Re: Docs for JENKINS_JAVA_OPTIONS

[Les Mikesell]

On Mon, Mar 5, 2012 at 11:44 PM, Frank Merrow  wrote:
>>I was referring to specifically to the way jenkins works, not about whether
> or not it is fun to write.
>>I don't see a lot of other languages that offer to self-install and execute
> across all the platforms that
>>run Jenkins as master or slaves.
>
> And I was pointing out that the IDE support for Java was weak compared to
> the Microsoft offerings . . .

IDE's don't matter if the resulting code won't run on the platform
where you want it.

> I was hoping you'd pipe up and tell me I was wrong "use Xyz" . . . was
> Eclipse that recommendation?
>
> The way you phrased it, it didn't sound like you even used it.

I'm more of a system administrator that uses perl as much as possible
to avoid dealing with compiler quirks at all, but the java developers
here do use eclipse and I would recommend it - just not from personal
experience.  The way it uses plugins makes it sort of a toolbox that
may need some custom setup for exactly the feature set you need,
though.  If you have looked at it and thought it was missing
something, you may have just needed a different setup.

-- 
   Les Mikesell
 lesmikes...@gmail.com

RE: Docs for JENKINS_JAVA_OPTIONS

[Frank Merrow]

>I was referring to specifically to the way jenkins works, not about whether
or not it is fun to write.
>I don't see a lot of other languages that offer to self-install and execute
across all the platforms that
>run Jenkins as master or slaves.

And I was pointing out that the IDE support for Java was weak compared to
the Microsoft offerings . . .

I was hoping you'd pipe up and tell me I was wrong "use Xyz" . . . was
Eclipse that recommendation?

The way you phrased it, it didn't sound like you even used it.

Frank

Re: Redundant syncs by Perforce plugin when using matrix jobs

[Rob Petti]

Sorry, apparently Google is REALLY slow at indexing groups, so I haven't 
yet received a notification for this post.

This is a problem with core, like many others have described. The main 
build will always sync first, and there's really no way (that I know of) to 
get around it.

I suggest you change your client spec name to "Jenkins_${JOB_NAME}" and 
turn off force syncing to make things move a bit faster.

On Thursday, 1 March 2012 04:34:21 UTC-7, Thomas Fields wrote:
>
> Hi there,
>
> I've been using the Perforce plugin with Jenkins for a while now and it's 
> been working great. However, recently I noticed that it's been doing what I 
> think is a redundant sync.
>
> I've got a matrix job and when the job triggers, the matrix parent gets a 
> list of the changes by calling "p4 describe" but I don't see why it 
> actually needs to sync the entire view.
>
> Here's a snippet of my console output for the matrix parent job:
>
> [LevelEditor] $ "C:\\Program Files\\Perforce\\p4.exe" counter change
> [LevelEditor] $ "C:\\Program Files\\Perforce\\p4.exe" -s changes -s submitted 
> //Jenkins_LevelEditor-Build1/...@167107,@167107
> [LevelEditor] $ "C:\\Program Files\\Perforce\\p4.exe" describe -s 167107
> [LevelEditor] $ "C:\\Program Files\\Perforce\\p4.exe" -G where //...
> Sync'ing workspace to changelist 167107 (forcing sync of unchanged files).
>
> [LevelEditor] $ "C:\\Program Files\\Perforce\\p4.exe" -s sync -f 
> //Jenkins_LevelEditor-Build1/...@167107
>
>
> Then each sub-element of the matrix does this:
>
> [x86] $ "C:\\Program Files\\Perforce\\p4.exe" counter change
> This is a matrix run, trying to use change number from parent/siblings...
> Latest change from parent is: 167107**[x86] $ "C:\\Program 
> Files\\Perforce\\p4.exe" -s changes -s submitted 
> //Jenkins_LevelEditor-CONFIG-Debug-TARGET-x86-Build2/...@167107,@167107**[x86]
>  $ "C:\\Program Files\\Perforce\\p4.exe" describe -s 167107**[x86] $ 
> "C:\\Program Files\\Perforce\\p4.exe" -G where //...**Sync'ing workspace to 
> changelist 167107 (forcing sync of unchanged files).**[x86] $ "C:\\Program 
> Files\\Perforce\\p4.exe" -s sync -f 
> //Jenkins_LevelEditor-CONFIG-Debug-TARGET-x86-Build2/...@167107
>
> Note the same sync happens again. 
>
> Is this a bug in the Perforce plugin? Is there anyway to remove that first 
> sync? As you can imagine, if your view is quite large, the sync could take a 
> while and slow down the build.
>
> Regards,
> Tom.
>

Re: Parameterized build not working - plz help

[Andrew Melo]

Hi there,

How are you submitting the job?

