I have a related agent scope request. Today, the agent is defined inside a
stage. If multiple stages use the same agent and the agent has just one
executor, then concurrent builds alternate stage-by-stage.
To prevent that, I'd like to set an agent for a couple of stages.
Regards,
Bert
On Wedne
Hello,
IMO, you want to file an argued JIRA for each evolution you'd like.
The things you ask for don't appear crazy, but declarative being about
straightforwardness and opinionated, almost by definition nothing will be
added without deep thinking basically. Hence why each addition request
somehow
Can we add *agent* as a child of *steps* so that we can decide which nodes
we want those steps to run on and that way we could then set the agent to
none for the overall stage?
And make *post* also available as a sibling of *stage* so that it can run
after a stage but not require a node?
I'm tr
When I specify an agent inside a stage, it then also applies to everything
in the 'post' section.
I'm not sure that is alway the behavior I want.
Is there a way around this?
Specifically, we have the ability to take a checkpoint (similar to the
enterprise edition checkpoint feature) but that s
