No relevant resource is found in the selected language.

This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read our privacy policy>Search

Reminder

To have a better experience, please upgrade your IE browser.

upgrade

NE40E V800R010C00 Feature Description - VPN 01

Rate and give feedback :
Huawei uses machine translation combined with human proofreading to translate this document to different languages in order to help you better understand the content of this document. Note: Even the most advanced machine translation cannot match the quality of professional translators. Huawei shall not bear any responsibility for translation accuracy and it is recommended that you refer to the English document (a link for which has been provided).
VPN GR

VPN GR

Graceful restart (GR) is a type of high availability (HA) technology, which comprises a comprehensive set of technologies such as fault-tolerant redundancy, link protection, faulty node recovery, and traffic engineering. As a fault-tolerant redundancy technology, GR ensures normal forwarding of data when the routing protocol restarts to prevent interruption of key services. Currently, GR has been widely applied to active/standby switchovers and system upgrade.

GR is usually used when the active route processor (RP) fails due to a software or hardware error, or used when an administrator performs a master/slave main control board switchover.

Prerequisite for GR Implementation

On a traditional router, a processor performs both control and forwarding. The processor finds routes based on routing protocols and maintains the routing and forwarding tables of a device. High- and medium-end devices generally use the multi-RP structure to improve forwarding performance and reliability. The processor responsible for routing protocols is mostly located on the main control board, whereas the processor responsible for data forwarding is located on the interface board. This design helps to ensure the continuity of packet forwarding on the interface board during the restart of the main processor. The technology that separates control from forwarding satisfies the prerequisite for GR implementation.

At present, a GR-capable device must have two main control boards. In addition, the interface board must have an independent processor and memory.

Related Concepts

GR involves the following concepts:

  • GR restarter: A GR-capable router that performs a master/slave main control board switchover upon the occurrence of a failure or under the instructions of an administrator.

  • GR helper: Neighbor of a GR restarter. A GR helper must support GR.

  • GR session: A session over which a GR restarter and a GR helper can negotiate GR capabilities.

  • GR time: Time when the GR helper keeps the topology information or routes obtained from the GR restarter after detecting that the GR restarter is Down.

NOTE:

Currently, the HUAWEI NetEngine40E can only function as a GR helper.

VPN GR Overview

VPN GR is the application of the GR technology on a VPN. VPN GR ensures that VPN traffic is not interrupted when a master/slave main control board switchover is performed on a router that transmits VPN services. VPN GR offers the following benefits:

  • Reduces the impact of VPNv4 route or BGP label route flapping on an entire network during the route processor switchover.

  • Decreases the packet loss ratio to almost 0%.

  • Reduces the impact on important VPN services.

  • Reduces PE or CE single-point failures to improve the reliability of an entire VPN.

To support VPN GR, a BGP/MPLS IP VPN must support IGP GR and BGP GR. When using an MPLS LDP LSP as a tunnel, the BGP/MPLS IP VPN must support MPLS LDP GR. If traffic engineering is used, the BGP/MPLS IP VPN must also support RSVP GR. After the master/slave main control board switchover is performed on a PE or CE, the router and its connected PE can keep the forwarding information of all VPN routes for a certain period to ensure that VPN traffic is not interrupted. In addition, the CE that connects to the PE on which the master/slave main control board switchover is performed also needs to keep the forwarding information of all VPN routes for a certain period.

On a common L3VPN, the master/slave main control board switchover can be performed on the router that functions as a PE, CE, or P.

Master/Slave Main Control Board Switchover of a PE

The master/slave main control board switchover of a PE consists of three phases:

  1. Before the switchover

    The PE negotiates the IGP GR and MPLS LDP GR capabilities with a P, and negotiates the IGP GR or BGP GR capabilities with the connected CE. The PE also negotiates BGP GR capabilities with the peer PE and sends the Open message containing the GR capability field of <AFI=Unicast, SAFI=VPNv4>.