-andrew

On Mon, Mar 5, 2012 at 8:15 PM, Jatin D Patel  wrote:

> Hi all,
>
> I have set up a string value
> Name : VERSION
> Default Value : 17
>
> Ref: http://wiki.hudson-ci.org/display/HUDSON/Parameterized+Build
>
> When i try to access, the version value, i get a blank response
>
> echo $JAVA_HOME
> returns Java Home location
>
> *echo $VERSION*
> *returns a blank line*
> *
> *
> *echo "$VERSION"*
> *also returns a blank line*
>
>
> How can i display the value 17 , while *executing shell script on remot
> ehost using ssh?*
> *
> *
> Please help.
>
> --
> Thanks,
> JD Patel
> You Got To Think HIGH to RISE.
>
>
>


-- 
--
Andrew Melo

Parameterized build not working - plz help

[Jatin D Patel]

Hi all,

I have set up a string value
Name : VERSION
Default Value : 17

Ref: http://wiki.hudson-ci.org/display/HUDSON/Parameterized+Build

When i try to access, the version value, i get a blank response

echo $JAVA_HOME
returns Java Home location

*echo $VERSION*
*returns a blank line*
*
*
*echo "$VERSION"*
*also returns a blank line*


How can i display the value 17 , while *executing shell script on remot
ehost using ssh?*
*
*
Please help.

-- 
Thanks,
JD Patel
You Got To Think HIGH to RISE.

Re: Upgraded to Jenkins, one Maven project fails to build with java.lang.NoSuchMethodError: org.apache.maven.project.MavenProject.getProjectBuildingRequest()

[TJ Rothwell]

Try updating the version of maven for the CI build.  I was using the
pre-configured one labeled "Maven3.0" and changed to "Maven3.0.4" which
resolved the issue for me.

-- TJ

On Mon, Mar 5, 2012 at 5:43 PM, Collin Peters wrote:

> No - I haven't been able to resolve it.  I even got the error trying to
> build a completely unrelated Maven project.  I have absolutely no idea what
> causes this
>
>
> On Mon, Mar 5, 2012 at 2:50 PM, TJ Rothwell  wrote:
>
>> Collin,
>>
>> Were you able to resolve this?
>>
>> -- TJ
>>
>>
>> On Friday, January 13, 2012 6:27:29 PM UTC-6, Nick wrote:
>>>
>>> Has anyone found a solution to this? I'm running into a very similar
>>> problem. Thank you! :)
>>
>>
>

Re: Upgraded to Jenkins, one Maven project fails to build with java.lang.NoSuchMethodError: org.apache.maven.project.MavenProject.getProjectBuildingRequest()

[Collin Peters]

No - I haven't been able to resolve it.  I even got the error trying to
build a completely unrelated Maven project.  I have absolutely no idea what
causes this

On Mon, Mar 5, 2012 at 2:50 PM, TJ Rothwell  wrote:

> Collin,
>
> Were you able to resolve this?
>
> -- TJ
>
>
> On Friday, January 13, 2012 6:27:29 PM UTC-6, Nick wrote:
>>
>> Has anyone found a solution to this? I'm running into a very similar
>> problem. Thank you! :)
>
>

Re: Upgraded to Jenkins, one Maven project fails to build with java.lang.NoSuchMethodError: org.apache.maven.project.MavenProject.getProjectBuildingRequest()

[TJ Rothwell]

Collin,

Were you able to resolve this?

-- TJ

On Friday, January 13, 2012 6:27:29 PM UTC-6, Nick wrote:
>
> Has anyone found a solution to this? I'm running into a very similar 
> problem. Thank you! :)

Re: "Your branch is behind 'origin/master' by 1 commit, and can be fast-forwarded"

[Sami Tikka]

So, I guess you could just do in your build step:

git checkout master
git merge origin/master
# do your thing
git commit
git push origin master

And then start thinking how to change your build process so that you
no longer need to make commits during the build... :)

-- Sami

2012/3/5 Matthias :
> Hi,
>
> I'm running into a problem with Git and Jenkins here and have no idea why
> that is. The thing is, we don't want to work in detached HEAD mode, since
> part of our build is keeping track of a ProGuard mapping file that is
> updated only ever during the build itself. This means we have to build on
> master, not detached, so that we can create a commit and push it.
>
> However, regardless what I do, Jenkins refuses to merge changes from origin
> before it executes the build. This seems very strange; why would it work on
> a stale local branch instead of the latest one? I even tried going to
> "Advanced" settings of the Git plugin and filled in "Merge before build"
> (with "origin" and "master"), although I believe this is only meant to be
> used for integration branches? Anyway, it has no effect either:
>
> Commencing build of Revision aa194e067103efa1ba23bc0a3a767bbb4846fca4
> (origin/master)
> Merging Revision aa194e067103efa1ba23bc0a3a767bbb4846fca4 (origin/master)
> onto $BRANCH
> Seen branch in repository origin/HEAD
> Seen branch in repository origin/acceptance_tests
> Seen branch in repository origin/acra_custom_error_notifications
> Seen branch in repository origin/actionbar
> Seen branch in repository origin/enhanced_upload_notification
> Seen branch in repository origin/ignited_lm
> Seen branch in repository origin/master
> Seen branch in repository origin/mwc
> Seen branch in repository origin/remove_old_rpx
> Seen branch in repository origin/robo_tests
> Seen branch in repository origin/vf2
> Seen branch in repository origin/vodafone
> [workspace] $ /bin/bash -x /tmp/hudson4018151321281891494.sh
> + git checkout master
> Previous HEAD position was aa194e0... Fix pro guard to not crash on 1.6
> devices on task errors due to NFC.
> Switched to branch 'master'
> Your branch is behind 'origin/master' by 1 commit, and can be
> fast-forwarded.
> Parsing POMs
>
> ...
>
>
>
> As you can see, the merge never happens.
>
> How can I tell the Git plugin to merge origin/master before performing the
> build? ^^
>
> Thanks,
> Matthias

