Border Gateway Protocol (BGP)

Topics

DC-BGP is a portable source code implementation of the Border Gateway Protocol (BGP) protocol designed explicitly to support the scalability, availability and functional requirements of OEMs building devices for next generation networks.

DC-BGP shares Metaswitch's unicast IP routing architecture. It is available for both IPv4 and IPv6 networks.

DC-BGP has comprehensive support for the relevant standards, with management interfaces and a design suited for the most demanding applications. Further details are provided below.

The DC-BGP architecture splits the product into modular components. This allows these components to be distributed onto line cards and supports multiple instances as required for virtual routers and VPNs.

DC-BGP Features

BGP Protocol Function

  • The DC-BGP software product includes the following BGP protocol function.
    • Full mandatory BGP v4 protocol support
    • RFC 1771
    • draft-ietf-idr-bgp4
    • IPv4 and IPv6 support
  • VPN support
    • BGP/MPLS VPNs (RFC 2547 and draft-ietf-ppvpn-rfc2547bis)
  • Authentication
    • TCP MD5 signature (RFC 2385)
  • Protocol enhancements including
    • BGP-4 multi-protocol extensions for IPv6 inter-domain routing (RFC 2545)
    • BGP-MPLS IP VPN extension for IPv6 VPN (draft-ietf-l3vpn-bgp-ipv6)
    • Capabilities advertisement (RFC 2842)
    • Route reflection (RFC 2796)
    • BGP communities (RFC 1997) and extended communities (draft-ietf-idr-bgp-ext-communities)
    • Route refresh (RFC 2918)
    • Route flap damping (RFC 2439)
    • Multi-protocol extensions (RFC 2858)
    • Sub-codes for the BGP Cease Notification message (draft-ietf-idr-cease-subcode)
  • DPoE support
    • Full control plane support for an IP Network Element in DPoE (DOCSIS Provisioning of EPON)

Management and Integration

  • Pre-integrated and transactional CLI, NETCONF, WEB UI, SNMP, REST, and scripted management solutions available through pre-integration and partnership with Tail-f ConfD
  • High-level Yang models map directly to operator-familiar semantics
  • Easily integrated with proprietary management solutions via alternative low-level MIB interfaces
    • Support for all BGP v4 MIB tables defined in RFC 1657
    • Extensions based on the latest BGP v4 MIB draft (draft-ietf-idr-bgp4-mibv2)
    • Proprietary extensions to allow MIB-based configuration of BGP routing policies
  • High availability and fault tolerance in management of configuration information (when used with Configuration Safe Store)

Carrier Class Design

  • Routing Table Manager
    • Provides common routing table across all IP routing protocols including BGP, OSPF and ISIS
    • Policies for route import/export between IP routing protocols
    • DC-BGP can be used with a third party Routing Table Manager
  • Scalability and distribution
    • No hard limits
    • Support for at least hundreds of peers
    • Support for at least half a million routes in the routing table
    • Support for at least a thousand Virtual Routing and Forwarding Tables (VRFs)
    • Lengthy routing calculations can be paused to ensure other processing is not blocked by the computationally intensive calculation
    • Multi-instance Routing Table Managers can be distributed to line cards to spread processing load across multiple CPUs
  • Redundancy and high availability
    • Integrated with DCL's high availability framework
    • Uninterrupted packet forwarding across hardware or software failures
    • Control plane restart (draft-ietf-idr-restart)
    • Availability and fault tolerance in management of configuration information possible with Configuration Safe Store