I am using the Jenkins Delivery Pipeline plugin. I have some parameterized builds that are manually triggered. If I trigger a build from the project page, the build shows a screen where parameters need to be defined (as expected). However, when I start a parameterized build from the Delivery Pipeline plugin, it just starts the build without offering a screen. This is strange behavior, is it possible to get the pipeline plugin to show the parameterized build screen?
Thanks for your help!
I'm answering this question in general.
You need to use the Parameterized Trigger Plugin, or use the Build Pipeline Plugin. This issue with Delivery Pipeline plugin is still being solved by the Jenkins Team. See this link for the update about the issue at https://issues.jenkins-ci.org/browse/JENKINS-25685. You can get to know when it gets resolved from that link.
You can use the Build Pipeline plugin if it can be incorporated in your use case. There was a similar issue for the Build Pipeline plugin which is fixed now. It got fixed within 22 days (https://issues.jenkins-ci.org/browse/JENKINS-25427, https://github.com/jenkinsci/build-pipeline-plugin/pull/56). You can just hope that the same issue with the Delivery Pipeline plugin is fixed soon.
Can you provide me the version of Jenkins, environment and the plugin version? I can update my answer based on the answers you give.
When you are using the Delivery Pipeline plugin, and you have manually-triggered parameterized builds, as long as you configure the upstream job to pass along the parameters to the downstream job, when the "build trigger button" is clicked on the pipeline view page, the parameters are automatically passed along.
For instance, let's say you have a setup like this:
Compile_Project ---> Deploy_Project
Let's assume that you are passing a variable called versionNumber from the Compile_Project to the Deploy_Project jobs. Let's also assume that you're using Subversion for your SCM, and your versionNumber looks like 1.0.${SVN_REVISION}. ${SVN_REVISION} is automatically provided by Jenkins, so your version number will look something like 1.0.1234, where 1234 is the commit number provided by Subversion.
On your Delivery Pipeline view, let's say that it's configured to show 3 pipeline instances, and that manual triggers are enabled in the view settings. Your pipeline view page might look something like this (yay ASCII art!):
Compile_Project ---> Deploy_Project (>)
Compile_Project ---> Deploy_Project (>)
Compile_Project ---> Deploy_Project (>)
In this case, I'm using (>) to represent the manual trigger button. The button on the bottom would attempt to deploy version 1.0.1234, the middle button would attempt to deploy version 1.0.1235, and the top button would attempt to deploy version 1.0.1236, assuming your project has gotten consecutive SVN commits.
In order to pass the parameter from the Compile_Project to the Deploy_Project job, you need to do the following. (Note: it sounds like you've already done this part, but I'm including it just in case you might have missed a step, and also for the sake of completeness.)
In the Compile_Project job's configuration, as a Post-Build Action, choose "Build other projects (manual step)". In the "Downstream project names" box, enter Deploy_Project, and then from the "Add Parameters" drop-down, select "Predefined Parameters". In the "Parameters" text area that appears, create a parameter to pass along, which I'll call VERSION_NUMBER. What you'll enter in the text area is then VERSION_NUMBER=1.0.${SVN_REVISION}. This will enable the parameter to get passed from the Compile_Project to the Deploy_Project. However, you're not quite done yet.
In the Deploy_Project job's configuration, you need to set it up to accept the parameter you're passing into the job. To do so, configure the Deploy_Project, and check the "This build is parameterized" checkbox. Then add a String parameter from the "Add parameter" drop-down. In the "Name" field, enter VERSION_NUMBER. At this point, you can then use ${VERSION_NUMBER} in the Deploy_Project's configuration wherever you need in order to specify the correct version number of the project to deploy.
Related
Is it possible to edit/save the code of "Job Configuration" (the structure and order of UI elements) ?
The pipeline job get the pipeline code from git, but this occurs only AFTER the job started. I want to edit the UI of job parameters (BEFORE user press "build")
Its necessary for 2 reasons :
its much easier to cut/paste text lines, than move elements up/down with the mouse.
can be saved and deployed easily on a new Jenkins machine
You can edit the configuration before the job is run to add the initial build parameters. If you then have the build parameters also defined in the pipeline, the parameters will be overwritten with the ones defined in the pipeline.
However, for that first build, if the build parameters defined in the job configuration and the ones in the pipeline match, the build parameter values will be used in that initial run.
I am switching from the github pull request builder plugin (for security reasons) and am trying to get the same functionality from Pipelines (using different plugin). I think I have just about everything, however I can't seem to find a way to re-trigger a build simply by a trigger phrase like in github pull request builder plugin. Is that possible via pipelines?
By trigger phrase, I mean that a user can make a comment on the PR saying "Jenkins re-test" and it will kick off the build again.
You can put a condition at the top of the build script to check for the message. You can access the changesets using currentBuild.ChangeSets. The last changeset is at the end of the array. Then you need to access the last element of that changeset. Finally you can access the message via message property. You can then search for your keyword.
