PLAYBOOK软件测试自学,帮助测试PR — 国内最专业的Ansible中文官方学习手册

For Those About To Test, We Salute You¶

At this point, you should be ready to begin testing!

If the PR is a bug-fix pull request, the first things to do are to run the suite of unit and integration tests, to ensure

the pull request does not break current functionality:

# Unit Tests

make tests

# Integration Tests

cd test/integration

make

Note

Ansible does provide integration tests for cloud-based modules as well, however we do not recommend using them for some users

due to the associated costs from the cloud providers. As such, typically it’s better to run specific parts of the integration battery

and skip these tests.

Integration tests aren’t the end all beat all - in many cases what is fixed might not HAVE a test, so determining if it works means

checking the functionality of the system and making sure it does what it said it would do.

Pull requests for bug-fixes should reference the bug issue number they are fixing.

We encourage users to provide playbook examples for bugs that show how to reproduce the error, and these playbooks should be used to verify the bugfix does resolve

the issue if available. You may wish to also do your own review to poke the corners of the change.

Since some reproducers can be quite involved, you might wish to create a testing directory with the issue # as a sub-

directory to keep things organized:

mkdir -p testing/XXXX # where XXXX is again the issue # for the original issue or PR

cd testing/XXXX

While it should go without saying, be sure to read any playbooks before you run them. VMs help with running untrusted content greatly,

though a playbook could still do something to your computing resources that you’d rather not like.

Once the files are in place, you can run the provided playbook (if there is one) to test the functionality:

ansible-playbook -vvv playbook_name.yml

If there’s not a playbook, you may have to copy and paste playbook snippets or run a ad-hoc command that was pasted in.

Our issue template also included sections for “Expected Output” and “Actual Output”, which should be used to gauge the output

from the provided examples.

If the pull request resolves the issue, please leave a comment on the pull request, showing the following information:

“Works for me!”

The output from ansible –version.

In some cases, you may wish to share playbook output from the test run as well.

Example!:

Works for me! Tested on `Ansible 1.7.1`. I verified this on CentOS 6.5 and also Ubuntu 14.04.

If the PR does not resolve the issue, or if you see any failures from the unit/integration tests, just include that output instead:

This doesn't work for me.

When I ran this my toaster started making loud noises!

Output from the toaster looked like this:

```

BLARG

StrackTrace

RRRARRGGG

```

When you are done testing a feature branch, you can remove it with the following command:

git branch -D someuser-feature_branch_name

We understand some users may be inexperienced with git, or other aspects of the above procedure, so feel free to stop by ansible-devel

list for questions and we’d be happy to help answer them.