RE: Passing information from one build to another

[John Arrizza]

Have your post build step write the build number etc to a file. The other 
builds read the file. Just use echo abc >> somefile.txt



 Original Message 
 From: Kayla Lewis 
 Sent: Mon, Mar 5, 2012 13:08
 To: jenkinsci-users@googlegroups.com
 CC: 
 Subject: Passing information from one build to another

>I want to have a post build action that's triggered only when the build 
>succeeds. The build number will need to be passed in to the post build script.
>Does anyone have suggestions on how to do this or what plugin to use?
>

Passing information from one build to another

[Kayla Lewis]

I want to have a post build action that's triggered only when the build 
succeeds. The build number will need to be passed in to the post build script.
Does anyone have suggestions on how to do this or what plugin to use?

Jenkins Selenium Plugin - License question

[Greg Hinson]

I have a question about the Selenium Plugin for Jenkins.  
https://wiki.jenkins-ci.org/display/JENKINS/Selenium+Plugin

When I looked at the source for the plugin on github (
https://github.com/jenkinsci/selenium-plugin) I didn’t see any license file 
associated with the plugin’s code.

Is the Selenium Plugin for Jenkins licensed under the same license as 
Jenkins?  If not, what is the license for the Selenium Plugin?

Thanks for any help on this. 

Re: Slow job configuration launch

[tah]

Using Firefox, so that's not the problem. Thanks for the suggestion
though!

K96 wrote:
> Hi,
>
> - Don't use IE8 browser
> - Disable unused plugins
>
> Regards,
> @kuhcrow
>
>
> On Mar 2, 2:22 pm, Didier Durand  wrote:
> > Hi,
> >
> > Do you keep a long history of builds for your jobs ? It may help if
> > you keep it smaller.
> >
> > regards
> >
> > didier
> >
> > On Mar 2, 6:12 am, tah  wrote:
> >
> >
> >
> > > Hello,
> > >         My Jenkins instance is generally fast, but launching any job
> > > configuration page is extremely slow. To be clear: Log in to Jenkins,
> > > nice and fast. Hit job link, nice and fast. Hit the 'Configure' link
> > > for that job, very very slow to bring up that page. I tried playing
> > > with the java heap settings, and no difference. Any pointers?
> >
> > > Thanks for any help!

Re: Slow job configuration launch

[tah]

No, the slowness happens with a brand new job with no history. In
fact, it happened right at initial install.

Didier Durand wrote:
> Hi,
>
> Do you keep a long history of builds for your jobs ? It may help if
> you keep it smaller.
>
> regards
>
> didier
>
> On Mar 2, 6:12 am, tah  wrote:
> > Hello,
> >         My Jenkins instance is generally fast, but launching any job
> > configuration page is extremely slow. To be clear: Log in to Jenkins,
> > nice and fast. Hit job link, nice and fast. Hit the 'Configure' link
> > for that job, very very slow to bring up that page. I tried playing
> > with the java heap settings, and no difference. Any pointers?
> >
> > Thanks for any help!

RE: Specifying files in warnings plugin

[Matthew.Webber]

Hi Ulli,
Thank you for your hint. “.*/materialize_workspace_git/scisoft/.*\.java” works 
for me.
Originally,  I was missing the “.*/” at the beginning of the regex.
Matthew

From: jenkinsci-users@googlegroups.com 
[mailto:jenkinsci-users@googlegroups.com] On Behalf Of Ullrich Hafner
Sent: 05 March 2012 16:11
To: jenkinsci-users@googlegroups.com
Subject: Re: Specifying files in warnings plugin

On 03/05/2012 04:08 PM, 
matthew.web...@diamond.ac.uk wrote:

We've just started using the warnings plugin, scanning the console log for 
Buckminster compiler warnings.



I want to only collect warnings for selected source files, so I clicked 
"Advanced" which offers options "Warnings to include" and "Warnings to ignore". 
I tried to specify a file path here (with wildcards), but that didn't seem to 
work. The help text is unclear, but it seems to suggest that only a filename, 
but not a file path, can be entered here.

