-
Improvement
-
Resolution: Duplicate
-
Minor
-
None
When you have multiple test runs using pipelines, it would be super awesome to be able to annotate the different test runs with a name to allow differentiation, e.g.
node('windows') { ... junit testResults:'target/tests.xml' testRunName: 'windows' } node('linux') { ... junit testResults:'target/tests.xml' testRunName: 'linux' }
This way you can tell which test run the tests came from.
The following screenshot illustrates how this would apply (by basically appending the test run name to each case)
- duplicates
-
JENKINS-27395 distinguish test reports filed from different parts of workflow
-
- Resolved
-
[JENKINS-41710] Annotate test results with a test run name
I'm not sure why it's marked as duplicated/resolved?
what is the way to achieve this?
Simplified Jenkinsfile
def jdks = ["jdk8", "jdk9"] def oss = ["linux","windows","osx"] def builds = [:] for (def os in oss) { for (def jdk in jdks) { builds[os+"_"+jdk] = getFullBuild( jdk, os ) } } def getFullBuild(jdk, os) { return { node(os) { // System Dependent Locations def mvntool = tool name: 'maven3', type: 'hudson.tasks.Maven$MavenInstallation' def jdktool = tool name: "$jdk", type: 'hudson.model.JDK' // Environment List mvnEnv = ["PATH+MVN=${mvntool}/bin", "PATH+JDK=${jdktool}/bin", "JAVA_HOME=${jdktool}/", "MAVEN_HOME=${mvntool}"] mvnEnv.add("MAVEN_OPTS=-Xms256m -Xmx1024m -Djava.awt.headless=true") withEnv(mvnEnv) { timeout(time: 90, unit: 'MINUTES') { // Run test phase / ignore test failures sh "mvn -V -B install -Dmaven.test.failure.ignore=true -Prun-its" // Report failures in the jenkins UI step([$class: 'JUnitResultArchiver', testResults: '**/target/surefire-reports/TEST-*.xml']) } } }
All test add aggregated in a single junit result and I cannot find which jdk/os failed
I would be happy to use something such
step([$class: 'JUnitResultArchiver', testResults: '**/target/surefire-reports/TEST-*.xml', testRunName: $jdk+"_"+$os])
Olivier Lamy
added a comment - - edited I'm not sure why it's marked as duplicated/resolved?
what is the way to achieve this?
Simplified Jenkinsfile
def jdks = [ "jdk8" , "jdk9" ]
def oss = [ "linux" , "windows" , "osx" ]
def builds = [:]
for (def os in oss) {
for (def jdk in jdks) {
builds[os+ "_" +jdk] = getFullBuild( jdk, os )
}
}
def getFullBuild(jdk, os) {
return {
node(os) {
// System Dependent Locations
def mvntool = tool name: 'maven3' , type: 'hudson.tasks.Maven$MavenInstallation'
def jdktool = tool name: "$jdk" , type: 'hudson.model.JDK'
// Environment
List mvnEnv = [ "PATH+MVN=${mvntool}/bin" , "PATH+JDK=${jdktool}/bin" , "JAVA_HOME=${jdktool}/" , "MAVEN_HOME=${mvntool}" ]
mvnEnv.add( "MAVEN_OPTS=-Xms256m -Xmx1024m -Djava.awt.headless= true " )
withEnv(mvnEnv) {
timeout(time: 90, unit: 'MINUTES' ) {
// Run test phase / ignore test failures
sh "mvn -V -B install -Dmaven.test.failure.ignore= true -Prun-its"
// Report failures in the jenkins UI
step([$class: 'JUnitResultArchiver' ,
testResults: '**/target/surefire-reports/TEST-*.xml' ])
}
}
}
All test add aggregated in a single junit result and I cannot find which jdk/os failed
I would be happy to use something such
step([$class: 'JUnitResultArchiver' ,
testResults: '**/target/surefire-reports/TEST-*.xml' , testRunName: $jdk+ "_" +$os])
This is a very good idea and makes replacing matrix jobs with (parallel) pipelines easier.