Re: Run a build concurrently on all slaves?

[Vincent Latombe]

Note that matrix job would work but you would need to maintain a static
list of nodes. Also, if one node gets stuck (or offline) it would block
further executions because the 'master' build would never end.  My solution
is more flexible, because it only relies on the labels declared on your
nodes, also have f one node gets unavailable, the corresponding run just
stays in the queue but doesn't block executions on other nodes.

Vincent
Le 12 janv. 2013 08:55, "Vincent Latombe"  a
écrit :

> I don't think matrix build would fit your need.
>
> Here is the pattern I applied successfully to run this kind of job.
>
> 1. Install node label parameter plugin and trigger parametrized job plugin
> 2. Define a job to do the work on one node. Define a node parameter on
> this job, this will allow to decide on which node it will run dynamically.
> Enable concurrent builds for this job
> 3. Define a second job containing a single build step calling your first
> job. Add a parameter factory and provide a label expression that covers all
> the nodes that you need. Make sure your parameter name matches the name of
> the parameter in your first job.
>
> When launching your aggregator job it will actually loop on all nodes
> matching your label expression and launch a run of your executor job for
> each of your modes.
>
> Voila!
>
> Vincent
> Le 11 janv. 2013 22:28,  a écrit :
>
>> On Fri, 11 Jan 2013 21:38:04 +0100
>> felix schwitzer  wrote:
>> > For this we use simply a multi-configuration project.
>>
>> On Fri, 11 Jan 2013 19:35:42 -
>> "KEVIN FLEMING (BLOOMBERG/ 731 LEXIN)"  wrote:
>> > This is exactly what Matrix builds are designed for. You define one
>> > job, with some number of axes in the matrix, and then corresponding
>> > labels on your nodes (slaves) to execute the spawned jobs.
>>
>> OK, thanks.
>>
>> I guess I'm still not entirely clear on the necessary steps, having read
>> as much documentation as I could find on multi-configuration projects.
>>
>> I started by labelling each node with relevant labels ("solaris",
>> "linux", "x86_64", etc) and gave each node a "cluster" label. Then,
>> I created a multi-configuration project and then added a "Label
>> expression" axis, where the name is "cluster" and the label expression
>> is also "cluster" (I realize those two aren't related).
>>
>> Then, I added a build step to build top-level Maven targets, and
>> a post-build step to aggregate test results at the end. The Maven goal
>> was set to "verify" (in order to build the project, run the unit tests,
>> produce a package, and verify the package).
>>
>> Then, I brought three nodes online and triggered a build. Two of the
>> nodes began building. The other node stayed idle.
>>
>> One of the two nodes finished much faster than the other, which looked
>> rather suspicious. Checking the console log for that build showed that
>> it simply checked out the source code from version control and declared
>> success (whilst the other node checked out the source code and actually
>> built it):
>>
>> Triggering cluster
>> cluster completed with result SUCCESS
>> Finished: SUCCESS
>>
>> Am I missing something incredibly obvious here?
>>
>> I'd obviously like *all* of the labelled nodes to run, and to actually
>> build code as opposed to ... not.
>>
>> Regards,
>> M
>>
>

Re: Run a build concurrently on all slaves?

[Vincent Latombe]

I don't think matrix build would fit your need.

Here is the pattern I applied successfully to run this kind of job.

1. Install node label parameter plugin and trigger parametrized job plugin
2. Define a job to do the work on one node. Define a node parameter on this
job, this will allow to decide on which node it will run dynamically.
Enable concurrent builds for this job
3. Define a second job containing a single build step calling your first
job. Add a parameter factory and provide a label expression that covers all
the nodes that you need. Make sure your parameter name matches the name of
the parameter in your first job.

When launching your aggregator job it will actually loop on all nodes
matching your label expression and launch a run of your executor job for
each of your modes.

Voila!

Vincent
Le 11 janv. 2013 22:28,  a écrit :

> On Fri, 11 Jan 2013 21:38:04 +0100
> felix schwitzer  wrote:
> > For this we use simply a multi-configuration project.
>
> On Fri, 11 Jan 2013 19:35:42 -
> "KEVIN FLEMING (BLOOMBERG/ 731 LEXIN)"  wrote:
> > This is exactly what Matrix builds are designed for. You define one
> > job, with some number of axes in the matrix, and then corresponding
> > labels on your nodes (slaves) to execute the spawned jobs.
>
> OK, thanks.
>
> I guess I'm still not entirely clear on the necessary steps, having read
> as much documentation as I could find on multi-configuration projects.
>
> I started by labelling each node with relevant labels ("solaris",
> "linux", "x86_64", etc) and gave each node a "cluster" label. Then,
> I created a multi-configuration project and then added a "Label
> expression" axis, where the name is "cluster" and the label expression
> is also "cluster" (I realize those two aren't related).
>
> Then, I added a build step to build top-level Maven targets, and
> a post-build step to aggregate test results at the end. The Maven goal
> was set to "verify" (in order to build the project, run the unit tests,
> produce a package, and verify the package).
>
> Then, I brought three nodes online and triggered a build. Two of the
> nodes began building. The other node stayed idle.
>
> One of the two nodes finished much faster than the other, which looked
> rather suspicious. Checking the console log for that build showed that
> it simply checked out the source code from version control and declared
> success (whilst the other node checked out the source code and actually
> built it):
>
> Triggering cluster
> cluster completed with result SUCCESS
> Finished: SUCCESS
>
> Am I missing something incredibly obvious here?
>
> I'd obviously like *all* of the labelled nodes to run, and to actually
> build code as opposed to ... not.
>
> Regards,
> M
>

Re: getting Exception in Job Import plugin

[Radha Mohan Maheshwari]

i am using Jenkins ver. 1.498 
and I tried with Jenkins ver. 1.474  also.


On Fri, Jan 11, 2013 at 7:41 PM, Clemens wrote:

> Hi, which Jenkins Version are you using ?
>



-- 
Radha Mohan Maheshwari

Jenkins console gives "Connection refused" when using :ext: for CVSROOT and CVS_RSH=ssh

[Cynsa Bonorris]

I'm trying to switch from :pserver: to :ext: and use ssh for secure 
checkouts. I've set up my keys and am connecting just fine with no password 
to the cvs server as my Jenkins/cvs user (builder).

However, Jenkins is giving me the following error on checkout, though when 
I paste the cvs command into a terminal as builder, the checkout completes 
just fine.

Started by an SCM change
Building on master in workspace /var/lib/jenkins/workspace/foo-co-test
[foo-co-test] $ cvs -Q -z3 -d :ext:builder:/cvsrepo/bmi co -P -r 
foo-branch -f configuration 
XX: Connection refused
cvs [checkout aborted]: end of file from server (consult above messages if any)
FATAL: CVS failed. exit code=1


(I've obscured our cvs server info in the above.) 


in my Jenkins master Global Properties I have the envar CVS_RSH with the value 
/usr/bin/ssh

but when I use groovyscript to display the envars, I don't see CVS_RSH 
populated:


TERM=xterm-256color
SHELL=/bin/bash
SHLVL=3
XFILESEARCHPATH=/usr/dt/app-defaults/%L/Dt
PATH=/sbin:/usr/sbin:/bin:/usr/bin
USER=builder
PWD=/
HOME=/build/buildsystem
LOGNAME=builder
_=/usr/bin/java
NLSPATH=/usr/dt/lib/nls/msg/%L/%N.cat
LD_LIBRARY_PATH=/usr/java/jdk1.6.0_27/jre/lib/i386/server:/usr/java/jdk1.6.0_27/jre/lib/i386:/usr/java/jdk1.6.0_27/jre/../lib/i386
LANG=en_US.UTF-8
Result: {TERM=xterm-256color, SHELL=/bin/bash, SHLVL=3, 
XFILESEARCHPATH=/usr/dt/app-defaults/%L/Dt, PATH=/sbin:/usr/sbin:/bin:/usr/bin, 
USER=builder, PWD=/, HOME=/build/buildsystem, LOGNAME=builder, _=/usr/bin/java, 
NLSPATH=/usr/dt/lib/nls/msg/%L/%N.cat, 
LD_LIBRARY_PATH=/usr/java/jdk1.6.0_27/jre/lib/i386/server:/usr/java/jdk1.6.0_27/jre/lib/i386:/usr/java/jdk1.6.0_27/jre/../lib/i386,
 LANG=en_US.UTF-8}


any assistance would be appreciated. how do I force Jenkins to export the 
CVS_RSH envar? I think that might be the problem, if groovyscript output is to 
be believed.


-cynsa

Git plugin and checking out from a specified point in time

[anthcourtney]

Hello all,

Does anyone have any pointers on how to configure a job within Jenkins (and 
probably more specifically, the Git plugin), so that the git checkout 
corresponds to a point in time. This point in time will be provided as a 
parameter to the build.

Googling around, I have seen many references to using either:

$ git checkout 'master@{2012-12-31 23:59'

or

$ git checkout `git rev-list -n 1 --before="2012-12-31 23:59" master`

to checkout a point in time revision, but I can't see a clear way in how to 
then configure the jenkins job or the git plugin.

Do I need to add a step like this as an execute shell step? Would I still 
use the git plugin to checkout master first? How will it work for 
subsequent builds of the same project?

cheers,
Anth

RE: Email-ext plugin. variable substitution in included file ?

[Alex Earl]

Yes, this is expected, the file token supports nested content.

Sent from my Windows Phone
From: chris_willia...@dell.com
Sent: 1/11/2013 3:02 PM
To: jenkinsci-users@googlegroups.com
Subject: Email-ext plugin. variable substitution in included file ?
We are using the "${FILE,path}" token to include the contents of a
file in the workspace to an email being sent by the Email-ext plugin.
We also have a parameter on this particular job (i.e. parameterized build).

