Given a structure like this:

parallel (
// job 1, 2 and 3 will be scheduled in parallel.
{ retry(3)

},
{ retry(2)

}
)

If I abord manally the main build, current builds for job1 and job2 will abort, but retry will ... retry to builds these jobs even though everything is aborted.

There is a 'worstAllowed' parameter, but given that ABORT is the 'worst' Result, I don't think I can catch it without accepting all other failures.

It feels like upon failure, retry should check the state of the current build before rescheduling a new build.