-
Bug
-
Resolution: Fixed
-
Critical
-
Jenkins 1.656
OpenStack Cloud Plugin 2.1
Launching a slave in OpenStack Havana
When launching a cloud slave, the VM is created and starts booting, but before it is available, Jenkins fails with this error.
[04/05/16 17:09:02] [SSH] Opening SSH connection to 10.0.2.39:22.
No route to host
ERROR: Unexpected error in launching a slave. This is probably a bug in Jenkins.
java.lang.IllegalStateException: Connection is not established!
at com.trilead.ssh2.Connection.getRemainingAuthMethods(Connection.java:1030)
at com.cloudbees.jenkins.plugins.sshcredentials.impl.TrileadSSHPublicKeyAuthenticator.getRemainingAuthMethods(TrileadSSHPublicKeyAuthenticator.java:88)
at com.cloudbees.jenkins.plugins.sshcredentials.impl.TrileadSSHPublicKeyAuthenticator.canAuthenticate(TrileadSSHPublicKeyAuthenticator.java:80)
at com.cloudbees.jenkins.plugins.sshcredentials.SSHAuthenticator.newInstance(SSHAuthenticator.java:207)
at com.cloudbees.jenkins.plugins.sshcredentials.SSHAuthenticator.newInstance(SSHAuthenticator.java:169)
at hudson.plugins.sshslaves.SSHLauncher.openConnection(SSHLauncher.java:1212)
at hudson.plugins.sshslaves.SSHLauncher$2.call(SSHLauncher.java:711)
at hudson.plugins.sshslaves.SSHLauncher$2.call(SSHLauncher.java:706)
at java.util.concurrent.FutureTask.run(FutureTask.java:266)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
at java.lang.Thread.run(Thread.java:745)
[04/05/16 17:09:05] Launch failed - cleaning up connection
[04/05/16 17:09:05] [SSH] Connection closed.
If I wait about a minute and click 'Launch Slave Agent' it works fine.
The solution, it seems, would be to permit retries on a 'No route to host' error, rather than failing immediately. Also, this worked fine a few versions back. I don't know exactly when it broke.
- depends on
-
JENKINS-41163 Launch retry not effective most of the time
-
- Resolved
-
[JENKINS-34100] Launch of cloud slave fails with "No route to host"
Component/s | New: ec2-plugin [ 15625 ] | |
Component/s | Original: core [ 21134 ] | |
Key |
Original:
|
New:
|
Workflow | Original: classic default workflow [ 170129 ] | New: JNJira [ 170129 ] |
Project | Original: Infrastructure [ 10301 ] | New: Jenkins [ 10172 ] |
Component/s | New: openstack-cloud-plugin [ 20631 ] | |
Component/s | Original: ec2-plugin [ 15625 ] |
Assignee | Original: R. Tyler Croy [ rtyler ] | New: Marat Mavlyutov [ mavlyutov ] |
Assignee | Original: Marat Mavlyutov [ mavlyutov ] | New: Oliver Gondža [ olivergondza ] |
Hmm, there ware some changes to the code in 2.1 where I expected to improve the situation. Please have a look into the service log for messages involving the slave name. Also, what is in the $JENKINS_HOME/logs/slaves/$SLAVE_NAME? Usually, there is couple of logs tracking how several instances of "no route to host" / "connection refused" turned into successful connection.