This Server Listens on Every Port

[@KJ7BZC]

Alright so we've got "Shouting in the Datacenter", "We can't send email more than 500 miles", and now "only odd numbered ports were working outbound"...

[@LANRanger]

Odd to see you here! Thanks for watching. Although I will say, I believe I am confused by the comment haha

[@KJ7BZC]

@@LANRanger Guess KZbin doesn't like the links I put in my reply a second ago... I'm just referring to the original port related issue being very strange like the other two I listed. "Shouting in the Datacenter" is a video posted by Bryan Cantrill where a hard disk missing a mounting screw was having bad latency due to vibration. In trying to replicate the issue for a blog post, they were able to shout at the JBOD array and cause just enough vibration for high latency across multiple disks to show up in the analytics software. "We can't send email more than 500 miles" is referring to a forum/message board post made by Trey Harris regarding an issue on the department mail server. Due to a mismatched configuration file with the version of sendmail installed on the server following an upgrade, the connection timeout value was being defaulted to 0. The time it took to connect to mail servers that were within an approximate 500 mile radius was short enough for it to not trigger the very low timeout value, but past that the delay in the networking infrastructure became long enough to cause it to fail.

[@VelcorHF]

Factory reset is scorched earth diplomacy. There can’t be a problem if there’s nothing left. Great idea, will check out the tool.

[@LANRanger]

When everything else checks out, it’s typically a bug. Although, rebuilding from backups is always nice!

[@88tx]

bruh no wonder when i looked up your ip on shodan it lit up like a Christmas tree, but all of the ports have similar responses.

[@thebinaryshadow]

@@88tx all ports being redirected to the same web server serving the exact same thing should always return the same response, so expected behaviour

[@foxfoxfoxfoxfoxfoxfoxfoxfoxfox]

Haha I just did this weeks ago. But not just for a single IP but for a whole internet addressable /24 subnet. I used xinetd to listen on a single port and execute a program to send random ASCII art to any opened connection. I then used iptables to redirect all traffic to that port. Exceptions were added for my management IP to make it through to port 22 on the server. I also intermittently had a webserver running and redirected all port 80 connections to it.

[@LANRanger]

Okay, that’s awesome! I never thought to send pseudo data back, I might have to try that.

[@nurmr]

So what was the reason that half the ports were being blocked?

[@LANRanger]

That’s the thing, it was determined to be a bug with the way the device passed traffic. The recommendation from TAC was to factory reset.

[@Deffcolony]

@LAN Ranger why did you made the Broadcasting my own cellular - it works! | PLTE w/ Open5GS, B210 video private? wanted to send it to my friend but now there is no way... do you have a download link instead?

[@LANRanger]

It’s online now

[@0ka354]

Why not just use tcp traceroute to see where your packet gets lost?

[@LANRanger]

Because traceroute isn’t going to tell me if a service is responding to a specific port.