This document is for a development version of Ceph.

Integration Tests using Teuthology Workflow

Scheduling Test Run

Getting binaries

Ceph binaries must be built for your branch before you can use teuthology to run integration tests on them. Follow these steps to build the Ceph binaries:

  1. Push the branch to the ceph-ci repository. This triggers the process of building the binaries on the Jenkins CI.

  2. To ensure that the build process has been initiated, confirm that the branch name has appeared in the list of “Latest Builds Available” at Shaman. Soon after you start the build process, the testing infrastructure adds other, similarly-named builds to the list of “Latest Builds Available”. The names of these new builds will contain the names of various Linux distributions of Linux and will be used to test your build against those Linux distributions.

  3. Wait for the packages to be built and uploaded to Chacra, and wait for the repositories offering the packages to be created. The entries for the branch names in the list of “Latest Builds Available” on Shaman will turn green to indicate that the packages have been uploaded to Chacra and to indicate that their repositories have been created. Wait until each entry is coloured green. This usually takes between two and three hours depending on the availability of the machines.

    The Chacra URL for a particular build can be queried from the Chacra site.


The branch to be pushed on ceph-ci can be any branch. The branch does not have to be a PR branch.


If you intend to push master or any other standard branch, check Shaman beforehand since it might already have completed builds for it.

Triggering Tests

After you have built Ceph binaries for your branch, you can run tests using teuthology. This procedure explains how to run tests using teuthology.

  1. Log in to the teuthology machine:

    ssh <username>

    This requires Sepia lab access. To request access to the Sepia lab, see:

  2. Run the teuthology-suite command:

    The options in the above command are defined here:






    machine name


    the name of the branch that was pushed on ceph-ci


    test-suite name


    the higher the number, the lower the priority of the job


    filter tests in a given suite. The argument passed to this filter specifies which test you want to run

    -e <email>

    When tests finish or time out, send an email to the specified address. Can also be specified in ~/.teuthology.yaml as ‘results_email’


    The priority number present in the command above is a placeholder. Do not use it in your own tests. See Testing Priority for information about recommended values.


    Do not issue a command without a priority number. The default value is 1000, a value so large that your job is unlikely ever to run.

    Run teuthology-suite --help to read descriptions of these and other available options.

  3. Wait for the tests to run. teuthology-suite prints a link to Pulpito where the test results can be viewed.

Other frequently used/useful options are -d (or --distro), --distroversion, --filter-out, --timeout, flavor, -rerun, -l (for limiting number of jobs) , -N (for how many times the job will run), and --subset (used to reduce the number of tests that are triggered). Run teuthology-suite --help to read descriptions of these and other options.

Testing QA changes (without re-building binaries)

If you are making changes only in the qa/ directory, you do not have to rebuild the binaries before you re-run tests. If you make changes only in qa/, you can use the binaries built for the ceph-ci branch to re-run tests. You just have to make sure to tell the teuthology-suite command to use a separate branch for running the tests.

If you made changes only in qa/ (, you do not need to rebuild the binaries. You can use existing binaries that are built periodically for master and other stable branches and run your test changes against them. Your branch with the qa changes can be tested by passing two extra arguments to the teuthology-suite command: (1) --suite-repo, specifying your ceph repo, and (2) --suite-branch, specifying your branch name.

For example, if you want to make changes in qa/ after testing branch-x (for which the ceph-ci branch is wip-username-branch-x), run the following command

teuthology-suite -v \
 -m smithi \
 -c wip-username-branch-x \
 -s fs \
 -p 50 \
 --filter cephfs-shell

Then make modifications locally, update the PR branch, and trigger tests from your PR branch as follows:

teuthology-suite -v \
 -m smithi \
 -c wip-username-branch-x \
 -s fs -p 50 \
 --filter cephfs-shell \
 --suite-repo$username/ceph \
 --suite-branch branch-x

You can verify that the tests were run using this branch by looking at the values for the keys suite_branch, suite_repo and suite_sha1 in the job config printed at the beginning of the teuthology job.


