There are many threads about similiar problems on stackoverflow but they couldn't help me with my specific case:
node{
def USER = "user"
def ADRESS= "adress"
def ARGUMENT= "argument"
stage('test'){
sshagent(['123123']) {
sh 'ssh ${USER}#${ADRESS} "command ${ARGUMENT} && command"'
}
}
}
In the console output ${USER} and ${ADRESS} will resolve to an empty string but ${ARGUMENT} will resolve to "argument". How do I have to format these placeholders that they work? I have tried everything I could find but nothing seems to work.
As #NoamHelmer pointed out:
sh "ssh ${USER}#${ADRESS} \""command ${ARGUMENT} && command\""
works.
Related
In my Jenkinsfile I want to dynamically find the unity version using a python script like so:
environment {
UNITY_EDITOR = bat(script: "py $WORKSPACE/get_versions.py --unity", returnStdout: true).trim()
UNITY_BASE = "C:/Program Files/Unity/Hub/Editor/$UNITY_EDITOR/Editor/Unity.exe"
UNITY_WRAPPER = "UnityBatchWrapper -silent-crashes -no-dialogs -batchmode -quit -unityPath \"$UNITY_BASE\""
}
post {
always {
script {
echo "Returning license"
licenseReturnStatus = bat (
script: "$UNITY_WRAPPER -returnlicense",
returnStatus: true
) == 0
}
}
From other stackoverflow answers this seems like it should work, but instead my Jenkins job errors out during the post-build step because $UNITY_WRAPPER isn't defined:
groovy.lang.MissingPropertyException: No such property: UNITY_WRAPPER for class: groovy.lang.Binding
I'm thinking the batch step is what's failing, even though Jenkins doesn't complain about it. I've also tried using $env.WORKSPACE and %WORKSPACE% and that doesn't work either.
I'm beginning to think $WORKSPACE doesn't exist til after the environments step...
Turns out I didn't have Python installed since it was an ephemeral GCP builder and I hadn't updated the node label yet.
For anyone reading this that has trouble with bat commands - be sure to put an # sign in front of your command like "#py ..." or else the command will be echoed in the output. Also trim your output so it doesn't have CRLF in it.
I'm newbie to Jenkins pipeline and writing a groovy script to parse a json file. However I'm facing an error which many have faced but none of the solutions worked for me. Below is my Jenkinsfile and error msg.
def envname = readJSON file: '${env.WORKSPACE}/manifest.json'
pipeline {
agent any
stages {
stage('Build') {
steps {
echo WORKSPACE
sh "ls -a ${WORKSPACE}"
}
}
}
}
[Pipeline] Start of Pipeline
[Pipeline] readJSON
[Pipeline] End of Pipeline
org.jenkinsci.plugins.workflow.steps.MissingContextVariableException:
Required context class hudson.FilePath is missing Perhaps you forgot
to surround the code with a step that provides this, such as: node at
org.jenkinsci.plugins.pipeline.utility.steps.AbstractFileOrTextStepExecution.run(AbstractFileOrTextStepExecution.java:30)
I even tried readJSON file: '${WORKSPACE}/manifest.json but that didn't work too. I'm sure the mistake is with the first line since when removing that line, there execution is successful. The docs are pretty helpful but I'm not able to track down where exactly I'm going wrong that is why posted here.
UPDATE:
I tried the following methods def envname = readJSON file: "./manifest.json" and def envname = readJSON file: "${env.WORKSPACE}/manifest.json" and even tried them defining under the steps block. Nothing worked. Below is the error msg I recieved when I defined them under step block
WorkflowScript: 5: Expected a step # line 7, column 13
def envname =
^
Below is the official syntax doc of readJson and I can see that I'm using the correct syntax only. but still doesn't work as expected.
https://www.jenkins.io/doc/pipeline/steps/pipeline-utility-steps/#readjson-read-json-from-files-in-the-workspace
'${env.WORKSPACE}/manifest.json' is interpolating the Groovy env map as a shell variable. You need to interpolate it as a Groovy variable like "${env.WORKSPACE}/manifest.json".
sh "ls -a ${WORKSPACE}" is interpolating the shell environment variable WORKSPACE as a Groovy variable. You need to interpolate it as a shell variable like sh 'ls -a ${WORKSPACE}'.
echo WORKSPACE is attempting to resolve the shell variable WORKSPACE as a first class Groovy variable expression. You need to use the Groovy env map instead like echo env.WORKSPACE.
As for the global variable indefinite type assignment on the first line: if it still throws the error above after making those fixes, then it may be due to invalid use of scripted syntax in a declarative syntax pipeline. You likely need to place it inside a step block within your pipeline in that case.
I've solved this myself with the help of "Matt Schuchard"'s below answer. I'm not sure whether this is the only way to solve but this worked for me.
pipeline {
agent any
stages {
stage('Json-Build') {
steps {
script {
def envname = readJSON file: "${env.WORKSPACE}/manifest.json"
element1 = "${envname.dev}"
echo element1
}
}
}
}
}
I have a pipeline job that runs a maven build. In the "post" section of the pipeline, I want to get the log file so that I can perform some failure analysis on it using some regexes. I have tried the following:
def logContent = Jenkins.getInstance()
.getItemByFullName(JOB_NAME)
.getBuildByNumber(
Integer.parseInt(BUILD_NUMBER))
.logFile.text
Error for the above code
Scripts not permitted to use staticMethod jenkins.model.Jenkins
getInstance
currentBuild.rawBuild.getLogFile()
Error for the above code
Scripts not permitted to use method hudson.model.Run getLogFile
From my research, when I encounter these, I should be able to go to the scriptApproval page and see a prompt to approve these scripts, but when I go to that page, there are no new prompts.
