0

I'm currently learning about CIDR, subnets and IPv4 classes.

I understood that IPv4 classes were used to route traffic in the beginning but nowadays we use "dynamic" subnet masks with the CIDR notation.

Let's say I'm a big corporation in Russia and I need about 2 millions IP addresses for my company's network. I go to the RIPE and ask them while justifying my need. They proceed to give (sell?) me the IP address 98.96.0.0/11 which grants me usage of the IP range 98.96.0.0 - 98.127.255.255.

Assuming all I said before is correct, what steps exactly does the RIPE takes to concretize this new ownership?

My guess is that they operate the "big boys" servers that route traffics on a regional basis and will update the routing table to forward everything falling into 98.96.0.0/11 to our servers but how do they do that?

For such a big address pool I assume we don't go trough an ISP so how do they know which router to forward traffic to? Do we... physically plug our router into themselves? I'm confused...

As I'm still learning, if I said something that was incorrect, please tell me.

1
  • 1
    "I'm a big corporation in Russia and I need about 2 millions IP addresses for my company's network. I go to the RIPE and ask them while justifying my need." Then you are too late because the RIRs have no more IPv4 addresses to assign. ARIN and RIPE are down to 0, and LACNIC has only 256 addresses left. APNIC and AfriNIC have some, but they are reserved for helping migration to IPv6. You will need to go on the open market for IPv4 addresses, but you still need RIR permission to get them assigned to you. – Ron Maupin Apr 29 at 17:42
2

From a RIR perspective, who has what address space is just paperwork. They don't police who is using what. When a bad actor announces someone else's address space, the proper holder can wave that paperwork in a vain attempt to get them to stop. This hijacking happens often, usually to otherwise unused/abandoned space no one will notice, and more importantly, blocks for which no one is likely to produce a letter-of-authority.

As far as how everyone knows where anything is, that's what BGP does. Once you have your properly authorized address block, you setup BGP on your router(s) to announce that block to your peers. (or you give a LOA to your ISP(s) to announce it for you.)

3

RIPE publishes a database indicating who is authorized to use address space. Other RIRs have similar databases. Outside entities also operate IRR databases for routing information. These systems are fragmented for historic reasons.

When BGP-speaking ISP networks exchange routes, often, these routes are filtered using prefix-lists generated based on these databases.

An updated mechanism includes RPKI -- cryptographic verification of who is allowed to announce address space.

3
  • But how do one get its newly assigned traffic when not going through an ISP? I assume Google for example does not go trough an ISP for its core services. How do they get packets inside their IP range forwarded to their servers? – geauser Apr 29 at 19:00
  • 1
    Of course you're going to go through other ISPs -- aka. peers. Your router isn't THE INTERNET. It will have to be connected to others for anyone else to reach you. You share routing information with your peers, and they to theirs, and so on. Thus, the global internet is created. – Ricky Apr 29 at 19:25
  • 1
    @geauser, the Internet is just companies (an ISP is a company) connecting to each other. There is no Internet backbone to which you connect. You peer with other companies, including ISP because the ISPs have the largest networks. A company needing two million addresses will peer with multiple ISPs. – Ron Maupin Apr 29 at 20:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.