  2. During the switchover

    The PE keeps the status of forwarding VPNv4 routes and the following procedures are involved:

    • MPLS LDP GR

      If a neighbor detects that the corresponding TCP session enters the Down state, the neighbor backs up all LSPs on the slave board and marks these LSPs as invalid.

    • BGP GR

      BGP session messages are lost during the switchover. Then, the PE does not keep any routing information but the forwarding information. GR-aware BGP peers mark all the routes related to the GR routers as Stale. The BGP peers, however, still forward packets based on these routes within the GR time.

  3. After the switchover

    The PE instructs all the IGP neighbors, BGP IPv4 peers, and private network IGP neighbors between the PE and CE to reestablish connections. The following procedures are involved:

    • IGP convergence

      To resynchronize the link state database (LSDB) of OSPF or IS-IS with the neighboring P, the PE sends a signal to each neighboring P and reestablishes the neighbor relationship list after receiving a response. If IS-IS or OSPF multi-instances are run between the PE and CE, the PE also needs to resynchronize the LSDB with the CE. The PE obtains the topology or routing information by establishing sessions with all the neighbors. After obtaining the topology and routing information, the PE recalculates the routing table and deletes the routes in the Stale state to complete IGP convergence.

    • BGP convergence

      The PE also exchanges routing information with BGP peers, including public network BGP peers, MP-BGP peers, and private network BGP peers. The PE then updates the routing table and the forwarding table according to the new routing information and replaces the invalid routing information to complete BGP convergence.

      After receiving the End-of-Rib message from a BGP peer on a public or private network, the PE notifies the routing management (RM) module. The End-of-Rib message is used to notify the peer that the first routing information update after a BGP session is established has been completed.

      Before all routing protocols complete the GR, only FIB information on the main control board is updated, and the FIB information on the interface board is not updated.

    After all routing protocols complete the GR, the RM module sends a message to notify each protocol that the GR is complete, and then updates the FIB information on the interface board. BGP sends BGP public network IPv4 routes, private network IPv4 routes, and VPNv4 routes to each peer. After sending the routes, BGP sends End-of-Rib messages.

The processing on routers connected to the PE is as follows:

  • After detecting the restart of the PE, the CE connected to this PE uses the same processing flow as that of the GR helper in the common IGP GR or BGP GR and keeps information about all IPv4 routes for a certain period.

  • After the P connected to this PE detects the restart of the PE, either of the following situations occurs:

    • If BGP is not configured, the P uses the same processing flow as that of the GR helper in the common IGP GR or MPLS LDP GR.

    • If BGP is configured, the BGP processing flow is the same as that of the GR helper in the common BGP GR except that the BGP processing flow includes additional IGP GR processing and MPLS LDP GR processing, and the P then keeps information about all the public IPv4 routes for a certain period.

  • After detecting the restart of the PE, the RRs reflecting VPNv4 routes and the other PEs (including ASBRs) connected to this PE use the same processing flow as that of the GR helper in the BGP GR. They then keep information about all the public IPv4 routes and VPNv4 routes for a certain period.

Master/Slave Main Control Board Switchover of a P

The processing flow of a P is the same as that of the GR restarter in common IGP GR, MPLS LDP GR, or BGP GR.

After detecting the restart of a P, other Ps and PEs that connect to the P use the same processing flow as that of the GR helper in common IGP GR or BGP GR. That means that they keep information about all the public network IPv4 routes for a certain period.

Master/Slave Main Control Board Switchover of a CE

The processing flow of a CE is the same as that of the GR restarter in common IGP GR or BGP GR.

After detecting the restart of a CE, the PEs that connect to the CE use the same processing flow as that of the GR helper in common IGP GR or BGP GR. That means that they keep information about all the private network IPv4 routes for a certain period.

Download
Updated: 2018-07-04

Document ID: EDOC1100027166

Views: 46322

Downloads: 178

Average rating:
This Document Applies to these Products
Related Documents
Related Version
Share
Previous Next