In the old configuration we had 2 jobs, test and build.
The build ran after test had run successfully, but we could manually trigger build if we want to skip the tests.
After we switched to multiple pipeline using Jenkinsfile, we had to put those 2 build jobs in to the same file:
stage('Running tests'){
...
}
stage('Build'){
...
}
So now the build step is only triggered after running tests successfully, and we cannot manually trigger build, without commenting out the test steps and commit to the repository.
I am wondering if there is a better approach/practise to utilise the Jenkinsfile to overcome this limitation?
Using pipeline and Jenkinsfile is becoming the standard and preferred way of running jobs on Jenkins now a days. So using a Jenkinsfile is certainly the way to go.
One way to solve the problem is to make the job parameterized:
// Set the parameter properties, this will be done at the first run so that we can trigger with parameters manually
properties([parameters([booleanParam(defaultValue: true, description: 'Testing will be done if this is checked', name: 'DO_TEST')])])
stage('Running tests'){
// Putting the check inside of the stage step so that we don't confuse the stage view
if (params['DO_TEST']) {
...
}
}
stage('Build'){
...
}
The first time the job runs, it will add a parameter to the job. After that we can trigger manually and select whether tests should run. The default value will be used when it's triggered by SCM.
Related
I have not found information about that, I want to trigger build but before it executes, I want it to trigger some other pipelines and wait for a file created by these other pipelines to pass to a main Pipeline, can I do this in Jenkins?
You can, yes, in multiple ways depending on your actual use case.
The simplest way would be to create the job you want to call and then add a step for calling that job, copy artifacts from the job and then continue with the pipeline.
Using Jenkins Build step:
stage ('Child job') {
steps {
build(job: 'foo', wait: true, propagate: true, parameters: [parameters_list])
}
}
The parameter wait makes it so your pipeline waits for the child to complete run before continuing and propagate means that the result of the job is shown in the parent pipeline as well. The parameters section is needed only if your child job requires parameters. See the parameter types in the Build step documentation.
You can also use the Jenkins pipeline snippet generator to create the call properly in your own Jenkins instance.
To get any build artifacts from the child job, the easiest way to go is to use the Copy Artifact plugin. You'll need to archiveArtifacts in the first job and then
copyArtifacts fingerprintArtifacts: true, projectName: 'foo', selector: lastCompleted()
When the artifacts are needed.
I am trying to setup various Jenkins pipelines whose last stage is always to run some acceptance tests. To cut a long story short, acceptance tests and test data (much of which is shared) for all products are checked into the same repository which is about 0.5 GB in size. It therefore seemed best to have a separate job for the acceptance tests and trigger it with a "build" step from each pipeline with the appropriate arguments to run the relevant tests. (It is also sometimes useful to rerun these tests without rebuilding the product)
stage('AcceptanceTest') {
steps {
build job: 'run-tests', parameters: ..., wait: true
}
}
So far I have seen that I can either:
trigger the job as normal. But this uses an extra agent/executor,
there doesn't seem to be a way to tell it to reuse the one from the
build (main pipeline). Both pipelines start with "agent { label 'master' }" but that
seems to mean "allocate a new agent on a node matching master".
trigger the job with the "wait: false" argument. This doesn't
block an executor but it does mean I can't report the results of the
tests in the main pipeline. It gives the impression that the test
stage has always succeeded.
Is there a better way?
I seem to have solved this, by adding "agent none" at the top of my main pipeline and moving "agent { label 'master' }" into the build stage. I can then leave my 'AcceptanceTest' stage without an agent and define it in the 'run-tests' job as before. I was under the impression from the docs that if you put agents in stages then all stages needed to have one, but it seems not to be the case. Which is lucky for this usecase...
I don't think that there's another way for declarative pipeline.
On the other hand for scripted pipeline you could execute this outside of node {} and it would just hold onto one executor on master releasing the one on slave.
stage("some") {
build job: 'test'
node {
...
