Pipelines and Jobs

Project pipelines

A pipeline is a group of jobs executed by GitLab CI.

Examples

List pipelines for a project:

pipelines = project.pipelines.list()

Get a pipeline for a project:

pipeline = project.pipelines.get(pipeline_id)

Get variables of a pipeline:

variables = pipeline.variables.list()

Create a pipeline for a particular reference:

pipeline = project.pipelines.create({'ref': 'master'})

Retry the failed builds for a pipeline:

pipeline.retry()

Cancel builds in a pipeline:

pipeline.cancel()

Delete a pipeline:

pipeline.delete()

Triggers

Triggers provide a way to interact with the GitLab CI. Using a trigger a user or an application can run a new build/job for a specific commit.

Examples

List triggers:

triggers = project.triggers.list()

Get a trigger:

trigger = project.triggers.get(trigger_token)

Create a trigger:

trigger = project.triggers.create({'description': 'mytrigger'})

Remove a trigger:

project.triggers.delete(trigger_token)
# or
trigger.delete()

Full example with wait for finish:

def get_or_create_trigger(project):
    trigger_decription = 'my_trigger_id'
    for t in project.triggers.list():
        if t.description == trigger_decription:
            return t
    return project.triggers.create({'description': trigger_decription})

trigger = get_or_create_trigger(project)
pipeline = project.trigger_pipeline('master', trigger.token, variables={"DEPLOY_ZONE": "us-west1"})
while pipeline.finished_at is None:
    pipeline.refresh()
    time.sleep(1)

You can trigger a pipeline using token authentication instead of user authentication. To do so create an anonymous Gitlab instance and use lazy objects to get the associated project:

gl = gitlab.Gitlab(URL)  # no authentication
project = gl.projects.get(project_id, lazy=True)  # no API call
project.trigger_pipeline('master', trigger_token)

Reference: https://docs.gitlab.com/ee/ci/triggers/#trigger-token

Pipeline schedule

You can schedule pipeline runs using a cron-like syntax. Variables can be associated with the scheduled pipelines.

Examples

List pipeline schedules:

scheds = project.pipelineschedules.list()

Get a single schedule:

sched = projects.pipelineschedules.get(schedule_id)

Create a new schedule:

sched = project.pipelineschedules.create({
    'ref': 'master',
    'description': 'Daily test',
    'cron': '0 1 * * *'})

Update a schedule:

sched.cron = '1 2 * * *'
sched.save()

Delete a schedule:

sched.delete()

List schedule variables:

# note: you need to use get() to retrieve the schedule variables. The
# attribute is not present in the response of a list() call
sched = projects.pipelineschedules.get(schedule_id)
vars = sched.attributes['variables']

Create a schedule variable:

var = sched.variables.create({'key': 'foo', 'value': 'bar'})

Edit a schedule variable:

var.value = 'new_value'
var.save()

Delete a schedule variable:

var.delete()

Projects and groups variables

You can associate variables to projects and groups to modify the build/job scripts behavior.

Examples

List variables:

p_variables = project.variables.list()
g_variables = group.variables.list()

Get a variable:

p_var = project.variables.get('key_name')
g_var = group.variables.get('key_name')

Create a variable:

var = project.variables.create({'key': 'key1', 'value': 'value1'})
var = group.variables.create({'key': 'key1', 'value': 'value1'})

Update a variable value:

var.value = 'new_value'
var.save()

Remove a variable:

project.variables.delete('key_name')
group.variables.delete('key_name')
# or
var.delete()

Jobs

Jobs are associated to projects, pipelines and commits. They provide information on the jobs that have been run, and methods to manipulate them.

Examples

Jobs are usually automatically triggered, but you can explicitly trigger a new job:

project.trigger_build('master', trigger_token,
                      {'extra_var1': 'foo', 'extra_var2': 'bar'})

List jobs for the project:

jobs = project.jobs.list()

Get a single job:

project.jobs.get(job_id)

List the jobs of a pipeline:

project = gl.projects.get(project_id)
pipeline = project.pipelines.get(pipeline_id)
jobs = pipeline.jobs.list()

Note

Job methods (play, cancel, and so on) are not available on ProjectPipelineJob objects. To use these methods create a ProjectJob object:

pipeline_job = pipeline.jobs.list()[0]
job = project.jobs.get(pipeline_job.id, lazy=True)
job.retry()

Get the artifacts of a job:

build_or_job.artifacts()

Warning

Artifacts are entirely stored in memory in this example.

You can download artifacts as a stream. Provide a callable to handle the stream:

class Foo(object):
    def __init__(self):
        self._fd = open('artifacts.zip', 'wb')

    def __call__(self, chunk):
        self._fd.write(chunk)

target = Foo()
build_or_job.artifacts(streamed=True, action=target)
del(target)  # flushes data on disk

You can also directly stream the output into a file, and unzip it afterwards:

zipfn = "___artifacts.zip"
with open(zipfn, "wb") as f:
    build_or_job.artifacts(streamed=True, action=f.write)
subprocess.run(["unzip", "-bo", zipfn])
os.unlink(zipfn)

Get a single artifact file:

build_or_job.artifact('path/to/file')

Get a single artifact file by branch and job:

project.artifact('branch', 'path/to/file', 'job')

Mark a job artifact as kept when expiration is set:

build_or_job.keep_artifacts()

Delete the artifacts of a job:

build_or_job.delete_artifacts()

Get a job trace:

build_or_job.trace()

Warning

Traces are entirely stored in memory unless you use the streaming feature. See the artifacts example.

Cancel/retry a job:

build_or_job.cancel()
build_or_job.retry()

Play (trigger) a job:

build_or_job.play()

Erase a job (artifacts and trace):

build_or_job.erase()