If you are making changes that are not in the qa/ directory, you must follow the standard process of triggering builds, waiting for the builds to finish, then triggering tests and waiting for the test results.

About Suites and Filters

See Suites Inventory for a list of available suites of integration tests. Each directory under qa/suites in the Ceph repository is an integration test suite, and arguments appropriate to follow -s can be found there.

Keywords for filtering tests can be found in qa/suites/<suite-name>/<subsuite-name>/tasks and can be used as arguments for --filter. Each YAML file in that directory can trigger tests; using the name of the file without its filename extension as an argument to the --filter triggers those tests.

For example, in the command above in the Testing QA Changes section, cephfs-shell is specified. This works because there is a file named cephfs-shell.yaml in qa/suites/fs/basic_functional/tasks/.

If the filename doesn’t suggest what kind of tests it triggers, search the contents of the file for the modules attribute. For cephfs-shell.yaml the modules attribute is tasks.cephfs.test_cephfs_shell. This means that it triggers all tests in qa/tasks/cephfs/

Viewing Test Results

Pulpito Dashboard

After the teuthology job is scheduled, the status and results of the test run can be checked at

Teuthology Archives

After the tests have finished running, the log for the job can be obtained by clicking on the job ID at the Pulpito page associated with your tests. It’s more convenient to download the log and then view it rather than viewing it in an internet browser since these logs can easily be up to 1 GB in size. It is easier to ssh into the teuthology machine ( and access the following path:


For example: for the above test ID, the path is:


This method can be used to view the log more quickly than would be possible through a browser.


To access archives more conveniently, /a/ has been symbolically linked to /ceph/teuthology-archive/. For instance, to access the previous example, we can use something like:


Killing Tests

teuthology-kill can be used to kill jobs that have been running unexpectedly for several hours, or when developers want to terminate tests before they complete.

Here is the command that terminates jobs:

teuthology-kill -r teuthology-2019-12-10_05:00:03-smoke-master-testing-basic-smithi

Let’s call the argument passed to -r as test ID. It can be found easily in the link to the Pulpito page for the tests you triggered. For example, for the above test ID, the link is -

Re-running Tests

The teuthology-suite command has a -r (or --rerun) option, which allows you to re-run tests. This is handy when your tests have failed or end up dead. The --rerun option takes the name of a teuthology run as an argument. Option -R (or --rerun-statuses) can be passed along with -r to choose which kind of tests should be picked from the run. For example, you can re-run only those tests from previous run which had ended up as dead. Following is a practical example:

teuthology-suite -v \
 -m smithi \
 -c wip-rishabh-fs-test_cephfs_shell-fix \
 -p 50 \
 --r teuthology-2019-12-10_05:00:03-smoke-master-testing-basic-smithi \
 -R fail,dead,queued \

Following’s the definition of new options introduced in this section:



-r, --rerun

Attempt to reschedule a run, selecting only those jobs whose status are mentioned by --rerun-status.

-R, --rerun-statuses

A comma-separated list of statuses to be used with --rerun. Supported statuses: ‘dead’, ‘fail’, ‘pass’, ‘queued’, ‘running’ and ‘waiting’. Default value: ‘fail,dead’

Naming the ceph-ci branch

Prepend your branch with your name before you push it to ceph-ci. For example, a branch named feature-x should be named wip-$yourname-feature-x, where $yourname is replaced with your name. Identifying your branch with your name makes your branch easily findable on Shaman and Pulpito.

If you are using one of the stable branches (quincy, pacific, etc.), include the name of that stable branch in your ceph-ci branch name. For example, the feature-x PR branch should be named wip-feature-x-nautilus. This is not just a convention. This ensures that your branch is built in the correct environment.

Delete the branch from ceph-ci when you no longer need it. If you are logged in to GitHub, all your branches on ceph-ci can be found here: