On Thu, 2 Mar 2017, Tate Belden wrote:
I have to assume it's something unique to your install, then.
I run about a dozen sshd instances on as many CentOS 7 boxes around here. To a one, systemctl restart sshd is all it takes to implement config changes.
Did you actually change something before doing the restart? If nothing changes it will restart every time. If I change the configuration to for instance have sshd listen on another port, that is when I have the problem. Oh and selinux is in permissive.
There is nothing special about the environment. a couple of VM's in question are brand new minimal installs. I spin them up for testing push new sshd configs to them with Ansible and tear them down.
Before someone says Ansible is the problem I have tried this by hand also.
Regards,