|
After a job failed there should be a possibility to run the next job.
My contribution to FlowDSL.groovy :
def buildAlways(Map args, String jobName) {
// ask for job with name ${name}
JobInvocation job = new JobInvocation(flowRun, jobName)
def p = job.getProject()
listener.logger.println("Trigger job " + HyperlinkNote.encodeTo('/'+ p.getUrl(), p.getFullDisplayName()))
Run r = flowRun.schedule(job, getActions(args));
listener.logger.println(HyperlinkNote.encodeTo('/'+ r.getUrl(), r.getFullDisplayName())+" completed")
return job;
}
def build(Map args, String jobName) {
if (flowRun.state.result.isWorseThan(SUCCESS)) {
fail()
}
buildAlways(args, jobName)
}
My workaround is:
parallel (
{
build("aaa")
flowRun.state.result = hudson.model.Result.SUCCESS
build("bbb")
flowRun.state.result = hudson.model.Result.SUCCESS
build("ccc")
},{
build("dddd")
flowRun.state.result = hudson.model.Result.SUCCESS
build("eeee")
flowRun.state.result = hudson.model.Result.SUCCESS
build("ffff")
} )
I run visual tests. Every test should run, even if there are errors.
Useful plugin! ![]()
|
