I started trial with Bamboo recently. I ran bamboo server on my personal PC and configure our build server to be a remote agent. Our build environment is in Windows. The build process is carried out via TCL script, which basically executes Windows batch command to build different stuffs. The remote agent execute that TCL script after source code checkout. However, the Windows commands that are suppoesed to execute do not actually execute. For example, if we try to execute msbuild batch command from windows, we will do
exec "cmd << msbuild **.sln"
in the TCL script. The result of this command is 'msbuild' is not recognized as an internal or external command, operable program or batch file. The problem was gone for two builds then re-appear right after.
I tried just simply called the script from CMD on the remote agent and aslo execute the job on my PC(which is the default agent) through Bamboo. The mentioned problem does not occur.
Can anyone tell me if there are some other configurations that I have missed?
Thank you.
Well, your command:
exec "cmd << msbuild **.sln"
That's not going to work (unless you have a program called exactly that, which I think is illegal on Windows; you can't have * in a file name). Instead, you need to invoke it either via cmd or using a direct Tcl style invoke.
Which one will be best? I don't know; try them and find out for yourself.
Invoking via cmd.exe
exec cmd /c msbuild *.sln
Invoking a more Tcl-ish style
exec msbuild {*}[glob *.sln]
(Note that I think ** is always unnecessary as * just means “any number of any character” when expanding a glob. If you're using it to mean “recursive search through a directory structure”, say so and I'll provide code to do it using the find command in the Tcllib fileutil package.)
Related
I have a C program and I need to run two batch file which manage all the compilation toolchain. The First one sets the enviroment variables and the next one to be executed performs the compilation and the linking. For some reason, eventhough I am setting all the needed enviroments variables with the first batch file, some .exe that are called by the second batch file are not recognized by the VS code internal terminal (extended error message: "file.exe" is not recognized as internal command or/and an executable program or a batch file"). I have tried to run in sequence this two batch files (by directly calling it, without any shortcut) in the Windows prompt opened in VS code and the C program is successfully compiled. Thus, Is there a way to run my build task (this batch file) in the Windows prompt opened in VS code by using the shortcut ctrl+shift+b? I have tried to set some option in the settings but they do not work.
The problem is that I was calling in sequence this two batch file, namely by defining in the task.json that the compilation batch file must be executed after the other batch. But in this way, the two batch files will be executed in two different instances of the prompt, so the enviroment variable will exist only in the first prompt. That is why when the compilation batch is executed I get that problem "'file.exe' is not recognized as internal command or/and an executable program or a batch file".
In order to execute these two batch files in the same prompt it is necessary to define a task in the task.json with this command:
command: "batch1 && batch2",
I have created a powershell script which will transfer(using winscp.dll) the files from Jenkins windows server to Linux server. In Jenkins batch command, I have executed that powershell script and it works fine.
But when i tried the same in Jenkins pipeline job, it calls the powershell script and comes to the next step. Its not waiting for powershell script response.
bat 'powershell.exe -ExecutionPolicy Bypass "D:\\Test\\Deploymentscripts\\PowerShellScript\\FileTransfer.ps1 $env:EndMarket $env:Environment"'
I have tried with another powershell script which will connect to Linux server and execute some commands. It works fine in pipeline job
Kindly guide me to fix this issue.
Interesting. Your problem doesn't seem to be in your script because you already explained that it works in a batch job.
I don't know how your pipeline is written, but I would suggest taking a look to Stage, Lock and Milestone which is probably what you need.
The stage step is a primary building block in Pipeline, dividing the
steps of a Pipeline into explicit units and helping to visualize the
progress using the "Stage View" plugin
I guess you could add a stage block like this one in your pipeline:
stage("Previous Step") {
// Some previous step
}
stage("Wait for Script Execution") {
// Call your script
bat 'powershell.exe -ExecutionPolicy Bypass "D:\\Test\\Deploymentscripts\\PowerShellScript\\FileTransfer.ps1 $env:EndMarket $env:Environment"'
}
stage("Next Step") {
// Script already finished its execution
}
But without your pipeline info is just a guessing. Also to improve your script compatibility avoiding "bat and ExcutionPolicy" and using the PowerShell plugin, with that plugin you could simplify your code like this:
powershell -File your_script.ps1
EDIT: I forgot to mention that you can try a different alternative to powershell and the winscp lib using "scp" direct compatibility between Windows and Linux, I'm talking about Cygwin.
With Cygwin installed (with scp) you can use scp as it was a Linux box and a bash script instead powershell:
D:/cygwin-64/bin/run.exe /usr/bin/bash -lic \"/home/user/file.sh\"
In this case I'm running a script silently through Cygwin within a Jenkins Project with a "Run Windows Batch" option. In the script you can add shell commands and the scp instructions you want.
