When configuring a build agent with the use of a PowerShell script, provided by an on-site TFS2015, the script errors out because it cannot find the agent pool on the server.
However, the agent pool 'default' definitely exists. On another server the same script works as expected, and builds run. See the script output below.
I have tried to following:
Create a new agent pool on the server, and reference that in the script. Same problem "Agent pool not found".
Installed Visual Studio 2015 with minimal features.
The user running the script is member of 'Build Administrators' and 'Build Service Accounts' security groups.
What is causing the script to fail on agent pool verification?
PS C:\Windows\system32> E:\Build\agentConfigureAgent.ps1
Enter the name for this agent (default is Agent-SRV001): BUILD002
Enter the URL for the Team Foundation Server (default is: http://[ip-address]:8080/tfs
Configure this agent against which agent pool? (default pool name is 'default'):
Enter the path of work folder for this agent (default is 'E:\Build\agent_work'): E:\Build\Work
Would you like to install the agent as a Windows Service (Y/N) (default is Y): n
Would you like to unconfigure any existing agent (Y/N) (default is N; the agent will be updated):
Configuring agent
Unblocking files Calling agent configure without /RunningAsService
Calling agent configure without /Force
Unable to find a agent pool with the name: default
WARNING: Configure agent failed, but you might fix this problem by configure with /Force.
Would you like to try agent configure again with parameter -Force (Y/N): y
Calling agent configure without /RunningAsService
Calling agent configure with /Force
Unable to find a agent pool with the name: default
E:\Build\agent\ConfigureAgent.ps1 : Configure agent failed. At line:1 char:1
+ E:\Build\agent\ConfigureAgent.ps1 + ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ CategoryInfo : NotSpecified: (:) [Write-Error], WriteErrorException
+ FullyQualifiedErrorId : Microsoft.PowerShell.Commands.WriteErrorException,ConfigureAgent.ps1
The user account running the build process script, needs to be part of the Agent Pool Administrator Accounts.
The error message from the script is misleading.
To resolve this error, I had to add the Windows user which is running the TFS build agent to the "Advanced" access level. This can be done under the root TFS Administer Server (top right gear) > Access levels UI. (http://your-tfs-server-here:8080/tfs/_admin/_licenses) My default access level for the server is set to Stakeholder.
The answer from #RoeIF worked for me, but I had a bit of trouble finding the page needed to add the account. Here's a screen grab to help:
Related
I'm new to Jenkins and I am trying to play around with it.
I'm trying to run a pipeline with a command that will run a simple dir on a remote windows fileserver (with a UNC path provided).
pipeline {
agent any
stages {
stage('Read File') {
steps {
bat 'whoami'
bat label: 'check directory', script: 'dir \\\\filesrv\\C$\\NewUser'
}
}
}
}
The whoami command returns the Jenkins AD user i configured to run the service on the slave
but after that I get an error Access is denied.
I tried giving the Jenkins AD service user local admin permission on the Jenkins master and slave servers and also on the file server. didn't help.
I also tried to explicitly giving that user full control permission on the folder I'm trying to access (located on the file server). didn't help.
I also tried giving permission to the computer accounts like many thread suggenst and point to this link https://serverfault.com/questions/135867/how-to-grant-network-access-to-localsystem-account also didn't help.
Will appreciate some assistance in understanding what permission is it missing?
Thanks in advance
i'll post the solution for any feature reference for those who are using windows environment...
the thing i was missing was making the target folder a shared folder.
so instead of \\\\filesrv\\C$\\NewUser' the path that worked is \\\\filesrv\\NewUser'
where NewUser is the name of the share
I am trying to create CI/CD pipeline using jenkins and my environment devops/deployment is windows servers 2016.
Pipeline script written in apache groovy. Which call batch file for deploy the code, that batch script is doing 2 tasks.
deploy code on another host machine (using robocopy command)
start w3svc service (using sc command).
