IPv6 data as JSON

Copyright and Permissions

This data is copyright APNIC.

The IPv6 data is made available on the usual 'hold harmless' basis with attribution.

Data Sources

• v6stats/v6as/
• v6stats/v6economyas/
• v6stats/v6economy/
• v6stats/v6organisation/
• v6stats/v6region/

Each source represents a data collection sub-type.

v6economyas are the data for each ASN seen in the experiment, broken out by economy. So, an ASN eg AS1221 (Telstra) May be seen originating prefixes from Australia or the USA so there exists a record for v6stats/v6economyas/AU/AS1221.json and v6stats/v6economyas/US/AS1221.json if the ASN is in both Australia and the USA.

v6as is the data by ASN for all economies, for example v6stats/v6as/AS1221.json

v6economy is the data by ISO3166 economy.

v6organisation is the data by various organisations eg the RIR, G8, G20 &c.

v6region is the data by UN (sub) region, mapped to private-use codes in the ISO3166 model.

• QM Northern Europe
• QN Southern Europe
• QO Western Europe
• QP Australia and New Zealand
• QQ Melanesia
• QR Micronesia
• QS Polynesia
• XA World
• XB Africa
• XC Americas
• XD Asia
• XE Europe
• XF Oceania
• XG Unclassified
• XH Eastern Africa
• XI Middle Africa
• XJ Northern Africa
• XK Southern Africa
• XL Western Africa
• XN Caribbean
• XO Central America
• XP South America
• XQ Northern America
• XR Central Asia
• XS Eastern Asia
• XT Southern Asia
• XU South-Eastern Asia
• XV Western Asia
• XW Eastern Europe

Data Format

{
  "copyright": "(C) APNIC Pty/Ltd. re-use with attribution permitted",
  "description": "bulk data from IPv6 measurement as a time series with raw and 30, 60 & 90 day simple smoothing",
  "docs": "https://labs.apnic.net/ipv6-data.html",
  "data": [
    { "10":      {"seen": 0.0, "preferred": 0.0, "preferred_pc": 0.0, "capable": 0.0, "capable_pc": 0.0}, 
      "30":      {"seen": 0.0, "preferred": 0.0, "preferred_pc": 0.0, "capable": 0.0, "capable_pc": 0.0}, 
      "60":      {"seen": 0.0, "preferred": 0.0, "preferred_pc": 0.0, "capable": 0.0, "capable_pc": 0.0}, 
      "90":      {"seen": 0.0, "preferred": 0.0, "preferred_pc": 0.0, "capable": 0.0, "capable_pc": 0.0}, 
      "raw":     {"seen": 0,   "preferred": 0,   "preferred_pc": 0.0, "capable": 0,   "capable_pc": 0.0}, 
      "type":    "v6as", 
      "date":    "2010-03-20", 
      "as":      1221, 
      "cc":      "AU", 
      "updated": "2018-06-04:04:18:16"
    },
    :
    :
    :
    :
    :
  ]
}

The data is presented as a JSON object with a simple header, and then an array of data elements. Each data element is a JSON object.

"raw" sample counts and instances are integer and are promoted to float in smoothing. percentages "_pc" are float.

"ASN" present as integers.

Economy, organisation and UN region names present as strings with optional spaces. Economy codes are derived from ISO3166 and region labels from the UN regional model.

Header Format

At this time, the header has the following named elements. The names may change, the "data" name will be preserved so that data can be accessed in code persisting in revisions to the file format (if possible).

• "copyright": simple assertion of copyright over the data
• "description": a one line description of the sub-type of the data
• "docs": a URL pointing to this file
• "data": the data of this collection, as an array of data elements. Each data element is a JSON object.

Data Element Format

The data elements represent IPv6 measurement for the given context, at a given time.

• "type": sub-type of this data. matches the URL path.
• "date": YYYY-MM-DD format date (UTC) of data in the time series.
• "as": (only present in AS data) the numeric ASN of the records.
• "cc": the ISO3166 economycode, augmented by private values for UN regions and sub-regions.
• "updated": date of production of this record.

For each ASN, since it may present in multiple economies, the specific economy is listed.

For each time interval, the data is presented in five versions:

• "raw": or unsmoothed: the measurement actually made on that date.
• "10": measurements smoothed in a 10 day window.
• "30": measurements smoothed in a 30 day window. This approximates to 'monthly'.
• "60": measurements smoothed in a 60 day window.
• "90": measurements smoothed in a 90 day window. This approximates to 'quarterly'

each version has the same basic sub-elements. Weight is not applied to ASN at this time, but is visible in Economy data. The weights are also shown as smoothed but are applied to raw.

• "weight": the imputed weight of this data series, under an APNIC derived population/internet-users model for the world (not in all data)
• "seen": count of experiments from this source.
• "preferred": count of experiments returning on IPv6 from a dual-stack label (A/AAAA) (under happy eyeballs if implemented by the client).
• "preferred_pc": percentage of "preferred" in "seen".
• "capable": count of experiments which can fetch an IPv6 only element.
• "capable_pc": percentage of "capable" in "seen"