It may seems a little bit more complex but adds more flexibility to perform Windows - Linux tasks.
I've detailed more examples in my blog, you may find it useful.
As avvi mentioned you should check if that variables are being loaded.
You are not calling the script correctly. You are passing in $Env in the powershell invocation which is a powershell variable.
You are in batch mode so should be passing in %.
bat 'powershell.exe -ExecutionPolicy Bypass "D:\\Test\\Deploymentscripts\\PowerShellScript\\FileTransfer.ps1" "%EndMarket%" "%Environment%"'
I am using Jenkins 2.46.1
I have a build pipeline plugin installed and I want to execute a Windows batch file in it. The batch file should execute in a new command window and not on jenkins console output. I give the below Jenkins pipeline groovy script:
node {
stage 'Init'
bat '''
call C:\\myprj\\mybat.bat stop
EXIT /B 0
'''
stage 'Deploy'
bat '''call C:\\myprj\\mybat.bat'''
}
In the init stage, I want to kill the process if it is already open and in stage deploy it should open a new command window and run my batch file.
The problem is that the above does not work. The build is succesful but no command window opens up. Pls suggest
Technically, to do what you're asking you should be able to run
bat 'start cmd.exe /c C:\\myprj\\mybat.bat'
This will launch a new command windows (cmd.exe) and run the batch file given. Depending how your Jenkins slave is running you may not see anything. (eg if it's running as a windows service or different user, you won't see anything)
Alternative solution if the agent is running in a service and you would like to get output:
bat(readFile("mybat.bat"))
Note: The bat file will need to be in your workspace.
Additional Note: You are no longer running the bat file from its original location. Instead it is being run from a temp location created by the underlying durable task system. This means things like %~dp0 in your script are not going to return the paths you might expect.
I have a Windows batch script that I use to build a module and the script in turn uses the ClearCase clearmake command to drive the actual compilations, directory creations and file manipulations, i.e. process the Makefile content. The batch script works flawlessly when invoked using a DOS window or from a "cmd /c ..." command line invocation. And it has been that way for some number of years.
I recently decided to move the script to Ant. The first step, out of simplicity, was to simply invoke the script unchanged using an Exec task (using cmd /c). Almost immediately, Ant fails while creating a directory. The error message reports something like:
mkdir: Cannot create the directory C:\\fred\\harry\\joe
I was able to verify that, using the DOS command prompt, the mkdir C:\\fred\\harry\\joe command works fine, so, as near as I can tell so far, Ant generating double backslash path separators combined with something inherent to clearmake and/or something in the Makefile is causing the failure.
The response I'm looking for is something along the lines: "Yes, clearmake is definitely the culprit because..." or "If you twiddle this thing or that thing in Ant, only a single backslash will be generated...". Should there be no simple and quick explanation, I will drill into the problem to determine what exactly is causing the failure.
Thanks,
I have seen similar error with:
dynamic views (more sensible to ownership than a snapshot view on C:\, which is your case)
resource handle conflict (the script tries to update a resource already taken by another process, which shouldn't be the case here with your script, since it was working outside of Ant Exec task)
error message (like you create a directory which already exists: the error get ignored in a classic script, while it could interrupt the ant task.
While the last cause is a good candidate, try first to simplify your script (leave only the mkdir for instance) in order to check that this line is indeed the issue (nd not "this line in conjunction with others actions taking place just before")
Folks, I'm using git tools such as git bisect run which need to call a command to build and test my project. My command to do is nant which is a windows program. Or a build.cmd script which calls nant.
It's easy to get the bash to call the nant build to run.
But the hard part is how to get the standard output written to a file?
I even installed the Windows PowerShell to try running a command from bash.
Again, it works but the standard output fill says "permission denied" when
I try to read it while the build is going on.
Update:
When running nant, the entire path is used. It is installed and runs fine. The problem is how to get the standard output when running from bash.
If running nant from the windows prompt with "> build.out" at the end of the line, you will get the standard out. But the same never works under bash. It just says build.out is locked, permissions denied.
Update:
Using tee as suggested below also doesn't solve the problem. In that case the file still report "access is denied" with any attempts to read it while the build runs. But also, the tee program never writes anything to standard output.
(If I am understanding your question correctly...)
You can probably use the 'tee' command to split the output to both a file and stdout. The line echo I am building something | tee build.out will both print the output on the console and save it to the file named "build.out".
The tee command is usually available in Cygwin, and also apparently in the Bash shell installed by msysgit (where I just tested it). Here's a good reference page for more details.
Okay, finally resolved this. It turns out the the nant build script was using git feature to erase all ignored files to cleanup. That was deleting the build.out file can causing these strange issues. Now, the process writes the build.out file to a parent directory so that it won't get deleted and now everything works smoothly as expected.