Why do we need 6PE and what is the basic idea to solve the problem?
if we have a IPv4 MPLS backbone and we want to connect to it some IPv6 networks the have to change the MPLS LERs.
The routers inside the backbone are not a problem because they look at the label but the router on the edge have to look at the address because the label is removed
idea: change the LER’s
Which components do we have in 6PE?
- P routers: they are the LSRs
- PE routers: they are the LERs
- CE router: it is the router that connects a PE to the DG of the IPv6 network
Which modification is made on LERs?
They become dual stack routers so they can both understand IPv4 and IPv6
Which is the relationship between PEs and Ps?
P routers know every PEs and they have a direct LSP to each of them (?)
How does a PE works with an external IPv6 network?
- the PE uses IPv6 standard protocols to communicate with the CE. Thanks to this PE discovers the needed addresses in the IPv6 network.
- Once it discovers them it communicates them to the other PEs saying “ill be your next hop if you want to reach this IPv6 location”
To advertised the other PEs it uses MP-BGP and with this protocol it can communicate both that it is the next hop and the label it has decided - PE1 advertises all its CEs neighbors about how to reach the discovered locations of the IPv6 network using IGP
Why PEs communicate with MP-BGP?
because it allows the communication between two routers that are not directly connected
If we have IPv6 - PE1- PE2 - IPv6, which has to be the address for the next hop in PE1?
It has to be PE2’s IPv6 address that is of the type unicast mapped IPv4
With the type of network we saw as example with 6PE, which protocols are used to distribute IPv4 and IPv6 labels?
- inside the backbone: IPv4 labels are distributed using LDPv4
- between PEs: IPv6 labels are distributed with mp-iBGP4