It appears that if the file being included uses a variable with the
same name as the job parameter, then the value of the parameter is
substituted for that variable in the sent email. Is this expected
behavior or should this be considered a bug ? We can work around the
problem by changing the parameter name, but that's a hack. I've also
experimented with trying to reset the parameter before the email is
sent, but it doesn't work or I'm doing it wrong.

Example:
1) Job has a parameter named "PASSWD".
2) Ext-email uses the token: ${FILE, path="./foo.sh"}
3) foo.sh contents:
echo ${PASSWD}

4) When the email is sent, the ${PASSWD} variable in foo.sh is
evaluated and the email contains the value that was typed in for the
PASSWD job parameter.

I'll open an issue if this is considered a bug or unexpected. An
additional data point is, that if I'm including a file named "foo.txt"
instead of "foo.sh", then the "${PASSWD}" string is not evaluated and
appears as expected in the email.

Thanks,
Chris

Chris Williams
Dell | PG Release Engineering

Re: Null pointer exception in authenticating against a subversion repository.

[Andrew Bayless]

A quick update, I have found a work around to this issue. 

In the .subversion directory under the Jenkins user's home directory, there 
is the "servers" file which specifies subversion server specific protocol 
parameters. I had previously configured this file with a group name and url 
for my subversion server. 
And in the section for the corresponding group name , I previously 
specified paths to the SSL authority files (PEMs)   and for the SSL client 
certificate file (P12), as was necessary to use SVN from the command line 
for this server. In my attempt to workaround this exception, I added the 
SSL client certificate password to this file, so that I could configure 
Jenkins to checkout from subversion via a shell command. I discovered by 
accident that after configuring it in this way, I was able to checkout from 
subversion using the SVNplugin as normal, which leads me to suspect that 
there must be a bug of some sort associated with the interface at /scm/SubversionSCM/enterCredential

While this workaround is acceptable, I would prefer not to have any client 
certificate passwords in the servers file. Does anyone here know of any 
issues regarding the interface I mentioned above ?

On Wednesday, January 9, 2013 10:03:16 AM UTC-8, Andrew Bayless wrote:
>
> Hello, 
>
> I am having problems authenticating against a local subversion repository 
> using the subversion plug-in. I am using Jenkins 1.497 with the Subversion 
> 1.44 plugin deployed on Winstone, running on CentOS 6.3. I am also using 
> JDK 1.7 to run Jenkins and Winstone. I am using a subversion repository, 
> version 1.7, configured to authenticate via HTTPS client certificates. 
>
> After entering the path to a PKCS12 certificate and providing the correct 
> password, I am receiving this exception : 
>
> Attempting an SSL client certificate authentcation
> Failed to authenticate: svn: E170001: SSL handshake failed: 'Received fatal 
> alert: handshake_failure'
>
> Tracing the stack of exceptions, the root of it appears to be this : 
>
> Caused by: java.lang.NullPointerException
>   at 
> org.tmatesoft.svn.core.internal.io.dav.http.HTTPSSLKeyManager.loadClientCertificate(HTTPSSLKeyManager.java:187)
>   at 
> org.tmatesoft.svn.core.internal.io.dav.http.HTTPSSLKeyManager.initialize(HTTPSSLKeyManager.java:441)
>   at 
> org.tmatesoft.svn.core.internal.io.dav.http.HTTPSSLKeyManager.initializeNoException(HTTPSSLKeyManager.java:406)
>   at 
> org.tmatesoft.svn.core.internal.io.dav.http.HTTPSSLKeyManager.chooseClientAlias(HTTPSSLKeyManager.java:302)
>   at 
> sun.security.ssl.AbstractKeyManagerWrapper.chooseClientAlias(SSLContextImpl.java:1029)
>   at 
> sun.security.ssl.ClientHandshaker.serverHelloDone(ClientHandshaker.java:703)
>   at 
> sun.security.ssl.ClientHandshaker.processMessage(ClientHandshaker.java:282)
>   at sun.security.ssl.Handshaker.processLoop(Handshaker.java:868)
>   at sun.security.ssl.Handshaker.process_record(Handshaker.java:804)
>   at sun.security.ssl.SSLSocketImpl.readRecord(SSLSocketImpl.java:966)
>   at 
> sun.security.ssl.SSLSocketImpl.performInitialHandshake(SSLSocketImpl.java:1262)
>   at sun.security.ssl.SSLSocketImpl.writeRecord(SSLSocketImpl.java:680)
>   at sun.security.ssl.AppOutputStream.write(AppOutputStream.java:85)
>   at 
> java.io.BufferedOutputStream.flushBuffer(BufferedOutputStream.java:82)
>   at java.io.BufferedOutputStream.flush(BufferedOutputStream.java:140)
>   at 
> org.tmatesoft.svn.core.internal.io.dav.http.HTTPConnection.sendData(HTTPConnection.java:238)
>   at 
> org.tmatesoft.svn.core.internal.io.dav.http.HTTPRequest.dispatch(HTTPRequest.java:168)
>   at 
> org.tmatesoft.svn.core.internal.io.dav.http.HTTPConnection._request(HTTPConnection.java:385)
>   ... 63 more
>
> I have attempted a similar configuration changing certain elements (however I 
> cannot change any configurations or software version on the local SVN repo). 
> I have attempted to deploy Jenkins via GlassFish, have attempted to running 
> it on JDK 1.6 update 37 
> on both Winstone and GlassFish, have attempted to downgrade the Subversion 
> plug-in and run it in the previous configurations. I was still running into 
> this exception in all these configurations, until I tried running everything 
> on a different machine on CentOS 5.8, 
> at which point I am not encountering this exception and am successfully able 
> to authenticate, check-out working copies and execute build jobs normally. 
>
> For now, I am able to use Jenkins successfully, but I might need to use 
> Jenkins to build a larger scale build server than what I am using now, in a 
> CentOS 6.3 enviroment. Is anyone here familiar with this exception ? Does 
> anyone here have any tips on how to work around this ?
> Is this an issue, or is there a mistake I am making somewhere ? 
>
> Thank You, 
> Andrew 
>
>

Email-ext plugin. variable substitution in included file ?

[Chris_Williams1]

We are using the "${FILE,path}" token to include the contents of a file in the 
workspace to an email being sent by the Email-ext plugin. 
We also have a parameter on this particular job (i.e. parameterized build). 

It appears that if the file being included uses a variable with the same name 
as the job parameter, then the value of the parameter is substituted for that 
variable in the sent email. Is this expected behavior or should this be 
considered a bug ? We can work around the problem by changing the parameter 
name, but that's a hack. I've also experimented with trying to reset the 
parameter before the email is sent, but it doesn't work or I'm doing it wrong.

Example:
1) Job has a parameter named "PASSWD". 
2) Ext-email uses the token: ${FILE, path="./foo.sh"}
3) foo.sh contents:
echo ${PASSWD}

4) When the email is sent, the ${PASSWD} variable in foo.sh is evaluated and 
the email contains the value that was typed in for the PASSWD job parameter.

I'll open an issue if this is considered a bug or unexpected. An additional 
data point is, that if I'm including a file named "foo.txt" instead of 
"foo.sh", then the "${PASSWD}" string is not evaluated and appears as expected 
in the email.

Thanks,
Chris

Chris Williams
Dell | PG Release Engineering

Re: Run a build concurrently on all slaves?

[org.jenkins-ci]

On Fri, 11 Jan 2013 21:38:04 +0100
felix schwitzer  wrote:
> For this we use simply a multi-configuration project.

On Fri, 11 Jan 2013 19:35:42 -
"KEVIN FLEMING (BLOOMBERG/ 731 LEXIN)"  wrote:
> This is exactly what Matrix builds are designed for. You define one
> job, with some number of axes in the matrix, and then corresponding
> labels on your nodes (slaves) to execute the spawned jobs.

OK, thanks.

I guess I'm still not entirely clear on the necessary steps, having read
as much documentation as I could find on multi-configuration projects.

I started by labelling each node with relevant labels ("solaris",
"linux", "x86_64", etc) and gave each node a "cluster" label. Then,
I created a multi-configuration project and then added a "Label
expression" axis, where the name is "cluster" and the label expression
is also "cluster" (I realize those two aren't related).

Then, I added a build step to build top-level Maven targets, and
a post-build step to aggregate test results at the end. The Maven goal
was set to "verify" (in order to build the project, run the unit tests,
produce a package, and verify the package).

Then, I brought three nodes online and triggered a build. Two of the
nodes began building. The other node stayed idle.

One of the two nodes finished much faster than the other, which looked
rather suspicious. Checking the console log for that build showed that
it simply checked out the source code from version control and declared
success (whilst the other node checked out the source code and actually
built it):

Triggering cluster
cluster completed with result SUCCESS
Finished: SUCCESS

Am I missing something incredibly obvious here?

I'd obviously like *all* of the labelled nodes to run, and to actually
build code as opposed to ... not.

Regards,
M

Re: How to pass parameters between jobs - quick question

[William Soula]

I think job B would see SVN_REVISION not SVN_REVISION_NUMBER because you are 
setting SVN_REVISION in the post-build section.  We use the "triggger/call 
builds on other projects" for the build step and it works perfectly fine to 
call job B from job A and echo the predefined parameters passed to it.  We use 
this to have a general deploy job that takes the server and instance as 
parameters from the more specific deploy job i.e. deploy_test calls deploy_web 
and passes the test server and instance as a parameter and this all works 
correctly.

Will

On 01/11/2013 02:05 PM, Stanley, Jason wrote:
I ran into this very same problem.

It appears (at least with my experience) the only variables passed from Job A 
-> Job B are the variables predefined by Jenkins.

