On Fri, May 2, 2014 at 11:56 AM, John Park <john.p...@sightlinesystems.com> wrote: > Silly question, does the build.xml that's on the slave have to be on the > master machine's root workspace directory in order for it to run on the > slave?
No, the slave has its own workspace and doesn't have anything to do with the filesystem the master sees. It can be confusing because the master web interface shows a 'workspace' folder icon with files - but that is actually a view from the node that did the last build. > I made changes to the node configuration based on your pointers, and re-ran > the project. On this try, I gave an explicit path to the build file, then > got this message: > > Started by upstream project "master_slave_test" build number 51 > originally caused by: > Started by user John Park > [EnvInject] - Loading node environment variables. > Building on master in workspace That wouldn't happen if your job has a 'restrict where this job can run' set to a slave node. > <http://build:80/job/master_slave_test/default/ws/> > FATAL: Unable to find build script at > <http://build:80/job/master_slave_test/default/ws/build.xml/"C>:\Builds\build.xml" > Build step 'Invoke Ant' marked build as failure > > Seems like there's a syntax error based on the above message, as it looks > to close the C: drive within the URL tag. If your build.xml is in the job's top-level directory you don't need to specify a path. > I then moved the build file to the jennkins/ directory on the slave machine > and re-ran, but got the same message: > > Started by upstream project "master_slave_test" build number 53 > originally caused by: > Started by user John Park > [EnvInject] - Loading node environment variables. > Building on master in workspace > <http://build:80/job/master_slave_test/default/ws/> > FATAL: Unable to find build script at > <http://build:80/job/master_slave_test/default/ws/build.xml> > Build step 'Invoke Ant' marked build as failure Are you not using a source control repository (subversion, git, etc.)? That is the typical and probably best way to handle things since it takes care of the transport of the code/build scripts etc. to whatever node/workspace needs it. Otherwise you'll need some other way to update the files in the workspace of the slave that is going to do the build. > Some other potential useful info: > > The path I gave to ant home and java home when configuring globally was only > to the apache ant and jdk folders, i did not include bin in the path Yes, I think it figures out where to execute ant from the ANT_HOME setting, and that in turn will find java by looking through an assortment of environment settings. > Same goes for configuring the path in the nodes config. > > Based on the error message, the job seems to be executing from the master, > and not the slave. Yes, fix the 'restrict where to build' first. And if your build.xml is at the top of the workspace, just remove your explicit reference to it. -- Les Mikesell lesmikes...@gmail.com -- 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/d/optout.