Is there a way to only collect warnings from certain files?




What patterns did you specify? Note that the pattern is a regular expression 
and not a shell wildcard expression.

The help text currently is: 'Comma separated list of regular 
expressions
 that specifies the files to exclude from the report (based on their absolute 
filename).'
I'm not a native speaker, so I would appreciate any improvements to this help 
text to make it clear for the user :-)

What I wanted to say with that help text is:
If the warning is in a file with absolute path "/path/to/file/file.txt" then 
e.g., the patterns ".*/to/.*" or ".*/file\.txt" will match that file. All files 
that match will be either included or excluded in the reports (depending in 
which field you enter the value).

Ulli





-- 

This e-mail and any attachments may contain confidential, copyright and or 
privileged material, and are for the use of the intended addressee only. If you 
are not the intended addressee or an authorised recipient of the addressee 
please notify us of receipt by returning the e-mail and do not use, copy, 
retain, distribute or disclose the information in or attached to the e-mail.

Any opinions expressed within this e-mail are those of the individual and not 
necessarily of Diamond Light Source Ltd. 

Diamond Light Source Ltd. cannot guarantee that this e-mail or any attachments 
are free from viruses and we cannot accept liability for any damage which you 
may sustain as a result of software viruses which may be transmitted in or with 
the message.

Diamond Light Source Limited (company no. 4375679). Registered in England and 
Wales with its registered office at Diamond House, Harwell Science and 
Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom

 

Re: groovy plugin?

[Les Mikesell]

On Mon, Mar 5, 2012 at 10:19 AM, Vojtech Juranek  wrote:
>
>> I was just missing something obvious.  Along with checking the
>> 'Install automatically' box in the global config (which might have
>> been set by default) you have to give it a name and then you have to
>> pick that name in the job config.   Before getting this right, the job
>> config only offered the 'default' verision, and until I went back to
>> the job and changed it, it didn't trigger the install.  And until I
>> clicked the drop-down on in the job config I don't realize there was a
>> new option there.
>
> thanks for feedback, IIRC this is standard behavior of any auto-installer, but
> I'll check if I missed something somewhere

It makes sense now, but since I didn't understand it at first I set up
a job with the 'default' groovy version which was the only choice, and
it continued to fail after giving the install a name.   It would have
helped to mention that the version names had to match in the ? help
link in the main and job configs where you enter them.   Or maybe even
to make the 'default' choice work with the auto-installed copy if you
don't find one earlier in PATH.   Shouldn't that become the default if
there is no other?

-- 
  Les Mikesell
lesmikes...@gmail.com

Re: Handing off builds between Jenkins masters

[Jacob Robertson]

We are wanting a similar thing.  We want a successful build to kick of
a particular parameterized build on another master.  Right now, we've
prototyped using wget (using "buildWithParameters") from within the
flexible publish plugin so that it can be a post-build action.  It
seems to work fine, but it really feels like a cludge.  If anyone does
end up writing some plugin code for this, I would be happy to test it.

On Sat, Mar 3, 2012 at 11:20 AM, Sami Tikka  wrote:
> I have recently used the ssh API because it allows one to trigger a build
> and wait for its completion, which was useful for one seldom-used task.
>
> -- Sami
>
> danny staple  kirjoitti 3.3.2012 kello 11.50:
>
> We've used the wget/curl methods quite successfully for this. It's worth
> getting to know as it also allows builds to start other builds in ways the
> plugins dont usually cover - although we are taking note of the ssh API.
>
> On 3 March 2012 07:33, domi  wrote:
>>
>> We do this with wget in a shell step
>>
>> $> wget http://someserver.com/job/jobA/build?delay=0sec
>>
>> …just the same way as if you would trigger it from a command line.
>> If you use the conditional-buildstep-plugin you can wrap the execution
>> into conditions too.
>> /Domi
>>
>>
>> On 02.03.2012, at 16:52, Jim McCaskey wrote:
>>
>> > Hello all,
>> >
>> > I've looked around for a bit for something to do what I want, and before
>> > I spend a whole lot of time hacking around, I thought I would ask.  I would
>> > like to be able to have one Jenkins master start a build on another Jenkins
>> > master.  Just like the existing "Build other projects" functionality, but
>> > instead of calling the other project on the local Jenkins Master, call it 
>> > on
>> > a remote Jenkins master.
>> >
>> > Why do I want to do this?  Our Jenkins use is growing internally (a good
>> > thing!) and other groups want to use the system for their own automation
>> > purposes.  These other groups are doing more detailed testing, integration
>> > work, and some system maintenance and the existing Jenkins administrators
>> > don't want to deal with that.  There's also the issue of passwords/security
>> > to deal with between the completely separate groups.  It seems like if 
>> > there
>> > was a way to spawn projects off on another Jenkins master than the other
>> > groups could have their own masters and work them as they see fit.
>> >
>> > I've thought of a few ways to handle this (the Jenkins CLI immediately
>> > comes to mind), but wanted to know if there was a cleaner implementation
>> > that could make this happen.
>> >
>> > Thanks!
>> >
>> > -Jim
>> >
>>
>
>
>
> --
> Danny Staple
>
> Director, ODM Solutions Ltd
> w: http://www.odmsolutions.co.uk
> Blog: http://orionrobots.co.uk/blog1-Danny-Staple
>
>

Re: groovy plugin?