I am doing the opposite (not triggering the build with a phrase) but never tried for pullrequests though.
Another idea is to use the "ignore builds with specific message" property and setting this message to be a regex with look ahead that accepts everything except the keyword. I don't really recall the syntax though :/
When I trigger a job remotely using the API the job on jenkins say: "Started by remote host IP". But my jobs can trigger for different reasons and from different sources so it would be great if I in Jenkins could look at a job and directly see the trigger reason.
I know I could pass a string parameter and then in each job view the parameters - but that's not very visible. I would like to preferably see it already in the list of all jobs. Something like: Reason: X Source: N.
Is there any way to do this ?
I have noticed that our release jobs take a version parameter that shows like I want - but I don't want to abuse that since these runs are not releases.
Build Name
I think you want to set the build name to a different one like here:
So you can use the Build Name Setter Plugin and use the build step Update build name.
For your case you can use:
#${BUILD_NUMBER} ${BUILD_CAUSE}
Cause Badge
Have a look on Build Trigger Badge Plugin, it represents with an icon the reason of the build.
Extract the job cause
One possibility to extract the job cause is to use groovy, currentBuild is injected into the runtime and includes information about the causes and so on.
This prints the first (!) cause:
println ("Cause: " + currentBuild.causes[0])
Unfortunately, even with the Build Trigger Badge Plugin activated http://<Your Jenkins>/job/<Your job's name>/api/xml doesn't show the trigger cause.
You could parse the job's HTML for:
<div class="middle-align build-badge" style="width: 100%;">
<img width="16" height="16" src="/plugin/buildtriggerbadge/images/user-cause.png"
alt="Trigger UserIdCause : Started by user YourUser" title="Started by user YourUser">
</div>
See also the EnvInject Plugin:
This plugin also exposes the cause of the current build as an environment variable.
The issue here is once the first Job in a Jenkins pipeline is done, we need to ask some inputs from user and based on the user Inputs to decide the next job to be triggered(Job2 or Job3)
tried build flow and parameterzied trigger plugin but didn't find any suitable option under these.
Any other plugin or jenkins feature which can help in achieving the above scenario?
There are a few plugins I have tried which collect user input on manually triggered jobs in a build pipeline: Active Choices Plug-in 1.2 and Extensible Choice Parameter 1.3.2.
With Active Choices you define a list of selections and a default value. With Extensible Choice Parameter you can have a text box and a default value.
This is how they work for me in Build Pipeline 1.4.8 on jenkins 1.628
If you run the manual step directly in the pipeline the default is used and other parameters propagate through correctly.
If you open the step there is an option to Build with Parameters which will ask for the user input. This works but other parameters like the build number do not propagate through so the pipeline is broken, and the pipeline screen does not show the status.
Jenkins will never pause and ask a user for inputs. It is an automated build system. It doesn't expect anyone sitting at the console watching the progress.
You can provide "inputs" or parameters when you manually trigger the job, i.e on the first job in your pipeline. You can them pass these parameters to downstream jobs, either through the Parameterized Trigger plugin or through a file copied between jobs.
If you need a human decision in the middle of your build flow, consider Promoted Builds plugin. With this plugin, a human can select a build, and then decide which "Promotion" to execute (which could branch your workflow as you need). The promotions can also be automated if needed, based on criteria and not human input.
In a single Jenkins job, we can trigger a build by specifying a schedule and also by polling. But then, in both the cases, the build is triggered, and the deploy operation that I have configured as a post-build step (using PostBuild Task plugin) also happens. I want that the build happens whenever a change is detected by polling, but deploy should happen only according to the schedule I have provided.
Is it possible to do it in a single job, or do I have to configure 2 separate jobs for them ?
You said you are using PostBuild Task plugin. This allows to do a regular expression on the console log to determine whether to execute a task or not.
Builds started by schedule will have Started by timer at the top of the log. All you need to do is add this expression to your PostBuild step under "Log Text" field. If you are already using some criteria in there, click "Add" button to add another "Log Text" field, and use the "AND" operator between them
It will be cleaner to do it in 2 jobs. However, if you really need to have it in one job, you could use a combination of Jenkins plugins to do the job.
Use EnvInject Plugin to expose the BUILD_CAUSE and/or BUILD_CAUSE_SCHEDULED* environment variable(s). (This may not be necessary, you might be able to reference the Jenkins variables within the Jenkins configuration by default)
Use Flexible Publish plugin, post-build action, to set up a conditional publish step when BUILD_CAUSE == SCHEDULED, or when BUILD_CAUSE_SCHEDULED == true. (Just test one condition.) Note that you'll need to use Jenkins' expression syntax, like so:
${ENV,var="BUILD_CAUSE_SCHEDULED"}
* BUILD_CAUSE_SCHEDULED is not its real name, you'll need to find this out on your own, sorry.