Related question: Jenkis - Trigger another pipeline job in same machine - without creating new "Executor"
I can call another jenkins job using the build command. Is there a way I can tell another job to do a branch scan?
A multibranch pipeline job has a UI button "Scan Repository Now". When you press this button, it will do a checkout of the configured SCM repository and detect all the branches and create subjobs for each branch.
I have a multibranch pipeline job for which I have selected the "Suppress automatic SCM triggering" option because I only want it to run when I call it from another job. Because this option is selected, the multibranch pipeline doesn't automatically detect when new branches are added to the repository. (If I click "Scan Repository Now" in the UI it will detect them.)
Essentially I have a multibranch pipeline job and I want to call it from another multibranch pipeline job that uses the same git repository.
node {
if(env.BRANCH_NAME == "the-branch-I-want" && other_criteria) {
//scanScm "../my-other-multibranch-job" <--- scanScm is a fake command I made up
build "../my-other-multibranch-job/${env.BRANCH_NAME}"
I get an error on that build line, because the target multibranch pipeline job does not yet know that BRANCH_NAME exists. I need a way to trigger an SCM re-scan in the target job from this current job.
Similar to what you figured out yourself, I can contribute my optimization that actually waits until the scan has finished (but is subject to Script Security):
// Helper functions to trigger branch indexing for a certain multibranch project.
// The permissions that this needs are pretty evil.. but there's currently no other choice
//
// Required permissions:
// - method jenkins.model.Jenkins getItemByFullName java.lang.String
// - staticMethod jenkins.model.Jenkins getInstance
//
// See:
// https://github.com/jenkinsci/pipeline-build-step-plugin/blob/3ff14391fe27c8ee9ccea9ba1977131fe3b26dbe/src/main/java/org/jenkinsci/plugins/workflow/support/steps/build/BuildTriggerStepExecution.java#L66
// https://stackoverflow.com/questions/41579229/triggering-branch-indexing-on-multibranch-pipelines-jenkins-git
void scanMultiBranchAndWaitForJob(String multibranchProject, String branch) {
String job = "${multibranchProject}/${branch}"
// the `build` step does not support waiting for branch indexing (ComputedFolder job type),
// so we need some black magic to poll and wait until the expected job appears
build job: multibranchProject, wait: false
echo "Waiting for job '${job}' to appear..."
while (Jenkins.instance.getItemByFullName(job) == null || Jenkins.instance.getItemByFullName(job).isDisabled()) {
sleep 3
}
}
Ended up figuring this out shortly after posting the question. Calling build against the base multibranch pipeline job as opposed to a branch causes it to re-scan. The solution to my above snippet would have ended up looking something like...
node {
if(env.BRANCH_NAME == "the-branch-I-want" && other_criteria) {
build job: "../my-other-multibranch-job", wait: false, propagate: false // scan for branches
sleep 2 // scanning takes time
build "../my-other-multibranch-job/${env.BRANCH_NAME}"
The wait: false is important because otherwise you get "ERROR: Waiting for non-job items is not supported". The multibranch "parent" job is closer to a folder than a job, but it's a folder that supports the build command, and it does so by scanning the SCM.
But solving this just led to another problem, which is that with wait: false we have no way of knowing when the SCM Scan finished. If you have a large repository (or you're short on jenkins agents), the branch won't get discovered until after the second build command has already failed due to the branch not existing. You could bump the sleep time even higher, but that doesn't scale.
Fortunately, it turns out manually initiating the SCM scan isn't even needed if you have github webhooks set up for your jenkins. The branch will be discovered more-or-less instantly, so for my purposes this is solved another way. The reason I was running into it is we don't have webhooks set up in our dev jenkins, but once I move this code to prod it will work fine.
If you're trying to use JobDSL to set up multibranches calling multibranches and you don't have webhooks or something equivalent, the better path is probably to abandon multibranch for your second tier of jobs and use JobDSL to create folders and manage the branch jobs yourself.
Prior Jenkins2 I was using Build Pipeline Plugin to build and manually deploy application to server.