[Vojtech Juranek]

> I was just missing something obvious.  Along with checking the
> 'Install automatically' box in the global config (which might have
> been set by default) you have to give it a name and then you have to
> pick that name in the job config.   Before getting this right, the job
> config only offered the 'default' verision, and until I went back to
> the job and changed it, it didn't trigger the install.  And until I
> clicked the drop-down on in the job config I don't realize there was a
> new option there.

thanks for feedback, IIRC this is standard behavior of any auto-installer, but 
I'll check if I missed something somewhere 

Re: Specifying files in warnings plugin

[Ullrich Hafner]

On 03/05/2012 04:08 PM, matthew.web...@diamond.ac.uk wrote:
> We've just started using the warnings plugin, scanning the console log for 
> Buckminster compiler warnings.
>
> I want to only collect warnings for selected source files, so I clicked 
> "Advanced" which offers options "Warnings to include" and "Warnings to 
> ignore". I tried to specify a file path here (with wildcards), but that 
> didn't seem to work. The help text is unclear, but it seems to suggest that 
> only a filename, but not a file path, can be entered here.
> Is there a way to only collect warnings from certain files?
>
>
What patterns did you specify? Note that the pattern is a regular
expression and not a shell wildcard expression.

The help text currently is: 'Comma separated list of regular expressions

that specifies the files to exclude from the report (based on their
absolute filename).'
I'm not a native speaker, so I would appreciate any improvements to this
help text to make it clear for the user :-)

What I wanted to say with that help text is:
If the warning is in a file with absolute path "/path/to/file/file.txt"
then e.g., the patterns ".*/to/.*" or ".*/file\.txt" will match that
file. All files that match will be either included or excluded in the
reports (depending in which field you enter the value).

Ulli

Re: groovy plugin?

[Les Mikesell]

On Mon, Mar 5, 2012 at 2:56 AM, Vojtech Juranek  wrote:
>> On Thu, Feb 23, 2012 at 5:43 PM, Vojtech Juranek 
> wrote:
>> >> Does the groovy plugin need anything special installed on the slaves?
>> >
>> > there are two options, groovy script and system groovy script. First one
>> > requires groovy installation on slaves (check global jenkins
>> > configuration, auto-install can be used), second doesn't, however it's
>> > not executed on slave, but in master JVM (it uses bundled groovy lib in
>> > the same way as you run groovy script via groovy console)
>>
>> I don't get it.  The script console in the node screen works and runs
>> on the slaves.  For example I can run 'println "ipconfig
>> /all".execute().text and I get results that are obviously coming from
>> the windows slave even though the jenkins server and master jvm are on
>> linux.  Or is that remoted with some java magic?
>
> exactly, slave agent has remote class loader which is able to load all
> necessary classes from master (if you are interested in technical details, see
> [1]). If you consider similar function/build step useful, please file a 
> feature
> JIRA request for me, it shouldn't be very difficult to implement it.
>
>> I checked the auto-install option but it doesn't seem to make any
>> difference.  Should the attempt to install be logged somewhere so I
>> can see if it failed?
>
> installation should be visible in the log (in log of the build which tries to
> install groovy before launching groovy commands). Could you provide some
> details or log a JIRA for it if you think it's a bug?
> Thanks

I was just missing something obvious.  Along with checking the
'Install automatically' box in the global config (which might have
been set by default) you have to give it a name and then you have to
pick that name in the job config.   Before getting this right, the job
config only offered the 'default' verision, and until I went back to
the job and changed it, it didn't trigger the install.  And until I
clicked the drop-down on in the job config I don't realize there was a
new option there.

-- 
  Les Mikesell
 lesmikes...@gmail.com

Removing deleted modules from workspaces

[David Pärsson]

Hi,

We're using Jenkins with multiple jobs running on multiple machines. Most 
jobs uses a SVN as SCM, with multiple modules in the same workspace.

When a module is removed from a job configuration, its files remain in the 
workspaces on our slaves until the workspaces are wiped. This causes some 
problems, for example with Checkstyle warning deltas. As a solution to 
this, we're wiping our workspaces every night, and while this works it 
isn't as fast as we'd like it to be.

Is there a way, by configuration or by using a plug-in, to automatically 
remove the deleted modules from the workspaces? 

Re: Redundant syncs by Perforce plugin when using matrix jobs

[Andrew Melo]

On Mon, Mar 5, 2012 at 9:26 AM, Thomas Fields wrote:

> I'm not sure about the SVN plugin but it's a performance killer for me
> so I'm hoping there is a fix. I've created
> https://issues.jenkins-ci.org/browse/JENKINS-12983 to see if I can get an
> official answer from the author of the Perforce plugin.
>

FWIW, I see the same behavior with git.

