Could it be that the errors happen in the activation script,
but not when actually starting nginx? I see in the code that
we appear to pass a “-p” flag already when starting nginx;
maybe we should simply do the same when testing the config
during activation?


Reply via email to