Old configuration:
That works great, but I want to use new Jenkins pipeline, generated from groovy script (Jenkinsfile), to create manual step.
So far I came up with input jenkins step.
Used jenkinsfile script:
node {
stage 'Checkout'
// Get some code from repository
stage 'Build'
// Run the build
}
stage 'deployment'
input 'Do you approve deployment?'
node {
//deploy things
}
But this waits for user input, noting that build is not completed. I could add timeout to input, but this won't allow me to pick/trigger a build and deploy it later on:
How can I achive same/similiar result for manual step/trigger with new jenkins-pipeline as prior with Build Pipeline Plugin?
This is a huge gap in the Jenkins Pipeline capabilities IMO. Definitely hard to provide due to the fact that a pipeline is a single job. One solution might be to "archive" the workspace as an "artifact" (tar and archive **/* as 'workspace.tar.gz'), and then have another pipeline copy the artifact and and untar it into the new workspace. This allows the second pipeline to pickup where the previous one left off. Of course there is no way to gauentee that the second pipeline cannot be executed out of turn or more than once. Which is too bad. The Delivery Pipeline Plugin really shines here. You execute a new pipeline right from the view - instead of the first job. Anyway - not much of an answer - but its the path I'm going to try.
EDIT: This plugin looks promising:
https://github.com/jenkinsci/external-workspace-manager-plugin/blob/master/doc/PIPELINE_EXAMPLES.md
I have a build pipeline in jenkins that builds and deploys the back-end components which expose a REST API. I have another build pipeline that builds and deploys the front-end components which call the back-end components. The back-end and front-end components live in seperate Git repositories.
The build job of each pipeline is kicked off when a commit occurs in each respective Git repository.
I would like to run automated functional tests at the end of the build pipeline of each build pipeline. But how do I know that both pipelines are finished and it should run the functional tests? Can it link the two pipelines together?
One approach is to use the Locks and Latches plugin and give each of the jobs on each pipeline their own Lock eg Pipeline-A and Pipeline-B, then the job that runs the tests is configured to obtain the lock on both Pipeline-A and Pipeline-B. This both prevents the test job running if any part of either pipeline is running, and blocks any changes on the pipeline whilst the tests are running.
If you'd only like to lock on the deploy jobs, you can use the same approach but only configure the deploy jobs with the locks; this will allow normal builds to run as normal, but deploy jobs queue up whilst the tests run.
Assumptions;
Any Deploy jobs are triggering a test execution
A second approach is to have your job pipelines setup such that before performing a deployment they trigger a single job in the following layout;
EndOfPipelineA -> SystemDeploymentController
EndOfPipelineB -> SystemDeploymentController
SystemDeploymentController -> DeployAppOne
SystemDeploymentController -> DeployAppTwo
DeployAppTwo -> TestExecution
DeployAppOne -> TestExecution
Then you use the Join plugin to only run the TestExecution job when both the deployments are complete AND successful.
The second approach allows you to:
conditionally control the execution of the test execution depending on the success of
deployments,
Have a single job that'll let you redeploy your whole system if you make any changes to the system it runs on, AND then run tests automatically.
Potentially make use of the Promotions plugin to highlight "good configurations" where both apps worked well together
However it is a bit trickier to manage.
Although this is an old question, you might consider restructuring your build pipeline using the Build Flow Plugin
https://wiki.jenkins-ci.org/display/JENKINS/Build+Flow+Plugin
It will have the advantage of keeping your pipeline logic in one place.
You can use build step for this. Let's say you have a pipeline named parent and child. In parent pipeline you can define:
pipeline {
agent any
stages {
stage ('call-child-pipeline') {
steps {
build job: 'child'
}
}
}
}
you can also pass some parameters to the child pipeline:
stage ('call-child-pipeline') {
steps {
build job: 'child', parameters: [string(name: 'my_param', value: "my_value")]
}
}
if you don't want to wait until the child pipeline is finished, add wait: false, e.g.
build job: 'child', wait: false