-Andrew



>
> On Saturday, 3 March 2012 15:58:16 UTC, Simon Stevenson wrote:
>
>> I think there may be a bug with the matrix builds... I have noticed with
>> SVN the parent does an unnecessary sync as well.
>>
>> On 1 March 2012 21:34, Thomas Fields  wrote:
>>
>>> Hi there,
>>>
>>> I've been using the Perforce plugin with Jenkins for a while now and
>>> it's been working great. However, recently I noticed that it's been doing
>>> what I think is a redundant sync.
>>>
>>> I've got a matrix job and when the job triggers, the matrix parent gets
>>> a list of the changes by calling "p4 describe" but I don't see why it
>>> actually needs to sync the entire view.
>>>
>>> Here's a snippet of my console output for the matrix parent job:
>>>
>>> [LevelEditor] $ "C:\\Program Files\\Perforce\\p4.exe" counter change
>>> [LevelEditor] $ "C:\\Program Files\\Perforce\\p4.exe" -s changes -s 
>>> submitted //Jenkins_LevelEditor-Build1/.**..@167107,@167107
>>> [LevelEditor] $ "C:\\Program Files\\Perforce\\p4.exe" describe -s 167107
>>> [LevelEditor] $ "C:\\Program Files\\Perforce\\p4.exe" -G where //...
>>> Sync'ing workspace to changelist 167107 (forcing sync of unchanged files).
>>>
>>> [LevelEditor] $ "C:\\Program Files\\Perforce\\p4.exe" -s sync -f 
>>> //Jenkins_LevelEditor-Build1/.**..@167107
>>>
>>>
>>> Then each sub-element of the matrix does this:
>>>
>>> [x86] $ "C:\\Program Files\\Perforce\\p4.exe" counter change
>>> This is a matrix run, trying to use change number from parent/siblings...
>>> Latest change from parent is: 167107
>>> [x86] $ "C:\\Program Files\\Perforce\\p4.exe" -s changes -s submitted 
>>> //Jenkins_LevelEditor-CONFIG-**Debug-TARGET-x86-Build2/...@**167107,@167107
>>> [x86] $ "C:\\Program Files\\Perforce\\p4.exe" describe -s 167107
>>> [x86] $ "C:\\Program Files\\Perforce\\p4.exe" -G where //...
>>> Sync'ing workspace to changelist 167107 (forcing sync of unchanged files).
>>> [x86] $ "C:\\Program Files\\Perforce\\p4.exe" -s sync -f 
>>> //Jenkins_LevelEditor-CONFIG-**Debug-TARGET-x86-Build2/...@**167107
>>>
>>> Note the same sync happens again.
>>>
>>> Is this a bug in the Perforce plugin? Is there anyway to remove that first 
>>> sync? As you can imagine, if your view is quite large, the sync could take 
>>> a while and slow down the build.
>>>
>>> Regards,
>>> Tom.
>>>
>>
>>


-- 
--
Andrew Melo

Re: Redundant syncs by Perforce plugin when using matrix jobs

[Thomas Fields]

I'm not sure about the SVN plugin but it's a performance killer for me 
so I'm hoping there is a fix. I've created 
https://issues.jenkins-ci.org/browse/JENKINS-12983 to see if I can get an 
official answer from the author of the Perforce plugin. 
 

On Saturday, 3 March 2012 15:58:16 UTC, Simon Stevenson wrote:

> I think there may be a bug with the matrix builds... I have noticed with 
> SVN the parent does an unnecessary sync as well.
>
> On 1 March 2012 21:34, Thomas Fields  wrote:
>
>> Hi there,
>>
>> I've been using the Perforce plugin with Jenkins for a while now and it's 
>> been working great. However, recently I noticed that it's been doing what I 
>> think is a redundant sync.
>>
>> I've got a matrix job and when the job triggers, the matrix parent gets a 
>> list of the changes by calling "p4 describe" but I don't see why it 
>> actually needs to sync the entire view.
>>
>> Here's a snippet of my console output for the matrix parent job:
>>
>> [LevelEditor] $ "C:\\Program Files\\Perforce\\p4.exe" counter change
>> [LevelEditor] $ "C:\\Program Files\\Perforce\\p4.exe" -s changes -s 
>> submitted //Jenkins_LevelEditor-Build1/...@167107,@167107
>> [LevelEditor] $ "C:\\Program Files\\Perforce\\p4.exe" describe -s 167107
>> [LevelEditor] $ "C:\\Program Files\\Perforce\\p4.exe" -G where //...
>> Sync'ing workspace to changelist 167107 (forcing sync of unchanged files).
>>
>> [LevelEditor] $ "C:\\Program Files\\Perforce\\p4.exe" -s sync -f 
>> //Jenkins_LevelEditor-Build1/...@167107
>>
>>
>> Then each sub-element of the matrix does this:
>>
>> [x86] $ "C:\\Program Files\\Perforce\\p4.exe" counter change
>> This is a matrix run, trying to use change number from parent/siblings...
>> Latest change from parent is: 167107
>> [x86] $ "C:\\Program Files\\Perforce\\p4.exe" -s changes -s submitted 
>> //Jenkins_LevelEditor-CONFIG-Debug-TARGET-x86-Build2/...@167107,@167107
>> [x86] $ "C:\\Program Files\\Perforce\\p4.exe" describe -s 167107
>> [x86] $ "C:\\Program Files\\Perforce\\p4.exe" -G where //...
>> Sync'ing workspace to changelist 167107 (forcing sync of unchanged files).
>> [x86] $ "C:\\Program Files\\Perforce\\p4.exe" -s sync -f 
>> //Jenkins_LevelEditor-CONFIG-Debug-TARGET-x86-Build2/...@167107
>>
>> Note the same sync happens again. 
>>
>> Is this a bug in the Perforce plugin? Is there anyway to remove that first 
>> sync? As you can imagine, if your view is quite large, the sync could take a 
>> while and slow down the build.
>>
>> Regards,
>> Tom.
>>
>
>