To work around the issue,  in my build shell for Job A, I wrote the 
variable=value into a file and used the “Parameters from properties file”  
Parameter in the “Trigger parameterized build on other projects” when running 
Job B.

Jason

From: jenkinsci-users@googlegroups.com 
[mailto:jenkinsci-users@googlegroups.com] On Behalf Of zw
Sent: Friday, January 11, 2013 2:50 PM
To: jenkinsci-users@googlegroups.com
Subject: How to pass parameters between jobs - quick question

Hi All

We have job A to kick off Job B
Job A uses in build section execute shell to get a svn revision number and sets 
SVN_REVISION.
(We understand that there's the option to use "This build is parameterized" 
String Parameter and set SVN_REVISION value as HEAD, but we have our reasons to 
not use that)
echo ${SVN_REVISION_NUM}
export SVN_REVISION=${SVN_REVISION_NUM}
echo ${SVN_REVISION}


output

+ SVN_REVISION_NUM=47337

:

+ echo 47337

47337

+ export SVN_REVISION=47337

+ SVN_REVISION=47337



In Post-build Actions section

we have Trigger parameterized build on other projects, and predefined parameter 
set to SVN_REVISION=${SVN_REVISION_NUM}



But in Job B console, it dosent see the revision number and Job B's SCM could 
retrieve the revision number

Location 'https://jobBhost.com/svn/src@${SVN_REVISION_NUM}' does not exist



What did we do wrong ?



Thanks

Re: Run a build concurrently on all slaves?

[felix schwitzer]

For this we use simply a multi-configuration project.

Regards
Felix

On 01/11/2013 05:23 PM, org.jenkins...@io7m.com wrote:

Hello.

I've set up Jenkins to allow me to test code on a wide range of
platforms. Essentially the setup is a master node with no executors
(it doesn't build code) that only dispatches jobs to slaves,
and a set of slaves (most of which are running in virtual machines and
frequently become unavailable).

I've set up a few jobs that are triggered whenever changes are detected
in a remote SCM, but I can't work out how to trigger a build to run
concurrently on all available nodes (leaving the job Pending on all
nodes that are currently offline). What's the correct way to do this?

I've seen a few posts that hint towards using
https://wiki.jenkins-ci.org/display/JENKINS/NodeLabel+Parameter+Plugin
but unfortunately I found the description incomprehensible and there
doesn't seem to be any other documentation.

Any help would be appreciated!
M

RE: How to pass parameters between jobs - quick question

[Stanley, Jason]

I ran into this very same problem.

It appears (at least with my experience) the only variables passed from Job A 
-> Job B are the variables predefined by Jenkins.

To work around the issue,  in my build shell for Job A, I wrote the 
variable=value into a file and used the “Parameters from properties file”  
Parameter in the “Trigger parameterized build on other projects” when running 
Job B.

Jason

From: jenkinsci-users@googlegroups.com 
[mailto:jenkinsci-users@googlegroups.com] On Behalf Of zw
Sent: Friday, January 11, 2013 2:50 PM
To: jenkinsci-users@googlegroups.com
Subject: How to pass parameters between jobs - quick question

Hi All

We have job A to kick off Job B
Job A uses in build section execute shell to get a svn revision number and sets 
SVN_REVISION.
(We understand that there's the option to use "This build is parameterized" 
String Parameter and set SVN_REVISION value as HEAD, but we have our reasons to 
not use that)
echo ${SVN_REVISION_NUM}
export SVN_REVISION=${SVN_REVISION_NUM}
echo ${SVN_REVISION}


output

+ SVN_REVISION_NUM=47337

:

+ echo 47337

47337

+ export SVN_REVISION=47337

+ SVN_REVISION=47337



In Post-build Actions section

we have Trigger parameterized build on other projects, and predefined parameter 
set to SVN_REVISION=${SVN_REVISION_NUM}



But in Job B console, it dosent see the revision number and Job B's SCM could 
retrieve the revision number

Location 'https://jobBhost.com/svn/src@${SVN_REVISION_NUM}' does not exist



What did we do wrong ?



Thanks

How to pass parameters between jobs - quick question

[zw]

Hi All

We have job A to kick off Job B
Job A uses in build section execute shell to get a svn revision number and 
sets SVN_REVISION. 
(We understand that there's the option to use "This build is parameterized" 
String Parameter and set SVN_REVISION value as HEAD, but we have our 
reasons to not use that)
echo ${SVN_REVISION_NUM}
export SVN_REVISION=${SVN_REVISION_NUM}
echo ${SVN_REVISION}

output

+ SVN_REVISION_NUM=47337
:

+ echo 47337
47337
+ export SVN_REVISION=47337
+ SVN_REVISION=47337


In Post-build Actions section

we have Trigger parameterized build on other projects, and predefined parameter 
set to SVN_REVISION=${SVN_REVISION_NUM}


But in Job B console, it dosent see the revision number and Job B's SCM could 
retrieve the revision number

Location 'https://jobBhost.com/svn/src@${SVN_REVISION_NUM}' does not exist


What did we do wrong ?


Thanks

Re: Run a build concurrently on all slaves?

[Les Mikesell]

On Fri, Jan 11, 2013 at 1:35 PM, KEVIN FLEMING (BLOOMBERG/ 731 LEXIN)
 wrote:
> This is exactly what Matrix builds are designed for. You define one job, with 
> some number of axes in the matrix, and then corresponding labels on your 
> nodes (slaves) to execute the spawned jobs.

But note that you can't add this to an existing job.  You have to
create the job as a'multi-configuration project' in the first place to
get that option.

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

matrix job aggregation

[Chris Withers]

(no response to this first time round, so trying again)

Hi,

What's the correct way to do aggregation in matrix jobs?

I've configured the email ext plugin to send out emails, and I'd expect 
either:


- one email from the whole multi-config job combining stuff from all the 
elements in the matrix


- one email from each of the elements in the matrix including only the 
stuff from that element


What I appear to be getting is one email from the whole multi-config job 
but with no test results, etc found from it.


Where can I find docs on how to set up multi-config jobs for either of 
the above scenarios?


Chris

--
Simplistix - Content Management, Batch Processing & Python Consulting
- http://www.simplistix.co.uk

__
This email has been scanned by the Symantec Email Security.cloud service.
For more information please visit http://www.symanteccloud.com
__

Re:Run a build concurrently on all slaves?

[KEVIN FLEMING (BLOOMBERG/ 731 LEXIN)]

This is exactly what Matrix builds are designed for. You define one job, with 
some number of axes in the matrix, and then corresponding labels on your nodes 
(slaves) to execute the spawned jobs.

- Original Message -
From: jenkinsci-users@googlegroups.com
To: jenkinsci-users@googlegroups.com
At: Jan 11 2013 11:24:12

Hello.

I've set up Jenkins to allow me to test code on a wide range of
platforms. Essentially the setup is a master node with no executors
(it doesn't build code) that only dispatches jobs to slaves,
and a set of slaves (most of which are running in virtual machines and
frequently become unavailable).

I've set up a few jobs that are triggered whenever changes are detected
in a remote SCM, but I can't work out how to trigger a build to run
concurrently on all available nodes (leaving the job Pending on all
nodes that are currently offline). What's the correct way to do this?

I've seen a few posts that hint towards using
https://wiki.jenkins-ci.org/display/JENKINS/NodeLabel+Parameter+Plugin
but unfortunately I found the description incomprehensible and there
doesn't seem to be any other documentation.

Any help would be appreciated!
M

GRADLE_HOME not available on slaves

[morty]

On the jenkins master I have configured that gradle should be autoinstalled
to:

*GRADLE_HOME: /var/jenkins/tools/mygradle*

on all slave nodes resulting in the gradle script being located here:

/var/jenkins/tools/mygradle/bin/gradle



This works fine and my gradle jobs builds on the slave nodes.

Now I need to call a .sh script (Build step - Execute shell) in a job. At
some point the script calls gradle and therefore requires that its in the
PATH. But when I execute the job (on one of the build slaves) gradle cannot
be found.

Is it possible to auto install gradle on all slaves and ALSO add gradle to
the PATH on all slaves?



--
View this message in context: 
http://jenkins.361315.n4.nabble.com/GRADLE-HOME-not-available-on-slaves-tp4651205.html
Sent from the Jenkins users mailing list archive at Nabble.com.

Re: Jenkins failed to sync from Perforce due to long file name

[Richard J]

> Yes, on Winsows, Perforce uses the Win32 interface to the file system 
> which is limited to something like a 250 character path length. have 
> suggested that they use the direct NT interface that allows 32K unicode 
> chars, but no luck so far.

Re: Re-key successful or not?

[flsobral]

I forgot to mention in my original post that I'm running Jenkins as Windows 
Service on Windows 7 as system user.


Fabio.


On Friday, January 11, 2013 1:39:50 PM UTC-2, Wolf wrote:
>
> Same here.
>
> Errors in the re-keying log like:
>
> ERROR: Failed to rewrite C:\Program Files\Jenkins\config.xml
>
> and
>
> ERROR: Failed to rewrite C:\Program Files\Jenkins\users\somebody\config.xml
>
> Running Jenkins as a Windows Service on Windows 7, 32-bit as the Local System 
> user.
>
>
> Wolf
>
>

Re: Changes since last success in Email Extension Plugin

[Dirk Kuypers]

Hi Alex,

I feared this answer.;-)

OK, so I have to dig into this.

