Issuing a SIGHUP to reload is certainly worth considering, but should be implemented in a separate branch to avoid bogging this fix down.
I don't know if there is a reason it was done with restart initially, or if this is a simple case of changing the set_state('prometheus.do-restart') to set_state('prometheus.do-reload'). Other handlers already do set_state('prometheus.do-reload').
Issuing a SIGHUP to reload is certainly worth considering, but should be implemented in a separate branch to avoid bogging this fix down.
I don't know if there is a reason it was done with restart initially, or if this is a simple case of changing the set_state( 'prometheus. do-restart' ) to set_state( 'prometheus. do-reload' ). Other handlers already do set_state( 'prometheus. do-reload' ).