On Wed, Mar 26, 2014 at 4:48 PM, Peter Savage wrote:
> Thats fine, but insisting that I dump the contents of that repository onto
> disk in a specific fashion may or may not help me use the contents. You keep
> insisting that a checkout into a directory controlled by the Jenkins server
> is a nece
Thats fine, but insisting that I dump the contents of that repository onto
disk in a specific fashion may or may not help me use the contents. You
keep insisting that a checkout into a directory controlled by the Jenkins
server is a necessary part of the way one must use the contents. I disagree.
I think a better answer to this is that this is not a typical use case for
people or teams using Jenkins. Although I can see some value to something
like this, it may not have enough support in general to change the behavior
of Jenkins. I had a situation in the past for a similar need. At that
t
On Wed, Mar 26, 2014 at 1:31 PM, Peter Savage wrote:
> "It seems odd to care if the contents of a repository changed if you
> aren't going to use them in the job that is triggered."
>
> That is so loaded with inherent assumptions I don't even know how to touch
> it.
Yes, and it matches the assump
"It seems odd to care if the contents of a repository changed if you
aren't going to use them in the job that is triggered."
That is so loaded with inherent assumptions I don't even know how to touch
it.
On Wed, Mar 26, 2014 at 11:08 AM, Les Mikesell wrote:
> On Wed, Mar 26, 2014 at 12:42 PM,
On Wed, Mar 26, 2014 at 12:42 PM, wrote:
> It seems pretty inflexible. Why not just separate polling from checkout? One
> step to poll repositories, and one step to checkout.
It seems odd to care if the contents of a repository changed if you
aren't going to use them in the job that is triggered
It seems pretty inflexible. Why not just separate polling from checkout?
One step to poll repositories, and one step to checkout.
On Friday, January 17, 2014 1:18:06 PM UTC-8, LesMikesell wrote:
>
> It is pretty much the point of the slave agent and the way jenkins
> works to do all of the remot
It is pretty much the point of the slave agent and the way jenkins
works to do all of the remote work for you. I wasn't sure if you did
not understand that or if you had some special requirement to run your
own remote ant script. If you want some things to be copied back to
the master, add a po
Thank you so much for your explanation!
It seems now easy to define a new slave, trigger svn update on it with SCM
poll, and execute the Ant script from that location.
Don't need even my Ant remote task anymore with jenkins slave agent...
On Thursday, January 16, 2014 10:05:12 PM UTC+1, LesMike
On 16.01.2014, at 15:16, kgiloo wrote:
> AFAIK if you configure the post-commit hook together with jenkins,
> you must activate "poll SCM" anyway, else it won't start the build.
Put the code to determine which change in the repository needs to trigger which
job into the post-commit hook, and
My builds are all done on slaves and even though the master controls
the SCM polling, it does not check out any source locally on the
master. However, with usual setup, the slave checks out a copy of
the target the job is configure to poll before executing anything in
the job.So I don't under
Sorry, i did not get your point... What do you mean exactly?
On Thursday, January 16, 2014 5:31:05 AM UTC+1, LesMikesell wrote:
>
> On Wed, Jan 15, 2014 at 2:42 AM, kgiloo >
> wrote:
> >
> > probably my description lacked precision:
> > I should have titled "A way to get rid of useless checkou
AFAIK if you configure the post-commit hook together with jenkins,
you must activate "poll SCM" anyway, else it won't start the build.
I followed instructions here:
https://wiki.jenkins-ci.org/display/JENKINS/Subversion+Plugin
What i want to avoid is the checkout/update of repo into my workspace
On Wed, Jan 15, 2014 at 2:42 AM, kgiloo wrote:
>
> probably my description lacked precision:
> I should have titled "A way to get rid of useless checkout of repository
> into my local workspace".
>
> I do not use any special slave configuration to trigger my builds on slaves,
> it is all driven by
kgiloo (2014-01-15 09:42):
probably my description lacked precision:
I should have titled "A way to get rid of useless checkout of
repository into my local workspace".
I do not use any special slave configuration to trigger my builds on
slaves, it is all driven by Ant.
In other words, check
kgiloo (2014-01-15 09:42):
probably my description lacked precision:
I should have titled "A way to get rid of useless checkout of
repository into my local workspace".
I do not use any special slave configuration to trigger my builds on
slaves, it is all driven by Ant.
In other words, check
probably my description lacked precision:
I should have titled "A way to get rid of useless checkout of repository
into my local workspace".
I do not use any special slave configuration to trigger my builds on
slaves, it is all driven by Ant.
In other words, checkout of repository is also per
If there's a workspace for these jobs on Jenkins master, but they're running on
slaves nodes due to a label expression in the job configuration, you should be
able to just delete the workspace on the Jenkins master. It's probably left
over from when the job was still running on the master.
On 1
>From my jenkins server, builds are triggered with Ant remotely on building
machines.
I inserted a "Build Trigger" with "Poll SCM" to listen to changes in my SVN
repository,
But as all my builds are run remotely i do not need a checkout of my repo
locally on my jenkins server.
How can i
19 matches
Mail list logo
