What exactly is a P and PE router<\/strong> :<\/p>\n\n\n\n In MPLS, a P router acts as a vital link within the core network, connecting different parts together efficiently. Unlike PE routers, P routers don’t deal directly with individual customers. This streamlined approach makes the network highly scalable because the core doesn’t get bogged down by the number of customers. While P routers handle the core backbone of the network, PE routers, equipped with their distinctive “Edge” functionality, act as the initial points of entry for customer traffic.<\/p>\n\n\n\n Typically, a true P router doesn’t run BGP but only IGP. However, sometimes the terms P and PE routers are used interchangeably, depending on whether routing lookup occurs via IP or Label. However, it’s important to note that a PE router is fundamentally distinct from a P router, as it incorporates the “Edge” functionality.<\/p>\n\n\n\n What looks like a Backbone with only PE routers (Collapsed P\/PE)<\/strong><\/p>\n\n\n\n In the diagram above, there are 4 Points of Presence (POP), each with 2 routers, resulting in a total of 6 physical interconnections to other POPs.<\/p>\n\n\n\n Now, imagine if we have 50, 100, or 200 POPs. We would then have an exponential increase in the number of physical interconnections between the POPs. For instance, with 50 POPs, the number of interconnections would be much higher than with just 4 POPs. Specifically, with 50 POPs, there would be approximately 2450 interconnections. With 100 POPs, this number would grow even further to around 9900 interconnections. This exponential growth in interconnections presents significant challenges in terms of network management, scalability, and maintenance.<\/p>\n\n\n\n What looks like a Backbone without P routers, but with Agreagation switches<\/strong> :<\/p>\n\n\n\n In this design, we can rely on central aggregation switches, which are less expensive than MPLS routers. However, despite the use of switches, the number of IGP adjacencies will remain the same as in the design without switches.<\/p>\n\n\n\n What looks like a Backbone with P routers :<\/strong><\/p>\n\n\n\n Now, consider the design where we deploy ‘P only routers,’ simplifying the architecture and enhancing the scalability, modularity, and maintainability of the backbone. This is because both the physical interconnections and the adjacencies that each PE router needs to maintain will be reduced.<\/p>\n\n\n\n We can also integrate multiple layers of P routers, forming aggregation and core layers, alongside other. The possibilities are virtually endless and vary depending on each constraint and situation (variations that may include aggregation switches…)<\/p>\n\n\n\n You can extrapolate that this principle applies to many other designs, such as having distinct “Access, Aggregation, Core layers” compared to having a collapsed Core and aggregation layer in a campus or Datacenter design.<\/p>\n\n\n\n In summary :<\/strong><\/p>\n\n\n\n Are you need an expert review of your future or upgrade deisgn of your network or do you have any questions about it? We can assist you. You can reach out to us through the\u00a0CONTACT-US<\/a>\u00a0page or by emailing us at\u00a0contact@lastopinion.io<\/a>\u00a0for advise assistance.<\/p>\n\n\n\n![\"\"](\"https:\/\/www.lastopinion.io\/wp-content\/uploads\/2024\/04\/PE_Collapsed_Network-1.png\")
<\/figure>\n\n\n\n![\"\"](\"https:\/\/www.lastopinion.io\/wp-content\/uploads\/2024\/04\/PE_Collapsed_Network_WithSwitch.png\")
<\/figure>\n\n\n\n![\"\"](\"https:\/\/www.lastopinion.io\/wp-content\/uploads\/2024\/04\/PE_Collapsed_Network_OnlyP-1.png\")
<\/figure>\n\n\n\n![\"\"](\"https:\/\/www.lastopinion.io\/wp-content\/uploads\/2024\/04\/PE_Collapsed_Network_WithPE_P_Aggr_NoCore-1-1024x422.png\")
<\/figure>\n\n\n\n\n
\n