Thanks for your hard work on the plugin!
Dirk

2013/1/11 Slide 

> In my opinion the tokens are only useful up to a certain point. My
> recommendation to people is to start using the Groovy templates more. You
> can interact with the Jenkins classes and get pretty much any information
> you want without having to have a new feature added to the plugin.
>
> slide
>
>
> On Fri, Jan 11, 2013 at 9:19 AM, Dirk Kuypers wrote:
>
>> Hi,
>>
>> as my developers keep asking about this I am re-sending my question.
>>
>> If a long-running test job is started after 2 or more faster compiles
>> went through, I am only getting the changes for the last compile inside the
>> email (with email-extension plugin). What I would like to have instead are
>> the changes in dependency if there has been more than 1 compile between the
>> test runs.
>>
>> So I would be happy If I could use something like
>>
>> ${CHANGES_IN_DEPENDENCY_SINCE_LAST_SUCCESS, showPaths=true,
>> pathformat="%p"}
>>
>> which falls back to
>>
>> ${CHANGES_SINCE_LAST_SUCCESS, showPaths=true, pathformat="%p"}
>>
>> if those are the same (1 to 1 relation between compile/test).
>>
>> Is this a request for a new feature or is it already possible to achieve
>> this via groovy templating?!?
>>
>> Thanks for helping out
>> Dirk
>>
>>
>> 2012/11/22 Dirk Kuypers 
>>
>>> Hi,
>>>
>>> I have a compile job that runs continuously triggered by SCM which
>>> starts about 40 or 50 test jobs after successful compilation.
>>> I am printing the changes since last success (coming from the upstream
>>> compile job) of a test job which is run after a compile job like this:
>>> ${CHANGES_SINCE_LAST_SUCCESS, showPaths=true, pathformat="%p"}
>>>
>>> This has worked until I began to allow the compile job to run
>>> concurrently because the feedback cycle was getting too long recently with
>>> compile times of 6 minutes and test times up to more than 20 minutes and
>>> changes of more than one developer getting mixed up into one build. Now it
>>> can happen that I have several compiles before a special test is running
>>> again. Looking at the build of the test Jenkins handles this correctly
>>> because it says:
>>>
>>> No changes from last build. Changes in dependency
>>>
>>>1. JobContinuous [image: Success]#9601 → [image: Success]
>>>#9605 (detail)
>>>
>>>
>>> Clicking on details I get all changes summarized for those 5 builds.
>>>
>>> The email only prints nothing. If there is a 1 to 1 relation between
>>> compile and test, emailing the changes works.
>>>
>>> Is this a bug/feature request? Am I doing something wrong there?
>>>
>>> BR
>>> Dirk
>>>
>>
>>
>
>
> --
> Website: http://earl-of-code.com
>

Run a build concurrently on all slaves?

[org.jenkins-ci]

Hello.

I've set up Jenkins to allow me to test code on a wide range of
platforms. Essentially the setup is a master node with no executors
(it doesn't build code) that only dispatches jobs to slaves,
and a set of slaves (most of which are running in virtual machines and
frequently become unavailable).

I've set up a few jobs that are triggered whenever changes are detected
in a remote SCM, but I can't work out how to trigger a build to run
concurrently on all available nodes (leaving the job Pending on all
nodes that are currently offline). What's the correct way to do this?

I've seen a few posts that hint towards using
https://wiki.jenkins-ci.org/display/JENKINS/NodeLabel+Parameter+Plugin
but unfortunately I found the description incomprehensible and there
doesn't seem to be any other documentation.

Any help would be appreciated!
M

Re: Changes since last success in Email Extension Plugin

[Slide]

In my opinion the tokens are only useful up to a certain point. My
recommendation to people is to start using the Groovy templates more. You
can interact with the Jenkins classes and get pretty much any information
you want without having to have a new feature added to the plugin.

slide


On Fri, Jan 11, 2013 at 9:19 AM, Dirk Kuypers wrote:

> Hi,
>
> as my developers keep asking about this I am re-sending my question.
>
> If a long-running test job is started after 2 or more faster compiles went
> through, I am only getting the changes for the last compile inside the
> email (with email-extension plugin). What I would like to have instead are
> the changes in dependency if there has been more than 1 compile between the
> test runs.
>
> So I would be happy If I could use something like
>
> ${CHANGES_IN_DEPENDENCY_SINCE_LAST_SUCCESS, showPaths=true,
> pathformat="%p"}
>
> which falls back to
>
> ${CHANGES_SINCE_LAST_SUCCESS, showPaths=true, pathformat="%p"}
>
> if those are the same (1 to 1 relation between compile/test).
>
> Is this a request for a new feature or is it already possible to achieve
> this via groovy templating?!?
>
> Thanks for helping out
> Dirk
>
>
> 2012/11/22 Dirk Kuypers 
>
>> Hi,
>>
>> I have a compile job that runs continuously triggered by SCM which starts
>> about 40 or 50 test jobs after successful compilation.
>> I am printing the changes since last success (coming from the upstream
>> compile job) of a test job which is run after a compile job like this:
>> ${CHANGES_SINCE_LAST_SUCCESS, showPaths=true, pathformat="%p"}
>>
>> This has worked until I began to allow the compile job to run
>> concurrently because the feedback cycle was getting too long recently with
>> compile times of 6 minutes and test times up to more than 20 minutes and
>> changes of more than one developer getting mixed up into one build. Now it
>> can happen that I have several compiles before a special test is running
>> again. Looking at the build of the test Jenkins handles this correctly
>> because it says:
>>
>> No changes from last build. Changes in dependency
>>
>>1. JobContinuous [image: Success]#9601 → [image: Success]
>>#9605 (detail)
>>
>>
>> Clicking on details I get all changes summarized for those 5 builds.
>>
>> The email only prints nothing. If there is a 1 to 1 relation between
>> compile and test, emailing the changes works.
>>
>> Is this a bug/feature request? Am I doing something wrong there?
>>
>> BR
>> Dirk
>>
>
>


-- 
Website: http://earl-of-code.com

Re: Changes since last success in Email Extension Plugin

[Dirk Kuypers]

Hi,

as my developers keep asking about this I am re-sending my question.

If a long-running test job is started after 2 or more faster compiles went
through, I am only getting the changes for the last compile inside the
email (with email-extension plugin). What I would like to have instead are
the changes in dependency if there has been more than 1 compile between the
test runs.

So I would be happy If I could use something like

${CHANGES_IN_DEPENDENCY_SINCE_LAST_SUCCESS, showPaths=true, pathformat="%p"}

which falls back to

${CHANGES_SINCE_LAST_SUCCESS, showPaths=true, pathformat="%p"}

if those are the same (1 to 1 relation between compile/test).

Is this a request for a new feature or is it already possible to achieve
this via groovy templating?!?

Thanks for helping out
Dirk

2012/11/22 Dirk Kuypers 

> Hi,
>
> I have a compile job that runs continuously triggered by SCM which starts
> about 40 or 50 test jobs after successful compilation.
> I am printing the changes since last success (coming from the upstream
> compile job) of a test job which is run after a compile job like this:
> ${CHANGES_SINCE_LAST_SUCCESS, showPaths=true, pathformat="%p"}
>
> This has worked until I began to allow the compile job to run concurrently
> because the feedback cycle was getting too long recently with compile times
> of 6 minutes and test times up to more than 20 minutes and changes of more
> than one developer getting mixed up into one build. Now it can happen that
> I have several compiles before a special test is running again. Looking at
> the build of the test Jenkins handles this correctly because it says:
>
> No changes from last build. Changes in dependency
>
>1. JobContinuous [image: Success]#9601 → [image: Success]#9605 (detail)
>
>
> Clicking on details I get all changes summarized for those 5 builds.
>
> The email only prints nothing. If there is a 1 to 1 relation between
> compile and test, emailing the changes works.
>
> Is this a bug/feature request? Am I doing something wrong there?
>
> BR
> Dirk
>

Maven claims "No compiler is provided in this environment" on XP

[org.jenkins-ci]

Hello.

I've just configured an XP VM as a slave service. I configured the
service to log in as Administrator (in Administrative Tools ->
Services). The environment appears to be correctly configured in that
JAVA_HOME is set to the location of the installed JDK, and the directory
containing the javac executable appears on the PATH. I can compile Maven
projects from the command line as the Administrator user, and can also
run the javac executable from the Groovy script console.

However, on attempting to build a checked out project in a job...

[INFO] Compiling 19 source files to
c:\jenkins\workspace\io7m-jaux\io7m-jaux-io7m-jaux-11\target\classes
[INFO] -
[ERROR] COMPILATION ERROR : [INFO]
- [ERROR]
No compiler is provided in this environment. Perhaps you are running on
a JRE rather than a JDK? [INFO] 1 error [INFO]
-
mojoFailed
org.apache.maven.plugins:maven-compiler-plugin:3.0(default-compile)
projectFailed com.io7m.jaux:io7m-jaux:2.4.0 sessionEnded [INFO]

[INFO] BUILD FAILURE [INFO]

[INFO] Total time: 9.264s [INFO] Finished at: Fri Jan 11 05:20:58 PST
2013 [INFO] Final Memory: 12M/30M
[INFO]

Projects to build: [MavenProject: com.io7m.jaux:io7m-jaux:2.4.0 @
c:\jenkins\workspace\io7m-jaux\io7m-jaux-io7m-jaux-11\pom.xml]
[JENKINS] Archiving
c:\jenkins\workspace\io7m-jaux\io7m-jaux-io7m-jaux-11\pom.xml
to 
/jenkins/.jenkins/jobs/io7m-jaux/modules/com.io7m.jaux$io7m-jaux/builds/2013-01-11_13-20-43/archive/com.io7m.jaux/io7m-jaux/2.4.0/io7m-jaux-2.4.0.pom
Waiting for Jenkins to finish collecting data mavenExecutionResult
exceptions not empty message : Failed to execute goal
org.apache.maven.plugins:maven-compiler-plugin:3.0:compile
(default-compile) on project io7m-jaux: Compilation failure No compiler
is provided in this environment. Perhaps you are running on a JRE
rather than a JDK?

