Hi list, Some days ago I updated my centos 7.3 to 7.4 on a dell server. After this update I noticed a strange behaviour by sshd. I noticed this problem because trying to connect to ssh I got connection refused or high latency between connecting and password req.
So I noticed that sshd will be restarted every ~2 min.
From log I get this:
Received signal 15; terminating Failed to start OpenSSH server daemon Unit sshd.service entered failed state sshd.service failed sshd.service holdoff time over, scheduling restart ....and sshd restart.
This happens so fast that if I run systemctl status I can't get the failed status.
Logs does not help also in debug facility. There is not any cronjob that kill sshd. I tried also to change some basic sshd_config directives and when running systemctl restart sshd it hangs until I type crtl-C and from log I see that sshd is restarted.
In sshd_config there aren't errors and the config is the same since 2016. On another machine in the same network I have the same configuration and same software installed (replica with native c7.4) but the problem does not occurs. I checked also md5 of relative server /usr/sbin/sshd and they match.
Selinux is disabled and using non standard port.
I notices also that if I'm connected to sshd and sshd crashes/restart my current session is not killed.
I have not stopped and restarted sshd because this is a dedicated server and until my provider does not attach a kvm switch I can't do other tests. I will try also to stop sshd and run from console using debug hoping that helps.
What I can check to know the reason of this behaviour? This is sshd problem or other?
Thanks in advance.