Both command are mentioned in batch file which is call through groovy script with simple 'bat batchfilelocation /name.bat'. here is batch script:
#echo [off]
net use "\\<servername>\<shareddirectory>" <password> /user:<domain\user> /persistent:no
robocopy "C:\jenkins\AngularPipeline\dist\<project>" \\$<Dir>eddistinationServer>\<shar /E /PURGE
sc \\$servername start w3svc
echo on
echo close batch script!
For sc command execuation it is giving below error:
C:\jenkins\AngularPipeline>sc \ start w3svc [SC]
StartService: OpenService FAILED 5:
Access is denied.
This script is working well while I directly called it through command prompt(non administrator). access denied only comes while this script run under the jenkins groovy.
I have checked below possibilities :
WMI permission on source and target machine
//Confirm that WinRM is properly configured. c:>Winrm quickconfig
//Make sure that the remote server allows commands from any machine. PS c:>Set-item wsman:localhost\client\trustedhosts -value *
Restart Windows Remote management service restarted on both machines
C:\Windows\system32>sc \ sdshow w3svc
D:(A;;CCLCSWRPWPDTLOCRRC;;;SY)(A;;CCDCLCSWRPWPDTLOCRSDRCWDWO;;;BA)(A;;CCLCSWLOCRRC;;;IU)(A;;CCLCSWLOCRRC;;;SU)S:(AU;FA;CCDCLCSWRPWPDTLOCRSDRCWDWO;;;WD)
Please suggest how to resolve the same.
Recommended Approach:
If you are not running this Jenkins job on an another agent then you could re-configure your Jenkins service to use local administrator account. To do so follow below steps
Open Start Menu
Search 'Services'
Find "Jenkins" service
Right click on Jenkins service
Open Properties
Click on "Log on" tab
Select "This Account"
Browse for local system admin account and select it
Add Password and confirm password for same id
click OK
Restart/Start Jenkins Service
Second Approach:
Create a bat file on the server/workstation where Jenkins is installed
Create a shortcut of the file
Right click on Shortcut and go to 'properties'
Click on 'Comp'
check the check box next to "Run as administrator"
Click OK and Then
Click Apply Now, Go to Control Panel -> System and Security ->
Action Center -> Change User Account Control Settings
Set the level to "never notify" Run that bat shortcut using Jenkins job
In this case the job will marked as failed but the script will get executed.
Deploying a PDF documentation Jenkins executes a simple batch file:
#echo off
REM
REM Adapt the filepath as needed
REM
set PDF_OUTPUT_DIR="F:\svnWorkingCopy\kostra\trunk\DITA\__out_pdf"
set PDF_DEPLOY_DIR="\\FILE-SRV01\Austausch\Setup_Masterbuild\KOSTRA_AKVS\Deutsch\KostraAKVS"
del %PDF_DEPLOY_DIR%\KostraUserManual.pdf /F /Q
copy %PDF_OUTPUT_DIR%\kostra.pdf %PDF_DEPLOY_DIR%\KostraUserManual.pdf /Y
It quits with permission denied. End of console output (in German) is:
[kostra] $ cmd /c call C:\Windows\TEMP\jenkins1093559218605334188.bat
F:\svnWorkingCopy\kostra>F:\svnWorkingCopy\kostra\trunk\DITA\kostra-pdf-deploy.bat
Zugriff verweigert
0 Datei(en) kopiert.
Build step 'Windows Batch-Datei ausführen' marked build as failure
Archiviere Artefakte
Discard old builds...
#111 is removed because old than numToKeep
Sending e-mails to: fjk#akgsoftware.de
Finished: FAILURE
The Jenkins user is the user logged on that machine (Win Server 2012 r2) and has all Jenkins rights.
If I open CMD and execute the batch file manually, everything goes OK. Any idea why Jenkins refuses to copy the file? Is there a way to get a more detailed error msg. from Jenkins?
I see you are running Jenkins on windows. First check that if the user that started Jenkins is allowed to write on your directory. If that is not the case restart Jenkins with a user allowed to write in that directory.
Found the solution for my environment (Jenkins as Windows Service):
The default system user assigned to the Jenkins service must be changed to the logged on windows user.
Even after doing so Jenkins as a service doesn't know about mapped network drives, so you must change these to UNC paths if you have to deploy to a local network share.