What's going on here?

Regards,
M

Re: Re-key successful or not?

[Wolf]

Same here.

Errors in the re-keying log like:

ERROR: Failed to rewrite C:\Program Files\Jenkins\config.xml

and

ERROR: Failed to rewrite C:\Program Files\Jenkins\users\somebody\config.xml

Running Jenkins as a Windows Service on Windows 7, 32-bit as the Local System 
user.


Wolf

Re: Jenkins Ant Compile Query

[Justin Mulcahy]

Thanks for the reply Chris.I have tried setting the ${source} variable to 
the path of my class file but Jenkins is still not compiling the class. 
Should i be putting something else in here?  

On Wednesday, 9 January 2013 00:21:56 UTC, Christopher Orr wrote:
>
> On 01/08/2013 11:40 AM, Justin Mulcahy wrote: 
> > I may be posting this on the wrong forum but here goes. I have set up a 
> > job (my First) in Jenkins to execute an ant build, the build completes 
> > successfully but does not seem to be compiling my java class. 
> > 
> > I have attached the Build.xml file, Jenkins job configuration, Jenkins 
> > JDK Installation and Jenkins Console output. Any help would be greatly 
> > appreciated. 
>
> Your "build-project" target runs  which looks for the Java files 
> in the ${source} variable. 
>
> But the "source" variable never appears to be set, so there's nothing to 
> compile. 
>
> Also, unrelated, the "-buildfile ..." part looks unnecessary in the 
> "Targets" configuration, as it seems you've specified the filename 
> (buildA.xml) already in the Advanced section? 
>
> Regards, 
> Chris 
>
>

Re: kicking off a job on a remote Jenkins - quick question

[Clemens]

Just a question, are you running any Security Settings on your Server ?
My command would only work on a "Not secured" Server.
And connected to your Error Message, i would guess you are having your 
Jenkins secured, am i right ?

Re: kicking off a job on a remote Jenkins - quick question

[Clemens]

Sry, go you wrong, my mistake..

Your command should look like this:
java -jar jenkins-cli.jar -s http://YOURSERVER:8080 build JOBNAME -p 
SVN=VALUE

While SVN would be a parameter at your job, thats how i call a job with
my jenkins-cli.jar and parameters

Re: Issue with Email-Ext plugin-Please help

[Slide]

You'll want to read up on Jelly and how to interact with the Jenkins
classes to get the information you want and add it to your template. I
would recommend using the groovy templates instead as they make interacting
with Jenkins classes much easier.

slide


On Fri, Jan 11, 2013 at 3:47 AM, uma shankar wrote:

> Hi Alex,
>
> Thank you very much for suggestion.
>
> I have used ${JELLY_SCRIPT, template="html"} and getting emails in below
> format:
>
> Build URLhttp://localhost:8080/job/test/14/Project: testDate of build:Fri,
> 11 Jan 2013 13:28:40 +0530Build duration:0.37 sec
> *CHANGES*  Revision *9* by *user:**(afsda)*  edit
> //streams/local/test2/test.txt   Revision *8* by *user:**(xzvz)*  edit
> //streams/local/test/test.txt  edit//streams/local/test2/test.txt
>
> I got what was expected, only thing is not getting the each file version.
> and i would also like the build executor name in email.
>
> Could you please help me.
>
>
> On Fri, Jan 11, 2013 at 12:22 AM, Alex Earl  wrote:
>
>> The default groovy template should have the info.
>>
>>
>> Sent from my Windows Phone
>> --
>> From: uma shankar
>> Sent: 1/10/2013 11:22 AM
>> To: Alex Earl
>> Cc: jenkinsci-users@googlegroups.com
>> Subject: Re: Issue with Email-Ext plugin-Please help
>>
>>  Hi Alex,
>>
>> I have tried in many ways to get that output, but did not get it through.
>> Will you be able to help me getting the template?
>>
>> Thank You.
>>
>>
>> On Thu, Jan 10, 2013 at 11:34 PM, Alex Earl wrote:
>>
>>> You may need to create your own template. Look at the wiki for more
>>> information.
>>>
>>> Sent from my Windows Phone
>>> --
>>> From: uma shankar
>>> Sent: 1/10/2013 10:53 AM
>>> To: jenkinsci-users@googlegroups.com
>>> Subject: Issue with Email-Ext plugin-Please help
>>>
>>>  Hi,
>>>
>>> ** **
>>>
>>> I am trying to use Jenkins for POC, and have enabled Email-ext plugin. I
>>> am facing some issue while getting perforce change list.
>>>
>>> ** **
>>>
>>> I am have been using perforce plugin for integration with SCM. I have
>>> some issue with the plugin.
>>> I use Email-ext plugin and trying get complete change list history in
>>> mail with revision no and workspace as display on build home page.
>>> Example: (*Expected to be in e-mail*)
>>> Changes
>>> 1436416 by umashankar@umashakar-test on Jan 10, 2013 8:44:54 AM
>>> Forcing a build
>>> Affected Files:
>>> //myproject/src/test/java/control.java(#16)
>>>
>>> //myproject/src/test/java/control1.java(#17)
>>>
>>> ** **
>>>
>>> ** **
>>>
>>> But I have used some keyword to get this things in email as below:
>>> *
>>>
>>> List of changed file(s):
>>> ${CHANGES, showPaths=true}
>>>
>>> ** **
>>>
>>> ${CHANGES_SINCE_LAST_SUCCESS, showPaths=true}
>>>
>>> *
>>>
>>> ** **
>>>
>>> And output comes in mail as below:
>>> List of changed file(s):
>>> [umashankar] Updates
>>> src/test/java/control.java
>>>
>>> src/test/java/control1.java
>>>
>>> ** **
>>>
>>> Changes for Build #16
>>> [umashankar] Updates
>>> src/test/java/control.java
>>>
>>> src/test/java/control1.java
>>>
>>> It is missing the full path, revision number, changelist workspace name.
>>> 
>>>
>>> ** **
>>>
>>> Early response would help to resolve this issue ASAP.
>>>
>>
>>
>


-- 
Website: http://earl-of-code.com

use tags instead of build numbers in URLs

[Brian J. Murrell]

Anyone know of any way or plugin that allows the use of tag names in
Jenkins URLs rather than build numbers.  i.e. if I have a job named
"foo", I can get build #10's result for it with:

http://jenkins/job/foo/10/

But if I've set a tag for that build as "1.2.4" I'd like to be able to use:

http://jenkins/job/foo/1.2.4/

To navigate to it.  Clearly there's a problem when two builds have the
same tag, in which case I'd prefer to always find the most recent,
barring any further nomenclature to differentiate between the multiply
same-named tags.  Multiply-used tags is not a real problem for me though.

Cheers,
b.



signature.asc
Description: OpenPGP digital signature

Re: kicking off a job on a remote Jenkins - quick question

[Clemens]

Hi,

you could start your Job via url and transmitting a parameter, which 
includes your svn revision Number.
See Example:
YOURSERVER/jobs/JOBNAME/buildWithParameters?PARMNAME=SVNREVISIONNUMBER

In case your remote job is parameterized your SVNREVISIONNUMBER will be set 
as parameter value and
is usable during your remote job.

Or check out the remote API : 
Link

RE: JNLP slave no longer works

[Steffen Breitbach]

We'll fix this in the next mainline release, and once all the fallout bugs
like this are fixed and verified, I'll plan on another LTS release.

 

My apologies for the problems in the mean time.



No problem! The workaround using the anonymous user is OK for me in the
meantime. :)



smime.p7s
Description: S/MIME cryptographic signature

Re: JNLP slave no longer works

[Kohsuke Kawaguchi]

2013/1/11 SBreitbach 

> The slave.jar from LTS 1.480.2 version has parameters "-auth" and
> "-jnlpCredentials". Neither of them seems to work, I will get a "Exception
> in thread "main" java.lang.NoClassDefFoundError:
> org/apache/commons/codec/binary/Base64" or "java.io.IOException: Failed to
> load
> http://bsjenkinsslave03.server.lan:8280/jenkins/computer/bsjenkinsslave03/slave-agent.jnlp:
> 403 Forbidden", depending on what combination of parameters I use.
>
> I consider the "manually download the jnlp file" as a very dirty
> quickhack... will we see a proper launching of the slaves again soon?
>

It's no less secure than using the -jnlpCredentials, and JNLP files do not
change as you change the slave configuration, so in my mind it is pretty
practical workaround, although it does require some extra step.

We'll fix this in the next mainline release, and once all the fallout bugs
like this are fixed and verified, I'll plan on another LTS release.

My apologies for the problems in the mean time.


-- 
Kohsuke Kawaguchi

RE: JNLP slave no longer works

[Matthew.Webber]

But what about when the slave is not installed as a service? I use headless 
jnlp 
(https://wiki.jenkins-ci.org/display/JENKINS/Distributed+builds#Distributedbuilds-Launchslaveagentheadlessly)
 in a script that makes sure slave.jar is up to date.

It looks like should be able to authenticate with -auth  or –jnlpCredentials, 
but neither seems to work  - ideally, I’d like to use my API token rather than 
my password …

Is this supposed to work, or is this a request for a new feature?

From: jenkinsci-users@googlegroups.com 
[mailto:jenkinsci-users@googlegroups.com] On Behalf Of Kohsuke Kawaguchi
Sent: 11 January 2013 14:16
To: jenkinsci-users@googlegroups.com
Subject: Re: JNLP slave no longer works

Slaves installed as a service stores the secret token information in the 
startup configuration file, such as /etc/init/jenkins-slave-*.conf (Linux), 
/System/Library/LaunchDaemons/org.jenkins-ci.slave.*.plist (OS X), or 
$SLAVE_ROOT/jenkins-slave.xml (Windows). Look for seemingly meaningless 
hex-encoded token in this file (which is passed as an argument to the the Java 
program).

