zulip/docs/testing/continuous-integration.md

Continuous integration (CI)

The Zulip server uses CircleCI and Travis CI for continuous integration. CircleCI is the primary CI, and runs frontend and backend tests across a wide range of Ubuntu distributions. Travis CI is legacy, used only for running the end-to-end production installer test. This page documents useful tools and tips to know about when using CircleCI and Travis CI and debugging issues with them.

Goals

The overall goal of our CI is to avoid regressions and minimize the total time spent debugging Zulip. We do that by trying to catch as many possible future bugs as possible, while minimizing both latency and false positives, both of which can waste a lot of developer time. There are a few implications of this overall goal:

• If a test is failing nondeterministically in CI, we consider that to be an urgent problem.
• If the tests become a lot slower, that is also an urgent problem.
• Everything we do in CI should also have a way to run it quickly (under 1 minute, preferably under 3 seconds), in order to iterate fast in development. Except when working on the CI configuration itself, a developer should never have to repeatedly wait 10 minutes for a full CI run to iteratively debug something.

CircleCI

Useful debugging tips and tools

• Zulip uses the ts tool to log the current time on every line of the output in our Travis CI scripts. You can use this output to determine which steps are actually consuming a lot of time.

• You can sign up your personal repo for CircleCI so that every remote branch you push will be tested, which can be helpful when debugging something complicated.

• With your personal repo signed up, CircleCI allows you to SSH into the job container if a job fails. SSHing into the containers can be helpful, especially in rare cases where the tests are passing in your computer but failing in the CI. Make sure that you have uploaded your SSH keys to GitHub: CircleCI uses those SSH keys for authentication.

Suites

The main CircleCI configuration file is ./circleci/config.yml. We currently run several jobs during a CircleCI build. They are:

• trusty-python-3.4
• xenial-python-3.5
• bionic-python-3.6

Each runs the Zulip backend test suites, using the indicated platform/OS and Python version. bionic-python-3.6 job for example runs the tests in Ubuntu Xenial with Python 3.6 pre-installed. Additionally, the trusty suite also runs the Zulip frontend test suites; since those are not platform-dependent, it doesn't make sense to run them on all platforms. Your build for the PR will pass only if all the 3 jobs are executed successfully.

Configuration

The remaining details in this section are primarily relevant for doing development on our CI system and/or provisioning process.

The first key of the job section is docker. The docker key specifies the image CircleCI should get from Docker Hub for running the job. Once CircleCI fetches the image from Docker Hub, it will spin up a docker container. See images section to know more about the images we use in CircleCI for testing.

After booting the container from the configured image, CircleCI will create the directory mentioned in working_directory and all the steps are be run from here.

The steps section describes describes everything: fetching the Zulip code, provisioning, fetching catched data, running tests and uploading coverage reports. The steps with prefix * reference aliases, which are defined in the aliases section at the top of the file.

Images

CircleCI tests are run in containers that are spun off from the images maintained by Zulip team. The Dockerfiles for the various images can be generated by running ./tools/circleci/generate-dockerfiles. This command will generate the Dockerfiles of the three Ubuntu releases in ./tools/circleci/images/{release_name} directories. Take a look at ./tools/circleci/images.yml to see how the Dockerfiles for the three releases differ from each other. To further generate images from the Dockerfiles and upload it to Docker Hub follow the instructions in the generated Dockerfiles.

Performance optimizations

Caching

An important element of making CircleCI perform effectively is caching the provisioning of a Zulip development environment. In particular, we cache the following.:

• Python virtualenvs
• node_modules directories

This has a huge impact on the performance of running tests in CircleCI CI; without these caches, the average test time would be several times longer.

We have designed these caches carefully (they are also used in production and the Zulip development environment) to ensure that each is named by a hash of its dependencies and ubuntu distribution name, so Zulip should always be using the same version of dependencies it would have used had the cache not existed. In practice, bugs are always possible, so be mindful of this possibility.

A consequence of this caching is that test jobs for branches which modify package.json, requirements/, and other key dependencies will be significantly slower than normal, because they won't get to benefit from the cache.

Travis CI

Configuration

The main Travis configuration file is .travis.yml. The specific test suites we have are listed in the matrix section, which has a matrix of Python versions and test suites ($TEST_SUITE).

Currently there is only the production test suite in this section as we have moved the backend and frontend suite to CircleCI. So the value of the variable $TEST_SUITE would be always production.

We've configured it to use a few helper scripts for each job:

• tools/ci/setup-$TEST_SUITE: This script sets up the test environment for the production suite. This is a complicated process because of all the packages Travis installs. See the comments in tools/ci/setup-production for details.
• tools/ci/$TEST_SUITE: The script that runs the actual test production test suite.

The main purpose of the distinction between the two is that if the setup-production job fails, Travis CI will report it as the suite having "Errored" (grey in their emails), whereas if the production job fails, it'll be reported as "Failed" failure (red in their emails). Note that Travis CI's web UI seems to make no visual distinction between these.

An important detail is that Travis CI will by default hide most phases other than the actual test; you can see this easily by looking at the line numbers in the Travis CI output. There are actually a bunch of phases (e.g. the project's setup job, downloading caches near the beginning, uploading caches at the end, etc.), and if you're debugging our configuration, you'll want to look at these closely.

Useful debugging tips and tools

• Zulip uses the ts tool to log the current time on every line of the output in our Travis CI scripts. You can use this output to determine which steps are actually consuming a lot of time.

• For performance issues, this statistics tool can give you test runtime history data that can help with determining when a performance issue was introduced and whether it was fixed. Note you need to click the "Run" button for it to do anything.

• You can sign up your personal repo for Travis CI so that every remote branch you push will be tested, which can be helpful when debugging something complicated.

Performance optimizations

Caching

We cache the following as well apart from what is mentioned in CircleCI caching section.

• Built/downloaded emoji sprite sheets and data.

This is probably worth eventually adding to the CircleCI caches, but because it only saves ~5s, it hasn't been a priority yet.

Uninstalling packages

In the production suite, we run apt-get upgrade at some point (effectively, because the Zulip installer does). This carries a huge performance cost in Travis CI, because (1) they don't keep their test systems up to date and (2) literally everything is installed in their build workers (e.g. several copies of Postgres, Java, MySQL, etc.).

In order to make Zulip's tests performance reasonably well, we uninstall (or mark with apt-mark hold) many of these dependencies that are irrelevant to Zulip in tools/ci/setup-production.