I came across this IP address: 154.30.27.228
As of Feb 17, 2024, we say this IP address is located in Ashburn, Virginia, United States.
https://ipinfo.io/154.30.27.228/json
{
"ip": "154.30.27.228",
"city": "Ashburn",
"region": "Virginia",
"country": "US",
"loc": "39.0437,-77.4875",
"org": "AS205659 UAB code200",
"postal": "20149",
"timezone": "America/New_York",
"readme": "https://ipinfo.io/missingauth"
}
However, some providers say the IP address is located in Djibouti.
Provider #1
Provider #2
Provider #3
Provider #4
Provider #5
Djibouti and Ashburn are a LOOOONG way apart. Five other IP geolocation providers say the IP address is located in Djibouti, while we are the only one who point to a location eleven thousand kilometers off!
Not agreeing with the majority does not mean we are inaccurate. In this instance, we are correct while everyone else is wrong. I will leave the rhetoric alone and go through the evidence.
Ping based geolocation
We use ping/probe geolocation backed by our proprietary ProbeNet infrastructure. This IP address is pingable. We can use a third-party ping provider to verify the location. We ping the IP address from multiple servers, then we can sort the average time and see which locations have the lowest RTT.
We can ping the IP addresses from a site like Ping.sx, which has the facility to ping an IP address from multiple locations. After the ping operation, we can sort the resulting table by average RTT in descending order. The closer a server is to the target IP address, the lower the average or minimum RTT is going to be.
Data from ping.sx
Note that Ping.sx does not have a probe server in Virginia; however, servers in the United States have the lowest ping times.
Also, there is a special note here: we use a combination of ping data and public internet records. Our ping data was able to verify the location information shared in the geofeed. By using a combination of ping and up-to-date internet records, we can ensure the highest accuracy possible.
Internet public records
From internet public records like Geofeed and WHOIS data, we can see that the IP address belongs and returns the following information.
PSINet, Inc. COGENT-153-30-16 (NET-154-30-0-0-1) 154.30.0.0 - 154.30.255.255
Aventice LLC AVENTICE-CGNT-NET-11 (NET-154-30-0-0-2) 154.30.0.0 - 154.30.31.255
Taking a closer look at the organization with a smaller range, we can see that the country of the organization points to the US. This WHOIS information can also be found on IPinfo: 154.30.24.0/22 IP range details - IPinfo.io
The organization even links the geofeed data in their WHOIS data.
I have no clue why the other geolocation provider chose Djibouti when the geofeed data points to the US, as the headquarters of the organization is in the US, and the most obvious evidence, ping data, points to the US. There is evidence that the IP address is located in the United States.
Maybe it is stale data, maybe it is from some Afrinic regional office, or maybe other providers just put that location just because. At the end of the day, they have placed the IP address 11,000 kilometers away from its actual location. Once some industry leaders did that, other providers blindly followed them, resulting in inaccuracies in consensus.
It may sound strange to say this, but consensus holds little significance in IP geolocation. We utilize probe-based techniques, while much of the industry relies on self-reported unverifiable public records. Just because a location is commonly reported by a bunch of geolocation provider does not mean it is correct. Individuals and large organizations alike use various geolocation data sources and often select locations based on the most frequently reported ones and here we can see how wrong this strategy is.
It must be noted that selecting from multiple providers and reaching a consensus does not guarantee accuracy. Our focus is on being right rather than conforming. Therefore, if you opt for the majority-reported locations, you may not always be choosing the accurately reported location data by IPinfo.