Now download the JNLP file again for that slave from Jenkins Web UI, then look 
for a similar seemingly meaningless hex-encoded token in that file. Overwrite 
the local version with what's now in JNLP, and it'd be good to go.

If you aren't sure, you can also simply re-install the slave agent again. 
Existing slave agent will do no harm since it will not be able to connect back.

2013/1/11 mailto:stuart.gr...@doccentrics.com>>

Thanks for the info, but how is this change applied to a linux slave?

On Tuesday, 8 January 2013 02:50:39 UTC, Kohsuke Kawaguchi wrote:
This is unfortunately a necessary measure.

See the "Fix" section in 
https://wiki.jenkins-ci.org/display/SECURITY/Jenkins+Security+Advisory+2013-01-04


--
Kohsuke Kawaguchi



--
Kohsuke Kawaguchi



-- 

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: JNLP slave no longer works

[Kohsuke Kawaguchi]

If your command line is something like "java -jar slave.jar -jnlpUrl
http://host/path/to/jnlp";, then the right thing to do is to use the
-jnlpCredentials option. But as SBreitBach discovered, due to a missing
dependency problem it turns out this doesn't work (Ugh!). This is being
fixed now.

In the mean time, what I recommend is to manually download the JNLP file
from your browser, then have -jnlpUrl point to the local copy via
file:///path/to/local.jnlp

I'm not really sure what the -auth option does, even though I added it :-(


2013/1/11 

>  From a message I sent to the list a few days ago, which worked for at
> least some people:
>
> ** **
>
> >>>** **
>
> I was in a slightly different situation (linux, slaves launched manually),
> but had the 403 as well. Fixed by going to Manage Jenkins ->  Configure
> Global Security, and under  Project-based Matrix Authorization Strategy I
> had to enable “connect” in the “slave” section, for user “Anonymous”.
>
> <<<** **
>
> ** **
>
> I’m not sure if that address your problem, of what the security
> implications are, but anyhow …
>
> ** **
>
> Matthew Webber
>
> ** **
>
> *From:* jenkinsci-users@googlegroups.com [mailto:
> jenkinsci-users@googlegroups.com] *On Behalf Of *
> stuart.gr...@doccentrics.com
> *Sent:* 11 January 2013 10:39
> *To:* jenkinsci-users@googlegroups.com
> *Subject:* Re: JNLP slave no longer works
>
> ** **
>
> ** **
>
> Thanks for the info, but how is this change applied to a linux slave?
>
>
> On Tuesday, 8 January 2013 02:50:39 UTC, Kohsuke Kawaguchi wrote:
>
> This is unfortunately a necessary measure.
>
> ** **
>
> See the "Fix" section in
> https://wiki.jenkins-ci.org/display/SECURITY/Jenkins+Security+Advisory+2013-01-04
> 
>
>
> 
>
> ** **
>
> --
> Kohsuke Kawaguchi 
>



-- 
Kohsuke Kawaguchi

Re: JNLP slave no longer works

[Kohsuke Kawaguchi]

If you want to be able to add slaves without configuring them first on the
master, you can use the swarm slave plugin.


2013/1/11 Les Mikesell 

> On Fri, Jan 11, 2013 at 7:12 AM, Brian J. Murrell 
> wrote:
> >>
> >> Fixed by going to Manage Jenkins ->  Configure Global Security, and
> under  Project-based Matrix Authorization Strategy I had to enable
> “connect” in the “slave” section, for user “Anonymous”.
> >
> > IIUC, that's going to leave your Jenkins server open to allowing any
> > random (i.e. blackhat) slave to connect, accept a build job and submit a
> > build result with a trojan inserted.
> >
> > So not really an acceptable solution, or even work-around.
>
> Doesn't the connecting node name already have to be configured in
> jenkins - and not already connected?   I'd actually like the ability
> to add new anonymous-but-labeled nodes on demand but didn't think it
> was possible.
>
> --
>   Les Mikesell
>  lesmikes...@gmail.com
>



-- 
Kohsuke Kawaguchi

Re: JNLP slave no longer works

[Kohsuke Kawaguchi]

Agree with Brian here. Only do it when you are trusting the users of the
network.


2013/1/11 Brian J. Murrell 

> On 13-01-11 05:55 AM,
> matthew.web...@diamond.ac.uk wrote:
> >
> > Fixed by going to Manage Jenkins ->  Configure Global Security, and
> under  Project-based Matrix Authorization Strategy I had to enable
> “connect” in the “slave” section, for user “Anonymous”.
>
> IIUC, that's going to leave your Jenkins server open to allowing any
> random (i.e. blackhat) slave to connect, accept a build job and submit a
> build result with a trojan inserted.
>
> So not really an acceptable solution, or even work-around.
>
> b.
>
>
>


-- 
Kohsuke Kawaguchi

Re: JNLP slave no longer works

[Kohsuke Kawaguchi]

Slaves installed as a service stores the secret token information in the
startup configuration file, such as /etc/init/jenkins-slave-*.conf (Linux),
/System/Library/LaunchDaemons/org.jenkins-ci.slave.*.plist (OS X), or
$SLAVE_ROOT/jenkins-slave.xml (Windows). Look for seemingly meaningless
hex-encoded token in this file (which is passed as an argument to the the
Java program).

Now download the JNLP file again for that slave from Jenkins Web UI, then
look for a similar seemingly meaningless hex-encoded token in that file.
Overwrite the local version with what's now in JNLP, and it'd be good to go.

If you aren't sure, you can also simply re-install the slave agent again.
Existing slave agent will do no harm since it will not be able to connect
back.


2013/1/11 

>
> Thanks for the info, but how is this change applied to a linux slave?
>
> On Tuesday, 8 January 2013 02:50:39 UTC, Kohsuke Kawaguchi wrote:
>
>> This is unfortunately a necessary measure.
>>
>> See the "Fix" section in
>> https://wiki.jenkins-ci.org/display/SECURITY/Jenkins+Security+Advisory+2013-01-04
>>
>>
>> --
>> Kohsuke Kawaguchi
>>
>


-- 
Kohsuke Kawaguchi

Re: getting Exception in Job Import plugin

[Clemens]

Hi, which Jenkins Version are you using ?

Re: Build a parameterized build while one or more parameters are empty

[Clemens]

Hi agedio,

Your Question : "What happens if you leave the option empty"
(Created a Test Job.)
I just cleared the default value content from the field and the returned 
value from this parameter was "". (not even a space)
The result would be *3.2* if you just 
clear your parameter fields.
Just the result of my test, if there are any other results tell me.
(Was starting the job manually)

Re: JNLP slave no longer works

[Kohsuke Kawaguchi]

Yes, you just need to re-download the JNLP file from the master and
overwrite the local copy at the slave.


2013/1/7 Jon Schewe 

> OK, so if I just keep the jnlp local now instead of always requesting it
> on startup, thing will continue to work, correct?
>
>
> On Monday, January 7, 2013, Kohsuke Kawaguchi wrote:
>
>> This is unfortunately a necessary measure.
>>
>> See the "Fix" section in
>> https://wiki.jenkins-ci.org/display/SECURITY/Jenkins+Security+Advisory+2013-01-04
>>
>>
>> 2013/1/7 Jon Schewe 
>>
>>> I just upgraded to 1.480.2 from 1.471 and my JNLP slaves can no longer
>>> connect. I get an error:
>>>
>>> JNLPException[category: Download Error : Exception: java.io.IOException:
>>> Server returned HTTP response code: 403 for URL:
>>> https://server:8010/computer/slave-jnlp/slave-agent.jnlp : LaunchDesc:
>>> null ]
>>> at com.sun.javaws.Main.launchApp(Unknown Source)
>>> at com.sun.javaws.Main.continueInSecureThread(Unknown Source)
>>> at com.sun.javaws.Main$1.run(Unknown Source)
>>> at java.lang.Thread.run(Unknown Source)
>>>
>>> I have security turned on, although anonymous has overall read. What
>>> other permissions do I need to change? What are my options?
>>>
>>> --
>>> http://mtu.net/~jpschewe
>>>
>>>
>>
>>
>> --
>> Kohsuke Kawaguchi
>>
>
>
> --
> On my mobile device - please excuse typos
>



-- 
Kohsuke Kawaguchi

Re: JNLP slave no longer works

[Brian J. Murrell]

On 13-01-11 08:44 AM, Les Mikesell wrote:
> 
> Doesn't the connecting node name already have to be configured in
> jenkins

Sure.  But if anonymous read is allowed, anyone can see the node names.

> - and not already connected?

Is disconnecting a node really that difficult?  You could try attacking
a vulnerability or even simply DDoS it.  Or even just wait for some
network/node flap and jump in while it's not connected.

b.




signature.asc
Description: OpenPGP digital signature

Re: JNLP slave no longer works

[Les Mikesell]

On Fri, Jan 11, 2013 at 7:12 AM, Brian J. Murrell  wrote:
>>
>> Fixed by going to Manage Jenkins ->  Configure Global Security, and under  
>> Project-based Matrix Authorization Strategy I had to enable “connect” in the 
>> “slave” section, for user “Anonymous”.
>
> IIUC, that's going to leave your Jenkins server open to allowing any
> random (i.e. blackhat) slave to connect, accept a build job and submit a
> build result with a trojan inserted.
>
> So not really an acceptable solution, or even work-around.

Doesn't the connecting node name already have to be configured in
jenkins - and not already connected?   I'd actually like the ability
to add new anonymous-but-labeled nodes on demand but didn't think it
was possible.

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

Job runs blocking on previous run of same job -- why?

