Yet another Hacker News

27 points Throwaway_sys 2 days ago 10 comments

don't want to file a decom report with a gap so I figured I would ask here.

On a contract job clearing out a data center doing routine stuff like taking inventory and audits before we decommission hardware. The issue is there is one node that keeps coming back that isn't in the documentation. ip is in the 46.28.x.x range Its not in the facilities registry though. Ran it through RIPE and ARIN to find nothing.

The latency is what is getting me though. 0.4 round trip every time. Tested from multiple machines including a phone on LTE to get the same response time. That should theoretically mean I am right next to the machine which doesn't make sense across three different connections.

Checked the physical hardware and it's nothing I've ever seen before. Not standard 1U or 2U ports maybe proprietary. serial format is:

CC-[4 digits]-[2 digits]-[6 alphanumeric]

CC prefix doesn't math Cisco, Ibm, Dec, 3com or anything. went back through the facility's historical logs. node appears in their earliest available records, which go back to 1994. facility was built in 1997.

has anyone seen a CC- serial prefix before? or have an explanation for the latency consistency?

mmastrac 1 hour ago | parent

CyberChron. If you don't need to know, don't ask.

And you're also assuming that all the pings are being returned by this box.

cbarrick 1 hour ago | parent

Well now I gotta ask. What's up with CyberChron?

The only thing I can find on Google is a website straight out of 1999 and lawsuit from 1995. They're obviously a US military contractor, but that's all I can tell.

devmor 1 hour ago | parent

Is it in what looks like a luggage/waterproof case? If so, that’s milspec networking hardware.

bombcar 1 hour ago | parent

Could the latency consistency be something designed to make it difficult to pinpoint its location? It sounds like you found the hardware and are just wondering what it is?

Well ain't this place a geographical oddity! 0.4ms from everywhere!

VladVladikoff 1 hour ago | parent

Wouldn’t that mean all upstream providers are all in on it? Participating in the charade and returning ICMP packets preemptively?

VladVladikoff 1 hour ago | parent

Lots of ASNs in 46.28.0.0/16 What’s the actual netblock?

protocolture 46 minutes ago | parent

>The latency is what is getting me though. 0.4 round trip every time. Tested from multiple machines including a phone on LTE to get the same response time. That should theoretically mean I am right next to the machine which doesn't make sense across three different connections.

It means your 3 different connections have decent connectivity to whatever host currently responds to ping for that IP. You cant really derive much more than that from a ping. If it has been there since 1994 it might have been decommed and the IP reassigned. I would suggest a scream test to be honest, especially if you have orders to remove it anyway, seeing if the pings stop responding when you remove the power or networking will tell you more.

bananamogul 45 minutes ago | parent

Would looking at the device's MAC address (which you can get from arp) help? That would give you at least the manufacturer of the network interface.

This is assuming you're on the same subnet.

userbinator 36 minutes ago | parent

The latency is what is getting me though. 0.4 round trip every time. Tested from multiple machines including a phone on LTE to get the same response time.

https://en.wikipedia.org/wiki/Anycast

Animats 28 minutes ago | parent

Is this the hardware you have? [1]

That's some kind of encryption box. It has a "zeroize" button, to clear the keys in an emergency. It might have something that forces uniform latency to make traffic analysis more difficult. Some cryptosystems are totally synchronous, and send random bits at a constant rate when there's no data.

[1] https://www.artisantg.com/TestMeasurement/89462-1/Cyberchron...