6-6 - Network Layer - OSPF and BGP Notes

Network Layer Functionality Recap

• Data Plane vs Control Plane
  • Data Plane: Handles packet forwarding based on a forwarding table.
  • Control Plane: Manages routing by implementing routing protocols and algorithms.

Routing and its Challenges

• Routing Algorithms

  • Common algorithms include Dijkstra's and Bellman-Ford.
  • Idealized assumptions:
  • All routers are identical.
  • Network is “flat.”
  • Real-world limitations:
  • Storage capacity of routing tables.
  • Excessive link-state information exchange can overload links.

• Administrative Autonomy

  • The Internet consists of multiple autonomous networks (Autonomous Systems - AS) managed independently.
  • Each AS may enforce its own routing policies affecting routing scalability.

Autonomous Systems (AS)

• Definition
  • An AS is a collection of routers under a common administration, also referred to as a domain.
• Routing Types:
  • Intra-AS Routing: Routing within a single AS.
  • Inter-AS Routing: Routing between different ASes.

Intra-AS Routing

• Requires all routers in an AS to use the same intra-domain protocol.
• Gateway routers connect the AS to other ASes and manage routing to external networks.

Inter-AS Routing

• Gateway routers handle external traffic and enforce routing policies based on administrative control.

OSPF (Open Shortest Path First)

• A link-state routing protocol.

• Characteristics:

  • OSPF is an open and publicly available protocol.
  • Routers exchange link-state advertisements (LSAs) directly over IP.
  • Each router maintains a complete view of the topology within the AS and uses Dijkstra’s algorithm for path computation.
  • Supports different metrics for link costs (e.g., bandwidth, delay).
  • Security measures include authenticating OSPF messages to prevent malicious activity.

• Scalability:

  • OSPF uses a hierarchical structure to enhance scalability:
  • Backbone and Areas: Link-state advertisements are contained within specific areas rather than flooding the entire AS, thus limiting routing update traffic.

BGP (Border Gateway Protocol)

• The protocol used for inter-domain routing in the Internet.

• Types of BGP:

  • External BGP (eBGP): Communicates between different ASes, exchanging reachability information.
  • Internal BGP (iBGP): Distributes reachability information among routers within a single AS.

• Path Vector Protocol:

  • BGP advertises routes by providing a list of ASes (AS-PATH) through which data will travel, allowing routers to make policy-based routing decisions.

• BGP Attributes include:

  • Prefix: The destination being advertised.
  • AS-PATH: The journey through which the advertisement has traveled.
  • NEXT-HOP: The specified router for the next segment of the journey.

BGP Policies vs OSPF

• Policy-oriented Routing:
  • Inter-AS routing (BGP) gives administrative control over how traffic is routed and can prioritize certain paths based on policy rather than just performance.
  • In contrast, intra-AS routing focuses primarily on optimizing performance.

OSPF vs BGP Comparison

Example Setup

• All area routers run OSPF, while only selected routers connect with BGP to manage inter-AS communication.
• Area Border Routers (ABRs) and AS Border Routers (ASBRs) play key roles in facilitating communication between different areas and ASes respectively.

Conclusion

• Understand the distinctions between OSPF for intra-domain routing and BGP for inter-domain routing.
• Recognize how hierarchical routing adds structure and helps manage the complexity and scalability of large-scale networks.
• Stay familiar with BGP's dependence on policy for routing decisions as opposed to the performance-focused approach of OSPF.