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

ME60 V800R010C10SPC500 Configuration Guide - IP Routing 01

This is ME60 V800R010C10SPC500 Configuration Guide - IP Routing
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).
Adjusting the BGP Network Convergence Speed

Adjusting the BGP Network Convergence Speed

You can adjust the BGP network convergence speed by adjusting BGP peer connection parameters to adapt to changes on large-scale networks.

Usage Scenario

BGP is used to transmit routing information on large-scale networks. Frequent network changes affect the establishment and maintenance of BGP peer relationships, which in turn affects the BGP network convergence speed.

The route dampening and triggered update functions of BGP suppress frequent route changes to a certain extent, but cannot minimize the impact of network flapping on BGP connections.

You can configure BGP timers and disable rapid EBGP connection reset to suppress BGP network flapping and speed up BGP network convergence.
  • BGP Keepalive and Hold timers

    BGP uses Keepalive messages to maintain BGP peer relationships and monitor connection status. After establishing a BGP connection, two peers send Keepalive messages periodically to each other to detect the BGP connection status based on the Keepalive timer. If a router does not receive any Keepalive message or any other types of packets from its peer within the hold time set by the Hold timer, the router considers the BGP connection interrupted and terminates the BGP connection.

  • BGP MinRouteAdvertisementIntervalTimer

    BGP does not periodically update a routing table. When BGP routes change, BGP updates the changed BGP routes in the BGP routing table by sending Update messages. If a route changes frequently, to prevent the router from sending Update messages upon every change, set the interval at which Update messages are sent.

  • Rapid EBGP peer reset

    Rapid EBGP connection reset is enabled by default so that EBGP can quickly detect the status of interfaces used to establish EBGP connections. If the interface status changes frequently, you can disable rapid EBGP connection reset so that direct EBGP sessions will not be reestablished and deleted as interface alternates between Up and Down, which speeds up network convergence.

  • BGP ConnectRetry Timer

    After BGP initiates a TCP connection, the ConnectRetry timer will be stopped if the TCP connection is established successfully. If the first attempt to establish a TCP connection fails, BGP re-establishes the TCP connection after the ConnectRetry timer expires.

    Setting a short ConnectRetry interval reduces the period BGP waits between attempts to establish a TCP connection, which speeds up the establishment of the TCP connection.

    Setting a long connectRetry interval suppresses routing flapping caused by peer relationship flapping.

After slow peer detection is configured on a device, the device identifies the slow BGP peer (if any) and removes it from the update peer-group to prevent this slow peer from affecting route advertisement to other peers in this update peer-group. Slow peer detection speeds up BGP network convergence.

Pre-configuration Tasks

Before setting parameters for a BGP peer connection, configure basic BGP functions.

Configuration Procedures

Perform one or more of the following configurations as required.

Configuring BGP Keepalive and Hold Timers

The values of BGP Keepalive and Hold timers determine the speed at which BGP detects network faults. You can adjust the values of these timers to improve network performance.

Context

BGP uses Keepalive messages to maintain peer relationships. After establishing a BGP connection, two peers periodically send Keepalive messages to each other to detect BGP peer relationship status. If a device receives no Keepalive message from its peer after the Hold timer expires, the device considers the BGP connection interrupted.

  • If short Keepalive time and holdtime are set, BGP can fast detect link faults. This speeds up BGP network convergence, but increases the number of Keepalive messages on the network and loads of ME devices, and consumes more network bandwidth resources.
  • If long Keepalive time and holdtime are set, the number of Keepalive messages on the network and loads of ME devices are reduced. If the Keepalive time is too long, BGP cannot fast detect link status changes, which slows down BGP network convergence and may cause packet loss.

Changing timer values using the timer command or the peer timer command interrupts BGP peer relationships between ME devices. Therefore, exercise caution before running either of the command.

Keepalive and Hold timers can be configured either for all peers or peer groups, or for a specific peer or peer group. Keepalive and Hold timers configured for a specific peer take precedence over those configured for the peer group of this peer. In addition, Keepalive and Hold timers configured for a specific peer or peer group take precedence over those configured for all peers or peer groups.

Procedure

  • Configure BGP timers for all peers or peer groups.
    1. Run system-view

      The system view is displayed.

    2. Run bgp as-number

      The BGP view is displayed.

