cron build status
Every weekday when the `check_cirrus_cron` github-actions workflow runs. It checks the status of all cirrus-cron jobs. If a build is found with a 'FAILED' status, it triggers an alert e-mail to be sent. However, the `test_image_build` is marked as a manually-triggered, resulting in a perpetual status of 'EXECUTING', even if there were failures. Fix this by only allowing the problematic task to run in pull requests without the `[CI:DOCS]` magic keyword. Signed-off-by: Chris Evich <cevich@redhat.com>
Diffstat (limited to '.cirrus.yml')
-rw-r--r--.cirrus.yml4
1 files changed, 2 insertions, 2 deletions
diff --git a/.cirrus.yml b/.cirrus.yml
index 772843dd7..44c5d425d 100644
--- a/.cirrus.yml
+++ b/.cirrus.yml
@@ -776,8 +776,8 @@ image_build_task: &image-build
test_image_build_task:
<<: *image-build
- # Allow this to run inside a PR
- only_if: $CI == $CI
+ # Allow this to run inside a PR w/ [CI:BUILD]
+ only_if: $CIRRUS_PR != '' && $CIRRUS_CHANGE_TITLE !=~ '.*CI:DOCS.*'
# This takes a LONG time, only run when requested. N/B: Any task
# made to depend on this one will block FOREVER unless triggered.
trigger_type: manual