[Mandeville, Rob]

I've been having problems where some of my builds appear to hang between 
finishing build actions and running the publisher.  Server and client are 
running on Java 1.7 on Red Hat Linux.  During these hangs, the slave node being 
"hung" isn't running anything, nor has file descriptors open to jobs it had 
launched (there is literally nothing but the slave node running on that given 
Linux account).

Per the thread dump, I get:

Executor #0 for build10_armada_5 : executing galleon_allIntegration #13982 : 
waiting for Check point hudson.plugins.templateproject.ProxyPublisher on 
galleon_allIntegration #13981
"Executor #0 for build10_armada_5 : executing galleon_allIntegration #13982 : 
waiting for Check point hudson.plugins.templateproject.ProxyPublisher on 
galleon_allIntegration #13981" Id=376 Group=main WAITING on 
hudson.model.Run$RunExecution$CheckpointSet@22d65fc7
at java.lang.Object.wait(Native Method)
-  waiting on hudson.model.Run$RunExecution$CheckpointSet@22d65fc7
at java.lang.Object.wait(Object.java:503)
at 
hudson.model.Run$RunExecution$CheckpointSet.waitForCheckPoint(Run.java:1363)
at hudson.model.Run.waitForCheckpoint(Run.java:1321)
at hudson.model.CheckPoint.block(CheckPoint.java:144)
at hudson.tasks.BuildStepMonitor$2.perform(BuildStepMonitor.java:25)
at 
hudson.model.AbstractBuild$AbstractBuildExecution.perform(AbstractBuild.java:717)
at 
hudson.model.AbstractBuild$AbstractBuildExecution.performAllBuildSteps(AbstractBuild.java:692)
at hudson.model.Build$BuildExecution.post2(Build.java:183)
at 
hudson.model.AbstractBuild$AbstractBuildExecution.post(AbstractBuild.java:639)
at hudson.model.Run.execute(Run.java:1527)
at hudson.model.FreeStyleBuild.run(FreeStyleBuild.java:46)
at hudson.model.ResourceController.execute(ResourceController.java:88)
at hudson.model.Executor.run(Executor.java:236)



Basically, this galleon_allIntegration job is a multi-hour integration test 
suite that gets run some 60-70 times a night on different branches of the same 
source tree.  When it gets hung like this, it's specifically after running all 
the build steps and before running a publisher taken from the template project 
plugin (these take a build action and a publisher from another project).  Per 
the CheckPoint class (which I don't understand due to my own ignorance), build 
#13982 is intentionally waiting for #13981 to reach some sort of checkpoint.  
As far as I'm concerned, these jobs are on different source branches and should 
be independent, not waiting for each other.

Is there any way to convince Jenkins that these different runs of 
galleon_allIntegration are independent so that they don't block on each other? 
Is launching 60-70 simultaneous multi-hour runs of the same project not 
particularly Jenkins-friendly?

Thanks in advance,

--Rob Mandeville
Litle & Co (part of the Vantiv family)

The information in this message is for the intended recipient(s) only and may 
be the proprietary and/or confidential property of Litle & Co., LLC, and thus 
protected from disclosure. If you are not the intended recipient(s), or an 
employee or agent responsible for delivering this message to the intended 
recipient, you are hereby notified that any use, dissemination, distribution or 
copying of this communication is prohibited. If you have received this 
communication in error, please notify Litle & Co. immediately by replying to 
this message and then promptly deleting it and your reply permanently from your 
computer.

Re: JNLP slave no longer works

[Brian J. Murrell]

On 13-01-11 05:55 AM,
matthew.web...@diamond.ac.uk wrote:
>
> Fixed by going to Manage Jenkins ->  Configure Global Security, and under  
> Project-based Matrix Authorization Strategy I had to enable “connect” in the 
> “slave” section, for user “Anonymous”.

IIUC, that's going to leave your Jenkins server open to allowing any
random (i.e. blackhat) slave to connect, accept a build job and submit a
build result with a trojan inserted.

So not really an acceptable solution, or even work-around.

b.




signature.asc
Description: OpenPGP digital signature

Re: JNLP slave no longer works

[Brian J. Murrell]

On 13-01-11 07:58 AM, SBreitbach wrote:
> The slave.jar from LTS 1.480.2 version has parameters "-auth" and 
> "-jnlpCredentials".

I discovered that the -auth in 1.498 also doesn't work.  I grabbed the
packets that the slave was sending to the jenkins server on connect and
there is absolutely no authentication credentials in them, even with
-auth account:password being set.

> I consider the "manually download the jnlp file" as a very dirty 
> quickhack...

+1

b.





signature.asc
Description: OpenPGP digital signature

Re: JNLP slave no longer works

[SBreitbach]

The slave.jar from LTS 1.480.2 version has parameters "-auth" and 
"-jnlpCredentials". Neither of them seems to work, I will get a "Exception 
in thread "main" java.lang.NoClassDefFoundError: 
org/apache/commons/codec/binary/Base64" or "java.io.IOException: Failed to 
load 
http://bsjenkinsslave03.server.lan:8280/jenkins/computer/bsjenkinsslave03/slave-agent.jnlp:
 
403 Forbidden", depending on what combination of parameters I use.

I consider the "manually download the jnlp file" as a very dirty 
quickhack... will we see a proper launching of the slaves again soon?

gerrit trigger build on push to refs/heads/*?

[Brian J. Murrell]

Hi,

I wonder if there is any way to configure the gerrit trigger to trigger
builds on pushes to refs/heads/* as was as the usual refs/for/*.

Cheers,
b.



signature.asc
Description: OpenPGP digital signature

Re: UI Not Showing Job Status

[rhythmicdevil]

I saw another thread in here that suggested restarting Jenkins. I followed 
the suggestion and the problem now appears to be "fixed".

Re: JNLP slave no longer works

[Richard Mortimer]

Hi,

I managed to fix this fairly simply on my windows slaves (tested on XP, 
Vista & Win 7). The fix should be similar for Linux slaves but I use a 
home-cooked script to start those so you'll have to adapt the solution 
as appropriate.


1 - autheticate as a Jenkins admin in a browser of your choice and 
download the slave-agent.jnlp file for the slave. The download URL will 
be something like

http:///computer//slave-agent.jnlp
and can be found in the jenkins-slave.xml file on the slave.

2 - transfer slave-agent.jnlp to the slave and store in the same folder 
as slave.jar and jenkins-slave.xml files


3 - edit jenkins-slave.xml and change the URL of slave-agent.jnlp to be
file:///%BASE%/slave-agent.jnlp
That means that the arguments line in jenkins-slave.xml will likely read

 -Xrs -jar "%BASE%\slave.jar" -jnlpUrl 
file:///%BASE%/slave-agent.jnlp


Now just start the slave service as normal and all should work.

Disclaimer: I haven't studied the security implications of storing 
slave-agent.jnlp at that location in the slave. I suspect it is fine but 
it would be good if someone could comment on whether there is a better 
location.


Richard

P.S. I just added this to
https://issues.jenkins-ci.org/browse/JENKINS-16273
because it seems to be an appropriate solution there.

On 11/01/2013 10:55, matthew.web...@diamond.ac.uk wrote:

 From a message I sent to the list a few days ago, which worked for at
least some people:





I was in a slightly different situation (linux, slaves launched
manually), but had the 403 as well. Fixed by going to Manage Jenkins ->
  Configure Global Security, and under  Project-based Matrix
Authorization Strategy I had to enable “connect” in the “slave” section,
for user “Anonymous”.

<<<

I’m not sure if that address your problem, of what the security
implications are, but anyhow …

Matthew Webber

*From:*jenkinsci-users@googlegroups.com
[mailto:jenkinsci-users@googlegroups.com] *On Behalf Of
*stuart.gr...@doccentrics.com
*Sent:* 11 January 2013 10:39
*To:* jenkinsci-users@googlegroups.com
*Subject:* Re: JNLP slave no longer works

Thanks for the info, but how is this change applied to a linux slave?


On Tuesday, 8 January 2013 02:50:39 UTC, Kohsuke Kawaguchi wrote:

This is unfortunately a necessary measure.

See the "Fix" section in
https://wiki.jenkins-ci.org/display/SECURITY/Jenkins+Security+Advisory+2013-01-04


--
Kohsuke Kawaguchi

Re: JNLP slave no longer works

[stuart . green]

Thanks Matthew, I'm a complete newbie with jenkins, I only started at this 
workplace a week ago and this update landed!  Going through my log I 
certainly had the 403 errors and ticked various anonymous access rights to 
get it to connect again.  I've just got it working after I noticed that it 
had moved onto a 401 error which prompted me to get the jnlp 
credentials(which look much like an api key) for our jenkins linux slave 
user account, I reapplied it as the password in the jenkins master and its 
now connecting, even with all the anonymous privileges revoked. 