    3. Run timer keepalive keepalive-time hold hold-time [ min-holdtime min-hold-value ]

      BGP timers are configured.

      The proper maximum interval at which Keepalive messages are sent is one third the holdtime and is not less than one second. Therefore, set the holdtime to 3s at least.
      NOTE:
      Setting the Keepalive time to 20s is recommended. If the Keepalive time is smaller than 20s, sessions between peers may be closed.

      When setting values of keepalive-time and hold-time, note the following points:

      • The keepalive-time and hold-time values cannot be both set to 0. Otherwise, the BGP timers become invalid, and BGP will not send Keepalive messages to detect connection status.

      • The hold-time value cannot be much greater than the keepalive-time value. For example, keepalive-time cannot be set to 1 while hold-time is set to 65535. If the hold-time value is too large, BGP cannot detect connection status in time.

      After a connection is established between peers, the peers negotiate the keepalive-time and hold-time values. The smaller one of the hold-time values carried by Open messages of both peers is used as the hold-time value. The smaller value of one third of the negotiated hold-time value and the locally configured keepalive-time value is used as the keepalive-time value.

    4. Run commit

      The configuration is committed.

  • Configure timers for a specific peer or peer group.
    1. Run system-view

      The system view is displayed.

    2. Run bgp as-number

      The BGP view is displayed.

    3. Run peer { ipv4-address | group-name } timer keepalive keepalive-time hold hold-time [ min-holdtime min-hold-value ]

      The Keepalive and Hold timer values are set for a specific peer or peer group.

      For details about the relationship between the keepalive-time and hold-time values, see Configure BGP timers for all peers or peer groups.
      NOTE:
      Setting the Keepalive time to 20s at least is recommended. If the Keepalive time is less than 20s, sessions between peers may be interrupted.

    4. Run commit

      The configuration is committed.

Configuring a MinRouteAdvertisementIntervalTimer

A proper MinRouteAdvertisementIntervalTimer can be configured to suppress frequent route changes, improving BGP network stability.

Context

BGP peers use update messages to exchange routing information. Update messages can be used to advertise reachable routes with the same attributes or delete unreachable routes.

BGP does not periodically update a routing table. When BGP routes change, BGP updates the changed BGP routes in the BGP routing table by sending Update messages. If a route changes frequently, to prevent the ME device from sending Update messages upon every change, configure a MinRouteAdvertisementIntervalTimer at which Update messages are sent.

Procedure

  1. Run system-view

    The system view is displayed.

  2. Run bgp as-number

    The BGP view is displayed.

  3. Run peer { ipv4-address | group-name } route-update-interval interval

    A MinRouteAdvertisementIntervalTimer is configured.

    ipv4-address specifies the address of a specific peer. while group-name specifies the name of a peer group. The MinRouteAdvertisementIntervalTimer configured for a peer takes precedence over that configured for a peer group.

  4. Run commit

    The configuration is committed.

Disabling Fast Reset of EBGP Connections

Disabling rapid EBGP connection reset can prevent frequent reestablishment and deletion of EBGP sessions if route flapping occurs, which speeds up BGP network convergence.

Context

With rapid EBGP connection reset, BGP can immediately respond to a fault on an interface and delete the direct EBGP sessions on the interface without waiting for the hold timer to expire, which speeds up BGP network convergence. Rapid EBGP connection reset is enabled by default.

NOTE:

Rapid EBGP connection reset enables BGP to quickly respond to interface faults, but not interface recovery. After the interface recovers, BGP uses its state machine to restore relevant sessions.

If the status of an interface used to establish an EBGP connection changes frequently, the EBGP session will be deleted and reestablished repeatedly, causing network flapping. To address this issue, disable rapid EBGP connection reset so that BGP will not delete direct EBGP sessions on the interface until the hold timer expires. Therefore, disabling rapid EBGP connection reset suppresses BGP network flapping, speed up BGP network convergence, and reduce network bandwidth consumption.

Procedure

  1. Run system-view

    The system view is displayed.

  2. Run bgp as-number

    The BGP view is displayed.

  3. Run undo ebgp-interface-sensitive

    Rapid EBGP connection reset is disabled.

    NOTE:

    Disable rapid EBGP connection reset only when the status of an interface used to establish an EBGP connection changes frequently. If the status of the interface becomes stable, run the ebgp-interface-sensitive command to enable rapid EBGP connection reset.