I am getting errors from the durable task plugin when I run my pipeline dsl jenkins job.
The error message suggests that I should use:
-Dorg.jenkinsci.plugins.durabletask.BourneShellScript.HEARTBEAT_CHECK_INTERVAL=300
This is the error I get:
\workspace\ne-sw-manifest_master-5ZF5EWBP7EVBXEBF6AS3C6UQLIXLCS3HRKYND6TPQAPIKZPFBDLQ#tmp\durable-252b3bfd
(JENKINS-48300: if on a laggy filesystem, consider -Dorg.jenkinsci.plugins.durabletask.BourneShellScript.HEARTBEAT_CHECK_INTERVAL=300)
I am not sure where to set this property.
I tried on Jenkins master -> Configure system -> Global properties -> Environment variables:
Name:org.jenkinsci.plugins.durabletask.BourneShellScript.HEARTBEAT_CHECK_INTERVAL
Value:300
But, I am not sure if this is the right place to add this property OR if it has come into effect.
Also, I haven't restarted the master or slave.
My jenkins set-up is Linux master (Jenkins ver. 2.107.1) and Linux and Windows Slaves.
My build is on a Windows slave (physical machine)
option 1:
Add in your pipeline
script {
System.setProperty("org.jenkinsci.plugins.durabletask.BourneShellScript.HEARTBEAT_CHECK_INTERVAL", "3800");
}
after Running the approve the script in security settings at Manage Jenkins – In-process Script approval.
Option 2:
go to Manage Jenkins -> Script Console
and run
System.setProperty("org.jenkinsci.plugins.durabletask.BourneShellScript.HEARTBEAT_CHECK_INTERVAL", "3800");
This CloudBees article explains how to set Jenkins Java arguments.
Note: you'll need to restart your Jenkins instance.
Edit: As per sirch's comment, I'm copying here the instructions for RedHat and Debian distro's.
Debian / Ubuntu based Linux distributions
If your configuration file is under /etc/default/ look for the argument JAVA_ARGS. It should look something like this:
JAVA_ARGS="-Djava.awt.headless=true"
Then, add the arguments:
JAVA_ARGS="-Xmx2048m -XX:MaxPermSize=512m -Djava.awt.headless=true"
RedHat Linux based distributions
If your configuration file is under /etc/sysconfig/ look for the argument JENKINS_JAVA_OPTIONS. It should look something like this:
JENKINS_JAVA_OPTIONS="-Djava.awt.headless=true"
Then, add the arguments:
JENKINS_JAVA_OPTIONS="-Xmx2048m -XX:MaxPermSize=512m -Djava.awt.headless=true"
set it either
JAVA_OPTS
or
JNLP_PROTOCOL_OPTS
which will be included in jenkins slave Startup Options
I am trying Jenkins to execute an ansible playbook.
But I am getting the unreachable host error which I don't get otherwise.
fatal: [vogo-alpha.cloudapp.net]: UNREACHABLE! => {"changed": false, "msg": "Authentication failure.", "unreachable": true}
I have given this variable in ansible hosts file,
ansible_ssh_private_key_file=/home/luvpreet/.ssh/id_rsa
I think it is because the user jenkins is playing those playbooks and it cannot read this private key file. I tried to make jenkins' user home folder but it was not successful.
It can be done if I switch to the user luvpreet and then run these playbooks.
How do I switch to another user via jenkins shell ?
OR
Is there any other way this problem can be solved ?
There are a couple of possibilities why your solution is working. Most likely because Ansible is trying to ssh to your target machine as the jenkins user which isn't on said machine. I'd approach the problem from a different angle.
First, I'd install the Ansible plugin for Jenkins. This allows you to use the built in credentials located at "Manage Jenkins > Manage Credentials". There you can copy and paste your key in (or point to a key file located on the jenkins server) and set the username that will ssh to the target machine. In your job configuration choose "Invoke Ansible Playbook" for your build step rather than shell. There will be a "Credentials" parameter where you can specify the ssh key you added earlier. The rest should be pretty self explanatory.