Cheers.
Stuart Green

On Friday, 11 January 2013 10:55:18 UTC, matthew...@diamond.ac.uk wrote:
>
>  From a message I sent to the list a few days ago, which worked for at 
> least some people:
>
>  
>
> >>> 
>
> I was in a slightly different situation (linux, slaves launched manually), 
> but had the 403 as well. Fixed by going to Manage Jenkins ->  Configure 
> Global Security, and under  Project-based Matrix Authorization Strategy I 
> had to enable “connect” in the “slave” section, for user “Anonymous”.
>
> <<< 
>
>  
>
> I’m not sure if that address your problem, of what the security 
> implications are, but anyhow …
>
>  
>
> Matthew Webber
>
>  
>   
> *From:* jenkins...@googlegroups.com  [mailto:
> jenkins...@googlegroups.com ] *On Behalf Of *
> stuart...@doccentrics.com 
> *Sent:* 11 January 2013 10:39
> *To:* jenkins...@googlegroups.com 
> *Subject:* Re: JNLP slave no longer works
>  
>  
>  
>  
>  
> Thanks for the info, but how is this change applied to a linux slave?
>  
>
> On Tuesday, 8 January 2013 02:50:39 UTC, Kohsuke Kawaguchi wrote:
>  
> This is unfortunately a necessary measure.
>  
>  
>  
> See the "Fix" section in 
> https://wiki.jenkins-ci.org/display/SECURITY/Jenkins+Security+Advisory+2013-01-04
>   
>
>   
>  
> -- 
> Kohsuke Kawaguchi 
>   

RE: JNLP slave no longer works

[Matthew.Webber]

From a message I sent to the list a few days ago, which worked for at least 
some people:

>>>
I was in a slightly different situation (linux, slaves launched manually), but 
had the 403 as well. Fixed by going to Manage Jenkins ->  Configure Global 
Security, and under  Project-based Matrix Authorization Strategy I had to 
enable “connect” in the “slave” section, for user “Anonymous”.
<<<

I’m not sure if that address your problem, of what the security implications 
are, but anyhow …

Matthew Webber

From: jenkinsci-users@googlegroups.com 
[mailto:jenkinsci-users@googlegroups.com] On Behalf Of 
stuart.gr...@doccentrics.com
Sent: 11 January 2013 10:39
To: jenkinsci-users@googlegroups.com
Subject: Re: JNLP slave no longer works


Thanks for the info, but how is this change applied to a linux slave?

On Tuesday, 8 January 2013 02:50:39 UTC, Kohsuke Kawaguchi wrote:
This is unfortunately a necessary measure.

See the "Fix" section in 
https://wiki.jenkins-ci.org/display/SECURITY/Jenkins+Security+Advisory+2013-01-04


--
Kohsuke Kawaguchi

Re: Issue with Email-Ext plugin-Please help

[uma shankar]

Hi Alex,

Thank you very much for suggestion.

I have used ${JELLY_SCRIPT, template="html"} and getting emails in below
format:

Build URLhttp://localhost:8080/job/test/14/Project:testDate of build:Fri,
11 Jan 2013 13:28:40 +0530Build duration:0.37 sec
*CHANGES*  Revision *9* by *user:**(afsda)*  edit
//streams/local/test2/test.txt  Revision *8* by *user:**(xzvz)*  edit
//streams/local/test/test.txt  edit//streams/local/test2/test.txt

I got what was expected, only thing is not getting the each file version.
and i would also like the build executor name in email.

Could you please help me.


On Fri, Jan 11, 2013 at 12:22 AM, Alex Earl  wrote:

> The default groovy template should have the info.
>
>
> Sent from my Windows Phone
> --
> From: uma shankar
> Sent: 1/10/2013 11:22 AM
> To: Alex Earl
> Cc: jenkinsci-users@googlegroups.com
> Subject: Re: Issue with Email-Ext plugin-Please help
>
> Hi Alex,
>
> I have tried in many ways to get that output, but did not get it through.
> Will you be able to help me getting the template?
>
> Thank You.
>
>
> On Thu, Jan 10, 2013 at 11:34 PM, Alex Earl  wrote:
>
>> You may need to create your own template. Look at the wiki for more
>> information.
>>
>> Sent from my Windows Phone
>> --
>> From: uma shankar
>> Sent: 1/10/2013 10:53 AM
>> To: jenkinsci-users@googlegroups.com
>> Subject: Issue with Email-Ext plugin-Please help
>>
>>  Hi,
>>
>> ** **
>>
>> I am trying to use Jenkins for POC, and have enabled Email-ext plugin. I
>> am facing some issue while getting perforce change list.
>>
>> ** **
>>
>> I am have been using perforce plugin for integration with SCM. I have
>> some issue with the plugin.
>> I use Email-ext plugin and trying get complete change list history in
>> mail with revision no and workspace as display on build home page.
>> Example: (*Expected to be in e-mail*)
>> Changes
>> 1436416 by umashankar@umashakar-test on Jan 10, 2013 8:44:54 AM
>> Forcing a build
>> Affected Files:
>> //myproject/src/test/java/control.java(#16)
>>
>> //myproject/src/test/java/control1.java(#17)
>>
>> ** **
>>
>> ** **
>>
>> But I have used some keyword to get this things in email as below:
>> *
>>
>> List of changed file(s):
>> ${CHANGES, showPaths=true}
>>
>> ** **
>>
>> ${CHANGES_SINCE_LAST_SUCCESS, showPaths=true}
>>
>> *
>>
>> ** **
>>
>> And output comes in mail as below:
>> List of changed file(s):
>> [umashankar] Updates
>> src/test/java/control.java
>>
>> src/test/java/control1.java
>>
>> ** **
>>
>> Changes for Build #16
>> [umashankar] Updates
>> src/test/java/control.java
>>
>> src/test/java/control1.java
>>
>> It is missing the full path, revision number, changelist workspace name.*
>> ***
>>
>> ** **
>>
>> Early response would help to resolve this issue ASAP.
>>
>
>

Re: JNLP slave no longer works

[stuart . green]

Thanks for the info, but how is this change applied to a linux slave?

On Tuesday, 8 January 2013 02:50:39 UTC, Kohsuke Kawaguchi wrote:
>
> This is unfortunately a necessary measure.
>
> See the "Fix" section in 
> https://wiki.jenkins-ci.org/display/SECURITY/Jenkins+Security+Advisory+2013-01-04
>
>
> -- 
> Kohsuke Kawaguchi 
>

Jenkins takes very long to build after upgrading to 1.480.2

[Ilko Iliev]

Hi,

we experience very long build times after we've done the upgrade to
1.480.2. Our infrastructure hasn't changed and as far as I can see no any
resource issues can be seen (cpu, memory usage, swapping). The build time
increased by factor of appr. 3-4 times - a job taking 11 mins before the
upgrade needs now 43 mins to complete. Another one takes 10 mins vs. 3 mins
before the upgrade. Someone else here havind such problems? Any hints to
solve them or what kind of info will be useful to determine the cause of
the problem?

The only exception I see in the log is about the missing POSIX lib, but I
assume it is "normal" and was there before the upgrade as well, considering
our infrastructure - jenkins is running on WAS installed on zlinux:

 [1/8/13 23:15:54:584 CET] 000e1396 SystemErr R Failed to load native
POSIX impl; falling back on Java impl. Stacktrace follows.
[1/8/13 23:15:54:584 CET] 000e1396 SystemErr R
java.lang.NoClassDefFoundError: com.sun.jna.Native (initialization failure)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
java.lang.J9VMInternals.initialize(J9VMInternals.java:168)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
org.jruby.ext.posix.POSIXFactory.loadLibC(POSIXFactory.java:96)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
org.jruby.ext.posix.POSIXFactory.loadLinuxPOSIX(POSIXFactory.java:65)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
org.jruby.ext.posix.POSIXFactory.getPOSIX(POSIXFactory.java:24)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
hudson.os.PosixAPI.(PosixAPI.java:41)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
java.lang.J9VMInternals.initializeImpl(Native Method)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
java.lang.J9VMInternals.initialize(J9VMInternals.java:228)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
hudson.Util.createSymlink(Util.java:1011)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
hudson.model.Run.execute(Run.java:1500)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
hudson.maven.MavenModuleSetBuild.run(MavenModuleSetBuild.java:477)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
hudson.model.ResourceController.execute(ResourceController.java:88)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
hudson.model.Executor.run(Executor.java:236)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R Caused by:
java.lang.UnsatisfiedLinkError: jnidispatch
(/com/sun/jna/linux-s390x/libjnidispatch.so) not found in resource path
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
com.sun.jna.Native.loadNativeLibraryFromJar(Native.java:697)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
com.sun.jna.Native.loadNativeLibrary(Native.java:674)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
com.sun.jna.Native.(Native.java:115)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
java.lang.J9VMInternals.initializeImpl(Native Method)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
java.lang.J9VMInternals.initialize(J9VMInternals.java:228)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
org.tmatesoft.svn.core.internal.util.jna.JNALibraryLoader.(JNALibraryLoader.java:69)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
java.lang.J9VMInternals.initializeImpl(Native Method)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
java.lang.J9VMInternals.initialize(J9VMInternals.java:228)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
org.tmatesoft.svn.core.internal.util.jna.SVNGnomeKeyring.isEnabled(SVNGnomeKeyring.java:104)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
org.tmatesoft.svn.core.internal.util.jna.SVNJNAUtil.isGnomeKeyringEnabled(SVNJNAUtil.java:180)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
org.tmatesoft.svn.core.internal.wc.DefaultSVNPersistentAuthenticationProvider.createPasswordStorages(DefaultSVNPersistentAuthenticationProvider.java:83)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
org.tmatesoft.svn.core.internal.wc.DefaultSVNPersistentAuthenticationProvider.(DefaultSVNPersistentAuthenticationProvider.java:69)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
org.tmatesoft.svn.core.internal.wc.DefaultSVNAuthenticationManager.createCacheAuthenticationProvider(DefaultSVNAuthenticationManager.java:352)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
org.tmatesoft.svn.core.internal.wc.DefaultSVNAuthenticationManager.(DefaultSVNAuthenticationManager.java:80)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
org.tmatesoft.svn.core.wc.SVNWCUtil.createDefaultAuthenticationManager(SVNWCUtil.java:202)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
org.tmatesoft.svn.core.wc.SVNWCUtil.createDefaultAuthenticationManager(SVNWCUtil.java:154)
[1/8/13 23:15:54:585 CET] 000e1396 SystemErr R  at
org.tmatesoft.svn.core.wc.SVNWCUtil.createDefaultAuthenticatio