Specifying files in warnings plugin

[Matthew.Webber]

We've just started using the warnings plugin, scanning the console log for 
Buckminster compiler warnings.

I want to only collect warnings for selected source files, so I clicked 
"Advanced" which offers options "Warnings to include" and "Warnings to ignore". 
I tried to specify a file path here (with wildcards), but that didn't seem to 
work. The help text is unclear, but it seems to suggest that only a filename, 
but not a file path, can be entered here.

Is there a way to only collect warnings from certain files?


--
This e-mail and any attachments may contain confidential, copyright and or 
privileged material, and are for the use of the intended addressee only. If you 
are not the intended addressee or an authorised recipient of the addressee 
please notify us of receipt by returning the e-mail and do not use, copy, 
retain, distribute or disclose the information in or attached to the e-mail.
Any opinions expressed within this e-mail are those of the individual and not 
necessarily of Diamond Light Source Ltd.
Diamond Light Source Ltd. cannot guarantee that this e-mail or any attachments 
are free from viruses and we cannot accept liability for any damage which you 
may sustain as a result of software viruses which may be transmitted in or with 
the message.
Diamond Light Source Limited (company no. 4375679). Registered in England and 
Wales with its registered office at Diamond House, Harwell Science and 
Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom

Sonar tests fail bevause xvnc plugin was terminated before execution

[MSF]

Hi,



we have some issues with one of our jobs on our Jenkins (1.424.3 = current LTS 
Version). We are using xvnc (1.10) and Sonar (1.8) plugins. The job fails 
because the generated xvnc session is terminated before Sonar is started. We 
tried different versions of the plugins, but the result is always the same. We 
even tried to recompile the xvnc plugin from source with a different ordinal 
number like:



@Extension(ordinal = 1)

   public static final class DescriptorImpl extends BuildWrapperDescriptor

or:

@Extension(ordinal = 1)

   public static final class DescriptorImpl extends BuildWrapperDescriptor

but this did also change nothing in the result.



Here is the error with stack trace:

[INFO] Nothing to compile - all classes are up to date
[INFO] [surefire:test {execution: default-test}]
[INFO] Surefire report directory: 
/home/jenkins/.jenkins/jobs/JOBNAME-Sonar/workspace/JOBNAME/JOBMODULE-tile/target/surefire-reports
org.apache.maven.surefire.booter.SurefireExecutionException: Can't connect to 
X11 window server using ':14' as the value of the DISPLAY variable.; nested 
exception is java.lang.InternalError: Can't connect to X11 window server using 
':14' as the value of the DISPLAY variable.
java.lang.InternalError: Can't connect to X11 window server using ':14' as the 
value of the DISPLAY variable.
 at sun.awt.X11GraphicsEnvironment.initDisplay(Native Method)
 at sun.awt.X11GraphicsEnvironment.access$100(X11GraphicsEnvironment.java:52)
...

[ERROR] There are test failures.



We're out of ideas right now, but it's quite important to get this running. Is 
there any way to have the xvnc plugin with a X session still running when Sonar 
is executed?

Any help would be greatly appreciated!



Thanks in advance.



BR,

Manuel

"Your branch is behind 'origin/master' by 1 commit, and can be fast-forwarded"

[Matthias]

Hi,

I'm running into a problem with Git and Jenkins here and have no idea why 
that is. The thing is, we don't want to work in detached HEAD mode, since 
part of our build is keeping track of a ProGuard mapping file that is 
updated only ever during the build itself. This means we have to build on 
master, not detached, so that we can create a commit and push it.

