From 5369e45470be371ab89e8b1e905ca9aa29e387ce Mon Sep 17 00:00:00 2001 From: Chris Evich <cevich@redhat.com> Date: Thu, 30 May 2019 12:55:51 -0400 Subject: docs: Add CI section and links Signed-off-by: Chris Evich <cevich@redhat.com> --- CONTRIBUTING.md | 43 +++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 43 insertions(+) diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md index 3778d6d7d..87efdeb8f 100644 --- a/CONTRIBUTING.md +++ b/CONTRIBUTING.md @@ -8,6 +8,7 @@ that we follow. * [Reporting Issues](#reporting-issues) * [Contributing to libpod](#contributing-to-libpod) +* [Continuous Integration](#continuous-integration) [](https://cirrus-ci.com/github/containers/libpod/master) * [Submitting Pull Requests](#submitting-pull-requests) * [Communications](#communications) @@ -296,6 +297,48 @@ be provided with PRs. For details on how to run the tests for Podman in your test environment, see the Integration Tests [README.md](test/README.md). +## Continuous Integration + +All pull requests and branch-merges automatically run: + +* Go format/lint checking +* Unit testing +* Integration Testing +* Special testing (like running inside a container, or as a regular user) + +For a more in-depth reference of the CI system, please [refer to it's dedicated +documentation.](contrib/cirrus/README.md) + +There is always additional complexity added by automation, and so it sometimes +can fail for any number of reasons. This includes post-merge testing on all +branches, which you may occasionally see [red bars on the status graph +.](https://cirrus-ci.com/github/containers/libpod/master) + +When the graph shows mostly green bars on the right, it's a good indication +the master branch is currently stable. Alternating red/green bars is indicitave +of a testing "flake", and should be examined (anybody can do this): + +* *One or a small handful of tests, on a single task, (i.e. specific distro/version) + where all others ran successfully:* Frequently the cause is networking or a brief + external service outage. The failed tasks may simply be re-run by pressing the + corresponding button on the task details page. + +* *Multiple tasks failing*: Logically this should be due to some shared/common element. + If that element is identifiable as a networking or external service (e.g. packaging + repository outage), a re-run should be attempted. + +* *All tasks are failing*: If a common element is **not** identifiable as + temporary (i.e. container registry outage), please seek assistance via + [the methods below](#communications) as this may be early indication of + a more serious problem. + +In the (hopefully) rare case there are multiple, contiguous red bars, this is +a ***very bad*** sign. It means additional merges are occurring despite an uncorrected +or persistently faulty condition. This risks additional bugs being introduced +and further complication of necessary corrective measures. Most likely people +are aware and working on this, but it doesn't hurt [to confirm and/or try and help +if possible.](#communications) + ## Communications For general questions and discussion, please use the -- cgit v1.2.3-54-g00ecf