Hello! Since yesterday, the ISPConfig URL on my server has been timing out. When I access the ISPConfig URL, my VPS stops responding to the ping command, and the SSH connection responds with "Connection timed out." All websites created in the ISPConfig interface are working normally. But ISPConfig on port 8080 and apps on port 8081 are not. I've tried disabling ufw, stopping fail2ban, and restarting nginx, but it didn't work. The last thing I did with ISPConfig was add a website. And upload, via FTP, about 500MB of music from a local artist to it. I suspect this might have something to do with it. Thank you very much for your help!
Seems more like a firewall problem with your hoster. If websites continue to work normally then the server isn't not responding and isn't having issues with the webserver, but more likely isn't receiving packages other then 80 and 443. My guess is your hoster is blocking the other traffic for whatever reason.
Run a curl from your local system and post the output. Code: curl -v yourispconfigurl.tld:8080 And is that scan run from your local system?
Yes. The scan is done from my personal computer to the provider's VPS. When ping is running in one terminal, running curl in another I get this response: Code: $ curl -v https://server1.example.com:8080/login/ * Host server1.example.com:8080 was resolved. * IPv6: (none) * IPv4: 177.xxx.xxx.xxx * Trying 177.xxx.xxx.xxx:8080... * Connected to server1.example.com (177.xxx.xxx.xxx) port 8080 * ALPN: curl offers h2,http/1.1 * TLSv1.3 (OUT), TLS handshake, Client hello (1): * CAfile: /etc/ssl/certs/ca-certificates.crt * CApath: /etc/ssl/certs * SSL connection timeout * Closing connection curl: (28) SSL connection timeout However, when the call with curl starts, the ping hangs and I can't log in via ssh. Strange because the server doesn't get overloaded. With the ping stuck, this is the response from curl: Code: $ curl -v https://server1.example.com:8080/login/ * Host server1.example.com:8080 was resolved. * IPv6: (none) * IPv4: 177.xxx.xxx.xxx * Trying 177.xxx.xxx.xxx:8080... * connect to 177.xxx.xxx.xxx port 8080 from 192.168.0.xxx port 46974 failed: Tempo esgotado para conexão * Failed to connect to server1.example.com port 8080 after 132727 ms: Couldn't connect to server * Closing connection curl: (28) Failed to connect to server1.example.com port 8080 after 132727 ms: Couldn't connect to server So, the ping gets stuck, about 5 minutes. And when it starts pinging again, I run curl in another terminal and the ping gets stuck again...
Did you install any other software on this server besides the usual ISPConfig stuff? For example any security solution. And when you get these stuck ping requests can you reach your server from another location, your phone when using the mobile network internet for example?
Software was installed only with the ISPConfig auto installer. When I try to connect to port 8080 of the VPS, through a mobile network, the connection times out returning the error ERR_CONNECTION_TIMED_OUT. However, websites created via ISPConfig respond normally...
Clueless about what's happening. For sure it's not the webserver, as other websites continue to function. And that's what makes it such a strange problem. It's not your ip being blocked (fail2ban could do that, which is installed by the auto installer) but specific traffic. And maybe even everything except 80 and 443. Is there any way to monitor traffic going towards the vps right in front of it?
The whole description of the issue sounds really weird. Let's get that straight, so we talk about the correct issue. - You can access any website hosted on this server just fine - You can't access the ISPConfig Webinterace at all - When running a constant ping and then try to access the Webinterface, the ping gets timed out and only recovers after about 30 minutes - Still druing this period you are able to reach the websites that are hosted on this server just fine Is that correct?
So, your VM has possibly ran out of space? That could be the reason why nothing works anymore. Do you remember how your VM was partitioned and how much space was left on the device where you uploaded the music to?
That would most likely have killed all hosted websites too and not just ISPC panel. Also that wouldn't kill ping and no space doesn't recover itself after about 5 minutes doing nothing. Yes as I understand it. Except the 30 minutes are 5 he said in this post. That's an understatement
he says he can't login using ssh when this happens, but that doesn't mean an existing ssh connection gets dropped. hell, i've lost local network connection and an active ssh session continued working when the local network came back.. so it could be worth having a few active ssh sessions open before trying to access the ispconfig interface.. i'd suggest having one running 'tail -f /var/log/syslog' and one running top or htop, possibly also iostat and see if they provide any info on what's happening as the issue starts possibly start those using screen, in case they do keep working but the network or ssh connection has problems, you may still see something useful after reconnecting to the screen session. only other thing that comes to mind is checking journalctl or /var/log/dmesg for clues. considering other websites continue working, the webserver is fine, possibly something specific to php-fastcgi, since that's used by the interface, not a version php-fpm like most sites would be set to, but don't see how that would affect ssh or icmp. is fail2ban definitely stopped? could be a jail monitoring 8080 or 8081 and blocking the source ip (or all ip's) on just those ports and icmp for 300 seconds. actually just re-read the opening post, was the last website added set to use php-fpm? or something else to run php scripts? i think my 1st bit of troubleshooting now would be to ssh to the server, go to /etc/apache2/sites-enabled or /etc/nginx/sites-enabled, delete the symlink for the site you added, and restart apache2 or nginx. and then try accessing the ispconfig interface.
