Hi, I have a question on pantsd and sandboxes. If I set

run_goal_use_sandbox

for a Python target, is pantsd involved here at all? The reason why I am asking is whether there is a performance drawback when also setting

--concurrent

for that same execution or if that is unrelated?

I believe pantsd is always used unless explicitly disabled, no matter what goal is being run. I’m not sure that’s answering your question, though

Any idea how to find out?

If you kill all pantsd processes (e.g.

pkill -f pantsd

or

rm -rf .pants.d/pids

, I think) and then run the goal, does it spend some time "initializing scheduler"? if so, that's starting pantsd

Actually, setting

concurrent

to true will disable pantsd (see this comment). I learned that the hard way 😄

That is true, but only for the current invocation