Uncommon IP notations

2019-02-19 06:20发布

问题:

I know that it is possible to write IPs in IPv4 as an integer e.g. 2130706433 instead of 127.0.0.1.

What is the reason for this possibility?

Is there a similar notation for IPv6?

I tried ping -6 1 as a try to ping ::1, but that don't work (the host would not exists).

回答1:

IPv4 addresses can be represented in multiple ways. For example the default loopback IP can be one of:

  • 127.0.0.1
  • 0177.0.0.1
  • 0x7f.0.0.1
  • 127.0.1
  • 127.1
  • 2130706433
  • 017700000001
  • 0x7f000001

The first notation (full 8-bit decimal dotted) is in wide usage, the remaining ones are seldom used but allowed by the inet_addr POSIX standard function. Only the first familiar notation has been retained in the newer inet_ntop/inet_pton POSIX standard functions which process both IPv4 and IPv6 addresses.

With IPv6, 16-bit hexadecimal dotted notation with an optional decimal dotted trailer (for embedded IPv4) and also an optional zero compression is what the standard defines.

eg:

  • 2001:0db8:85a3:0000:0000:8a2e:9370:7334
  • 2001:0db8:85a3::8a2e:9370:7334

There are then still multiple representations of a single address. To avoid the resulting confusion RFC 5952 recommends a canonical form that allows a unique notation.



回答2:

An IPv4 address is just a 32bit number. You could represent it in any way you can represent such a number (decimal, hex, octal). The dotted notation a.b.c.d is just much more practical.

You can do the same thing with IPv6 addresses, except that those are 128bit numbers - even harder to grok in decimal form.

The usual tools will only deal with usual notations. Decimal isn't one of those.