However, regardless what I do, Jenkins refuses to merge changes from origin 
before it executes the build. This seems very strange; why would it work on 
a stale local branch instead of the latest one? I even tried going to 
"Advanced" settings of the Git plugin and filled in "Merge before build" 
(with "origin" and "master"), although I believe this is only meant to be 
used for integration branches? Anyway, it has no effect either:

Commencing build of Revision aa194e067103efa1ba23bc0a3a767bbb4846fca4 
(origin/master)
Merging Revision aa194e067103efa1ba23bc0a3a767bbb4846fca4 (origin/master) onto 
$BRANCH
Seen branch in repository origin/HEAD
Seen branch in repository origin/acceptance_tests
Seen branch in repository origin/acra_custom_error_notifications
Seen branch in repository origin/actionbar
Seen branch in repository origin/enhanced_upload_notification
Seen branch in repository origin/ignited_lm
Seen branch in repository origin/master
Seen branch in repository origin/mwc
Seen branch in repository origin/remove_old_rpx
Seen branch in repository origin/robo_tests
Seen branch in repository origin/vf2
Seen branch in repository origin/vodafone
[workspace] $ /bin/bash -x /tmp/hudson4018151321281891494.sh
+ git checkout master
Previous HEAD position was aa194e0... Fix pro guard to not crash on 1.6 devices 
on task errors due to NFC.
Switched to branch 'master'
Your branch is behind 'origin/master' by 1 commit, and can be fast-forwarded.
Parsing POMs

...



As you can see, the merge never happens.

How can I tell the Git plugin to merge origin/master before performing the 
build? ^^

Thanks,
Matthias

junit stylesheet transformation before jenkins collects junit result

[andy w]

Hi,

i was wondering if it is possible to perform a custom stylesheet
transformation against  the junit report files before jenkins collects
the junit result? Or are there other ways to modify the junit report
xml before the report is collected by jenkins. we are using maven and
the surefire plugin for the build. Problem is that jenkins immediately
collects the result, before i have a chance to perform my stylesheet
transformation (even if the transformation happens within the same
maven goal).

background: we are running non parameterized tests, however we want to
print some information in the name of the testcase so it actually
looks like a parameterized test. the parameters are dynamic, this is
why we cannot write the unit tests as parameterized. for easier
handling we want to put the parameters on which the test fails in the
name of the test:

e.g.
...

...

The parameters marketCountry and vehicle id are added via a stylesheet
transformation after the test has finished (information is gathered
from the stacktrace).

Any ideas how i can perform the transformation before jenkins collects
the results.?

thanks in advance,
Cheers,
Andy

Strange behavior for Cobertura-plugin

[Asmann, Roland]

Hi all,

I recently added 2 new projects to our Jenkins and the cobertura-plugin 
is behaving oddly on these projects.
On both projects, I see that there is actually code-coverage being 
registered, but as soon as I go into the detail-page, I get:
- one project that says that the source-code is not available
- the other project shows all lines as not having coverage

The projects are similar to all other projects I have on Jenkins (which 
work). The only difference is the fact that the other projects are a 
little older and we had some updates to both Jenkins and plug-ins lately.

Does anybody have any idea what I could do to get Jenkins to correctly 
show my code-coverage?

-- 
Roland Asmann
Senior Software Engineer

adesso Austria GmbH
Floridotower 26. Stock  T +43 1 2198790-27
Floridsdorfer Hauptstr. 1   F +43 1 2198790-927
A-1210 Wien M +43 664 88657566
E roland.asm...@adesso.at
W www.adesso.at

-
 >>> business. people. technology. <<<
-

Re: groovy plugin?

[Vojtech Juranek]

> On Thu, Feb 23, 2012 at 5:43 PM, Vojtech Juranek  
wrote:
> >> Does the groovy plugin need anything special installed on the slaves?
> > 
> > there are two options, groovy script and system groovy script. First one
> > requires groovy installation on slaves (check global jenkins
> > configuration, auto-install can be used), second doesn't, however it's
> > not executed on slave, but in master JVM (it uses bundled groovy lib in
> > the same way as you run groovy script via groovy console)
> 
> I don't get it.  The script console in the node screen works and runs
> on the slaves.  For example I can run 'println "ipconfig
> /all".execute().text and I get results that are obviously coming from
> the windows slave even though the jenkins server and master jvm are on
> linux.  Or is that remoted with some java magic?

exactly, slave agent has remote class loader which is able to load all 
necessary classes from master (if you are interested in technical details, see 
[1]). If you consider similar function/build step useful, please file a feature 
JIRA request for me, it shouldn't be very difficult to implement it.

> I checked the auto-install option but it doesn't seem to make any
> difference.  Should the attempt to install be logged somewhere so I
> can see if it failed?

installation should be visible in the log (in log of the build which tries to 
install groovy before launching groovy commands). Could you provide some 
details or log a JIRA for it if you think it's a bug?
Thanks


[1] 
https://github.com/jenkinsci/jenkins/blob/master/core/src/main/java/hudson/util/RemotingDiagnostics.java#L115