You are here: version6.ru » Welcome! » Dividing the Indivisible
Differences
This shows you the differences between two versions of the page.
en:dividing-the-indivisible [2013-04-01 17:39 UTC] rm style |
en:dividing-the-indivisible [2021-09-08 20:23 UTC] (current) rm |
||
---|---|---|---|
Line 1: | Line 1: | ||
====== Dividing the Indivisible ====== | ====== Dividing the Indivisible ====== | ||
+ | //Initially published on 2013-04-01.// | ||
===== How to live with a /64, if that's all they gave you? ===== | ===== How to live with a /64, if that's all they gave you? ===== | ||
Line 11: | Line 12: | ||
===== Network configuration ===== | ===== Network configuration ===== | ||
- | Let's assume we have the following. | + | Let's assume we have the following on the router. |
- | | IPv6 routed subnet from the provider: | ''2001:db8:abcd:cdef::/64'' | | + | | An IPv6 subnet from the provider: | ''2001:db8:abcd:cdef::/64'' | |
- | | One primary internal interface, where we want SLAAC to function: | ''eth0'' | | + | | NIC on the primary internal network, where we want SLAAC to function: | ''eth0'' | |
- | | Three (or more) secondary internal interfaces, which also \\ need IPv6, but they will have to live without SLAAC: | ''eth1'', ''eth2'', ''eth3'' | | + | | NICs on three (or more) secondary internal networks, which also \\ need IPv6, but these will have to live without SLAAC: | ''eth1'', ''eth2'', ''eth3'' | |
What we do is... | What we do is... | ||
Line 80: | Line 81: | ||
Simple. | Simple. | ||
* ''eth0: 2001:db8:abcd:cdef::1/64'' | * ''eth0: 2001:db8:abcd:cdef::1/64'' | ||
- | * ''eth1: 2001:db8:abcd:cdef:ffff:ffff:fff1::1/112'' | + | * ''eth1: 2001:db8:abcd:cdef:ffff:ffff:fff1:1/112'' |
- | * ''eth2: 2001:db8:abcd:cdef:ffff:ffff:fff2::1/112'' | + | * ''eth2: 2001:db8:abcd:cdef:ffff:ffff:fff2:1/112'' |
- | * ''eth3: 2001:db8:abcd:cdef:ffff:ffff:fff3::1/112'' | + | * ''eth3: 2001:db8:abcd:cdef:ffff:ffff:fff3:1/112'' |
+ | |||
+ | ===== On the clients ===== | ||
+ | On the GNU/Linux clients connected to the "primary" LAN, the following change may be required for them to start accepting the extra more-specific route: | ||
+ | |||
+ | <file>echo 108 > /proc/sys/net/ipv6/conf/[INTERFACENAME]/accept_ra_rt_info_max_plen</file> | ||
===== Address collisions? ===== | ===== Address collisions? ===== | ||
Hosts from your primary LAN assume that the whole /64 is their own personal playground, and that they can pick any address they want. In theory this can lead to address collisions with some hosts on your secondary networks. But: | Hosts from your primary LAN assume that the whole /64 is their own personal playground, and that they can pick any address they want. In theory this can lead to address collisions with some hosts on your secondary networks. But: | ||
- | - We ruled out possibility of collision with a EUI-64 address by using ''ff:ff'' in the middle of our /108 (those will always have ''ff:fe''); | + | - We have ruled out the possibility of collision with a EUI-64 address by using ''ff:ff'' in the middle of our /108 (those will always have ''ff:fe''); |
- The probability of a "Privacy Extensions" address suddenly having ''ffff:ffff:fff'' in it, is... well, calculate yourself, seems "pretty low" to me. | - The probability of a "Privacy Extensions" address suddenly having ''ffff:ffff:fff'' in it, is... well, calculate yourself, seems "pretty low" to me. | ||
- |
en/dividing-the-indivisible.1364837944.txt.gz · Last modified: 2013-04-01 17:39 UTC by rm