Search code examples
jenkinsvirtual-machineslave

Jenkins slave fails to execute after VM reboot

I had a perfectly working AWS VM, which I duplicated in order to have similar environements for stage and production.
I stopped the VM to create a snapshot, but now, when I launch a build with the jenkins slave on it, I get this error :

Building remotely on <nodename> in workspace /home/ubuntu/workspace/<workspaceName>
Installing JDK 
java.io.IOException: Unable to find JDK with ID=
    at hudson.tools.JDKInstaller.locate(JDKInstaller.java:422)
    at hudson.tools.JDKInstaller.performInstallation(JDKInstaller.java:134)
    at hudson.tools.InstallerTranslator.getToolHome(InstallerTranslator.java:68)
    at hudson.tools.ToolLocationNodeProperty.getToolHome(ToolLocationNodeProperty.java:107)
    at hudson.tools.ToolInstallation.translateFor(ToolInstallation.java:205)
    at hudson.model.JDK.forNode(JDK.java:130)
    at hudson.model.AbstractProject.getEnvironment(AbstractProject.java:355)
    at hudson.model.Run.getEnvironment(Run.java:2228)
    at hudson.model.AbstractBuild.getEnvironment(AbstractBuild.java:932)
    at hudson.scm.SubversionSCM.checkout(SubversionSCM.java:828)
    at hudson.scm.SCM.checkout(SCM.java:484)
    at hudson.model.AbstractProject.checkout(AbstractProject.java:1274)
    at hudson.model.AbstractBuild$AbstractBuildExecution.defaultCheckout(AbstractBuild.java:609)
    at jenkins.scm.SCMCheckoutStrategy.checkout(SCMCheckoutStrategy.java:86)
    at hudson.model.AbstractBuild$AbstractBuildExecution.run(AbstractBuild.java:531)
    at hudson.model.Run.execute(Run.java:1738)
    at hudson.model.FreeStyleBuild.run(FreeStyleBuild.java:43)
    at hudson.model.ResourceController.execute(ResourceController.java:98)
    at hudson.model.Executor.run(Executor.java:381)
Finished: FAILURE

Alternatively, I sometimes got this precision
(between "Installing JDK" and "java.io.IOException: Unable to find JDK with ID=")

FATAL: Failed to get Environment Vars from build
java.lang.RuntimeException: Failed to get Environment Vars from build
    at jenkins.plugins.publish_over.BPPlugin.getEnvironmentVariables(BPPlugin.java:84)
    at jenkins.plugins.publish_over.BPPlugin.perform(BPPlugin.java:93)
    at jenkins.plugins.publish_over_ssh.BapSshPreBuildWrapper.setUp(BapSshPreBuildWrapper.java:64)
    at hudson.model.Build$BuildExecution.doRun(Build.java:154)
    at hudson.model.AbstractBuild$AbstractBuildExecution.run(AbstractBuild.java:536)
    at hudson.model.Run.execute(Run.java:1738)
    at hudson.model.FreeStyleBuild.run(FreeStyleBuild.java:43)
    at hudson.model.ResourceController.execute(ResourceController.java:98)
    at hudson.model.Executor.run(Executor.java:381)

I didn't notice what causes those extra lines to appear, since I tried everything I thought about, including

  • Flushing every files in my user directory (.jenkins, workspace, and every jdk related file from slave.jar and the jdk dir to the install scripts...)
  • Recreating my jenkins slave and / or build, either from scratch, or as a copy of my initial build.
  • Recreating again the VM from the snapshot (just to be sure I didn't screw anything up in-between)

JDK is already here, of course. Everything worked just fine before this reboot. So my first question is "Why does it try to install a JDK ?"

I have no idea about the "Environment Vars" it's trying to get, and am not aware of any change in the VM config, so I'm quite out of ideas.

Solution

  • Looks like the problem was "just" that the configured JDK setting value disappeared from Jenkins.

    I have absolutely no idea WHY it did, since no one would have modified this setting, and it's in a submenu which needs to be opened first, making it protected from manipulation mistakes.

    Anyway, here's how I fixed it. It's as simple as setting the JAVA_HOME value in Jenkins:


    • Go to Manage Jenkins -> Configure system
    • Click on "Installations JDK"
    • Untick "install automatically" (My environnement was already set up, I just needed to set the variable again)
    • Set JAVA_HOME (in my case, /usr/lib/jvm/java-7-openjdk-amd64 ), and optionally, name your install

    EDIT : since some other builds continued to fail, I had to create a symlink on one of my older servers, which had java installed in /usr/bin/java
    mkdir -p /usr/lib/jvm/java-7-openjdk-amd64/bin/
    ln -s /usr/bin/java /usr/lib/jvm/java-7-openjdk-amd64/bin/java
    I also added a JAVA_HOME environment variable with the same /usr/lib/jvm/java-7-openjdk-amd64 value in Jenkin's conf. Probably redundant

    Our Jenkins version is 1.609.3 (checked in /var/lib/jenkins/config.xml ),
    Please, comment if you experience the same problem, since I really doubt this mistake originated from a user.