0

I have a CentOS 7.3.1611 VPS which runs plesk Onyx 17.5.3

Last time I tunneled to access my admin panel with this command:

ssh -fNL 2546:localhost:8443 myvps sleep 3600

(Now I realize it's wrong to put both -N and sleep, and delay is overkill) Which worked fine, but since I cannot access my VPS through ssh, tunnel or not, I instead get a "Connection timed out" message after a waiting time (~5min).

I rebooted my VPS and both SELinux, Fail2ban and mod_security are on "permissive" mode, so they don't enforce rules. With ps ax | grep ssh I find sshd running with -D option, what does it mean?

Also I've followed this article about ssh troubleshooting:

  • netstat -rn | grep UG gave me eth0 as my gateway, which is my public ethernet interface
  • 'tcpdump' capture, receive and don't drop packets from my ssh attempts
  • permissions are 700 ~/, 700 ~/.ssh, 600 ~/.ssh/authorized_keys
  • running /usr/sbin/sshd -d an another port doesn't capture any incoming packets while I'm trying to connect on this other port

I've checked my firewalls and the needed ports are open, even though nmap only find them as "filtered".

Also I've tried to disable tcp_window_scaling but that doesn't change a thing.

If anyone has an idea about what to do right now it would be very welcome, thanks!

raaaahman
  • 13
  • 7
  • What about `firewalld`? – Jakuje May 02 '17 at 18:13
  • You're right, it was a firewalld related problem: I changed `eth0` from my custom "pleskserver" zone to the default configured "public" zone and now it's working again. But when I `--list-all` the two zones, I see no differences except the plesk specific ports I opened for the "pleskserver" zone ( 6308/tcp and 8447/tcp ). Has anyone got an explanation for this? – raaaahman May 04 '17 at 08:59

0 Answers0