On 03.05.2013 14:44, Paul Weiss wrote:
But how do you guarantee that the second job runs on the same slave as
the first, unless you pin both jobs to a single slave?


Weeeell - on first sight, no clue, really. :-)

Then, again, a short search yields:

https://wiki.jenkins-ci.org/display/JENKINS/NodeLabel+Parameter+Plugin

so it appears one should be able to set this up:

In inital job:
* first delete parameter file
* once point reached where we need to clean up:
* write parameter file with the appropriate options and
* with the current node info
* downstream job should be triggered on correct node

I haven't tested this, though.




On May 3, 2013 12:12 AM, "Martin Ba" <0xcdcdc...@gmx.at
<mailto:0xcdcdc...@gmx.at>> wrote:

    On 03.05.2013 03:36, Eric Blom wrote:

        Hello Everyone,

        We started using Jenkins about a year ago and have been very
        happy with it. We keep finding more way to use it! Some of our
        jobs use hardware connected to the slave and when that slave job
        is canceled we need to execute a clean up step, how can this be
        done?

        I'm looking for a solution for both Linux and Windows. On Linux
        I've tried using the trap command, but, I can't find any signal
        that trap can trap. On windows I haven't found any solution yet.

        ...
        Can anyone offer me some advice on how to implement a clean up
        on cancel for Linux and Windows?


    Idea: Move the cleanup to a second job that you run downstream from
    the first one. You start the second job always ("Complete (always
    trigger)") and it will run even when the first job was aborted.

    And if you use the parameterized trigger plugin and load the
    parameters for job#2 from a file, you can have it starting depend on
    the file existing. ("Don't trigger if any files are missing.")

    cheers,
    Martin


--
You received this message because you are subscribed to the Google Groups "Jenkins 
Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to jenkinsci-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.


Reply via email to