  4. Run commit

    The configuration is committed.

Configuring a BGP ConnectRetry Timer

You can control the speed at which BGP peer relationships are established by changing the BGP ConnectRetry timer value.

Context

After BGP initiates a TCP connection, the ConnectRetry timer will be stopped if the TCP connection is established successfully. If the first attempt to establish a TCP connection fails, BGP re-establishes the TCP connection after the ConnectRetry timer expires.
  • Setting a short ConnectRetry interval reduces the period BGP waits between attempts to establish a TCP connection, which speeds up the establishment of the TCP connection.
  • Setting a long connectRetry interval suppresses routing flapping caused by peer relationship flapping.

A ConnectRetry timer can be configured either for all peers or peer groups, or for a specific peer or peer group. A ConnectRetry timer configured for a specific peer takes precedence over that configured for the peer group of this peer. In addition, a ConnectRetry timer configured for a specific peer or peer group takes precedence over that configured for all peers or peer groups.

Procedure

  • Configure a BGP ConnectRetry timer for all peers or peer groups.

    Perform the following steps on a BGP ME device:

    1. Run system-view

      The system view is displayed.

    2. Run bgp as-number

      The BGP view is displayed.

    3. Run timer connect-retry connect-retry-time

      A BGP ConnectRetry timer is configured for all peers or peer groups.

    4. Run commit

      The configuration is committed.

  • Configure a BGP ConnectRetry timer for a peer or peer group.

    Perform the following steps on a BGP ME device:

    1. Run system-view

      The system view is displayed.

    2. Run bgp as-number

      The BGP view is displayed.

    3. Run peer { group-name | ipv4-address } timer connect-retry connect-retry-time

      A BGP ConnectRetry timer is configured for a peer or peer group.

    4. Run commit

      The configuration is committed.

Enabling Slow Peer Detection

After slow peer detection is configured on a device, the device identifies the slow BGP peer (if any) and removes it from the update peer-group to prevent this slow peer from affecting route advertisement to other peers in this update peer-group. Slow peer detection speeds up BGP network convergence.

Context

An update peer-group may consist of multiple BGP peers. If a network problem (congestion for example) occurs and slows down the speed at which the local device advertises routes to a BGP peer in the update peer-group, the speed at which the local device advertises routes to other BGP peers in the update peer-group is affected. To address this problem, enable slow peer detection.

After slow peer detection is enabled, the local device calculates the difference between the time taken to send packets to each BGP peer and the shortest time taken to send packets to a BGP peer in the group. If the difference between the time taken to send packets to BGP peer 1 and the shortest time is greater than the threshold, the local device considers BGP peer 1 as a slow peer and removes it from the update peer-group, which prevents this slow peer from affecting route advertisement to other peers in the group.

Procedure

  1. Run system-view

    The system view is displayed.

  2. Run bgp { as-number-plain | as-number-dot }

    The BGP view is displayed.

  3. (Optional) Enter a desired address family view using one of the following commands.

    • ipv4-family { unicast | vpnv4-multicast | multicast | vpnv4 [ unicast ] | vpn-instance vpn-instance-name | mvpn | mdt }
    • ipv6-family [ [ vpnv6 ] [ unicast ] | vpn-instance vpn-instance-name ]
    • l2vpn-ad-family
    • l2vpn-family evpn
    • link-state-family unicast

  4. Run slow-peer detection threshold threshold-value

    Slow peer detection is enabled.

    threshold threshold-value specifies a slow peer detection threshold. If the difference between the time taken to send packets to BGP peer 1 and the shortest time taken to send packets to BGP peer 2 is greater than the threshold-value, the local device considers BGP peer 1 as a slow peer and removes it from the update peer-group.

  5. Run commit

    The configuration is committed.

Verifying the Configuration of the BGP Network Convergence Speed

After the BGP network convergence speed is adjusted, you can view information about BGP peers and peer groups.

Prerequisites

Adjusting the BGP network convergence speed has been configured.

Procedure

  • Run the display bgp peer [ verbose ] command to check information about BGP peers.
  • Run the display bgp group [ group-name ] command to check information about BGP peer groups.
  • Run the display bgp slow-peer command to check information about slow BGP peers.
Translation
Download
Updated: 2019-01-04

Document ID: EDOC1100059437

Views: 20897

Downloads: 15

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