The "*agent any*" declaration at the top level of your pipeline indeed
allocates an executor for the whole pipeline.
You can avoid this allocation with something like :
pipeline {
agent none
parameters {
choice(choices: 'SIT\nUAT', description: 'What environment?', name:
'environme
Yeah. Thats right. Not tied to using that though. Not sure why I need to
push to using it yet though.
pipeline {
agent any
parameters {
choice(choices: 'SIT\nUAT', description: 'What environment?', name:
'environment')
}
tools {
maven 'maven 3.3.9'
//j
It seems like you are using the Declarative way to write your pipeline
right ? Can you share the full script please. I guess that you are
allocating an agent at the top level.
2017-07-04 11:06 GMT+02:00 paul b :
> Brill. I have managed to get that to work. I have the following stage
>
>
Brill. I have managed to get that to work. I have the following stage
stage ('test') {
steps {
echo 'test it.'
input 'Ready to go?'
}
}
The problem we have now is that the build consumes one of the executors. I
guess I can farm that off somehow???
On Tuesday, 4 July 2017 0
Brill. Thanks. I will look into it.
On Tuesday, 4 July 2017 08:59:13 UTC+1, mpapo - Michael Pailloncy wrote:
>
> Yeah, seems like you want to use the *input* step (interaction through
> Jenkins) : https://jenkins.io/doc/pipeline/steps/pipeline-input-step/
> You can find a simple example here wit
Yeah, seems like you want to use the *input* step (interaction through
Jenkins) : https://jenkins.io/doc/pipeline/steps/pipeline-input-step/
You can find a simple example here with some explanations :
https://github.com/jenkinsci/pipeline-plugin/blob/master/TUTORIAL.md#pausing-flyweight-vs-heavywei
I would like to use the multibranch pipeline plugin and wondered if it can
be configured such that when it gets to a stage, lets says deploying to
environment, it then interacts with a user. That interaction being a
through the jenkins, email or even jira! Then after the user interaction
resp
