It might be a systemd issue, but I am using the official yum repo on Postgres 12, not postgres 13, and not doing anything fancy with systemd, so I doubt it.
@musicman I see that you've resolved it. Did you have to do anything special, or did it just take a long time to start working?
If you're using screen, try "CTRL+A C" and run "systemctl status postgresql-12" to see where it is in the start / restart process. (If you're using tmux, replace CTRL+A with CTRL+B. I haven't used dtach in a while, so I don't recall what command it uses.)