I've also tried loading the script in from a separate file and running it on a different node with no luck.
I'm not sure what else to try at this point, so that's why I'm here. Any help is greatly appreciated.
P.S. I'm aware of the BFA tool, and I've tried manually triggering the analysis early, but in order to do that, I need to be able to access the log file, so I run into the same issue.
You can use pipeline step httpRequest from here
pipeline {
agent any
stages {
stage('Build') {
steps {
echo 'Test fetch build log'
}
post {
always {
script {
def logUrl = env.BUILD_URL + 'consoleText'
def response = httpRequest(
url: logUrl,
authentication: '<credentialsId of jenkins user>',
ignoreSslErrors: true
)
def log = response.content
echo 'Build log: ' + log
}
}
}
}
}
}
If your jenkins job can run on linux machine, you can use curl to archive same goal.
pipeline {
agent any
stages {
stage('Build') {
environment {
JENKINS_AUTH = credentials('< credentialsId of jenkins user')
}
steps {
sh 'pwd'
}
post {
always {
script {
def logUrl = env.BUILD_URL + 'consoleText'
def cmd = 'curl -u ${JENKINS_AUTH} -k ' + logUrl
def log = sh(returnStdout: true, script: cmd).trim()
echo 'Build log: ' +
echo log
}
}
}
}
}
}
Above two approaches both require the credentials is Username and password format. More detail about what is it and how to add in Jenkins, please look at here
Currently this is not possible via the RunWrapper object that is made available. See https://issues.jenkins.io/browse/JENKINS-46376 for a request to add this.
So the only options are:
explicitly whitelisting the methods
read the log via the URL as described in the other answer, but this requires either anonymous read access or using proper credentials.
I'm trying to call a remote job from one Jenkins server to another, I have this working fine via a shell script. However, trying to translate it into a Jenkinsfile is causing me issues. The environment variable is always "null" when used inside of a stage, even thought this article says it should be globally available?
pipeline {
agent any
/* get crumb for CSRF */
environment {
def crumb = sh 'curl https://jenkins-remote/crumbIssuer/'
}
stages {
/* call remote job */
stage("remote") {
steps {
sh "curl -X POST -H ${env.crumb} https://jenkins-remote/buildWithParameters?foo"
}
}
}
}
The trimmed output looks like:
[remote_pipeline] Running shell script
+ curl -X POST -H null
I am using Jenkins v2.89.4, new "Pipeline" job, "pipeline script".
Thanks to #TrafLaf for pointing out the variable is null because it does not get set to the output of the curl command. My hacky solution was this:
environment {
def crumbRequest = sh 'curl https://jenkins-remote/crumbIssuer/ > crumbHeader'
crumbHeader = readFile('crumbHeader').trim()
}
As per the official documentation, This is how you define environment variables.
pipeline {
agent any
environment {
DISABLE_AUTH = 'true'
DB_ENGINE = 'sqlite'
}
stages {
stage('Build') {
steps {
echo "${env.DB_ENGINE}" # to access
}
}
}
}
But you have coded wrong,
environment {
def crumb = sh 'curl https://jenkins-remote/crumbIssuer/'
}
So please do the rest.
The sh task can now return output, so in theory the following should work (untested):
environment {
crumb = sh(script: 'curl https://jenkins-remote/crumbIssuer/',
returnStdout: true).trim()
}
For example:
var output=sh "echo foo";
echo "output=$output";
I will get:
output=0
So, apparently I get the exit code rather than the stdout. Is it possible to capture the stdout into a pipeline variable, such that I could get:
output=foo
as my result?
Now, the sh step supports returning stdout by supplying the parameter returnStdout.
// These should all be performed at the point where you've
// checked out your sources on the slave. A 'git' executable
// must be available.
// Most typical, if you're not cloning into a sub directory
gitCommit = sh(returnStdout: true, script: 'git rev-parse HEAD').trim()
// short SHA, possibly better for chat notifications, etc.
shortCommit = gitCommit.take(6)
See this example.
Note: The linked Jenkins issue has since been solved.
As mention in JENKINS-26133 it was not possible to get shell output as a variable. As a workaround suggested using of writ-read from temporary file. So, your example would have looked like:
sh "echo foo > result";
def output=readFile('result').trim()
echo "output=$output";
Try this:
def get_git_sha(git_dir='') {
dir(git_dir) {
return sh(returnStdout: true, script: 'git rev-parse HEAD').trim()
}
}
node(BUILD_NODE) {
...
repo_SHA = get_git_sha('src/FooBar.git')
echo repo_SHA
...
}
Tested on:
Jenkins ver. 2.19.1
Pipeline 2.4
You can try to use as well this functions to capture StdErr StdOut and return code.
def runShell(String command){
def responseCode = sh returnStatus: true, script: "${command} &> tmp.txt"
def output = readFile(file: "tmp.txt")
if (responseCode != 0){
println "[ERROR] ${output}"
throw new Exception("${output}")
}else{
return "${output}"
}
}
Notice:
&>name means 1>name 2>name -- redirect stdout and stderr to the file name
I had the same issue and tried almost everything then found after I came to know I was trying it in the wrong block. I was trying it in steps block whereas it needs to be in the environment block.
stage('Release') {
environment {
my_var = sh(script: "/bin/bash ${assign_version} || ls ", , returnStdout: true).trim()
}
steps {
println my_var
}
}
A short version would be:
echo sh(script: 'ls -al', returnStdout: true).result
def listing = sh script: 'ls -la /', returnStdout:true
Reference : http://shop.oreilly.com/product/0636920064602.do Page 433