[Top] [Table Of Contents] [Prev] [Next] [Index]
12 out of 25 total pages
 

BGP Routing   9

  This chapter describes the commands you use to configure PortMaster 4, when you are using the Border Gateway Protocol (BGP) as a routing protocol. Lucent implements version 4 of BGP, as defined in RFC 1771, with updates from the draft standard number 5 of January 1997. Also supported are the BGP communities attribute, defined in RFC 1997, BGP autonomous system confederations, defined in RFC 1965, and BGP route reflection, defined in RFC 1966.
  See the PortMaster Routing Guide for BGP configuration instructions and examples before attempting to configure BGP.

  Note ¯ After making any changes to the BGP configuration, you must use the save all  and reset bgp  commands to ensure the changes take effect, and are retained after PortMaster reboots. If you are changing only peer-specific policy information, however, you need only reset the affected individual peers with the reset bgp peer  Ipaddress command.

 

       Displaying BGP Information

  To display BGP information on the console, use the following commands:
 

       Summary of BGP Commands

  BGP commands, shown in Table 9-1, allow you to configure the PortMaster for BGP routing.

  Table 9-1 BGP Commands 

 
  Command Syntax     
 add bgp peer Ipaddress(sr) Ipaddress(dest) ASN  - see page 9-4
 add bgp policy Policyname  - see page 9-5
 add bgp summarization Prefix/NM  - see page 9-5
 add propagation Protocol(src) Protocol(dest) Metric Filtername  - see page 7-3
 delete bgp peer Ipaddress(dest)  - see page 9-6
 delete bgp policy Policyname|all  - see page 9-6
 delete bgp summarization Prefix/NM  - see page 9-7
 delete propagation Protocol(src) Protocol(dest)  - see page 7-3
 reset bgp [peer Ipaddress]  - see page 9-8
 reset propagation  - see page 7-5
 save bgp  - see page 9-8
 set bgp as ASN  - see page 9-9
 set bgp cluster-id Ipaddress  - see page 9-9
 set bgp cma ASN  - see page 9-10
 set bgp connect-retry-interval Seconds  - see page 9-10
 set bgp enable|disable  - see page 9-11
 set bgp hold-time Seconds  - see page 9-11
 set bgp id Ipaddress  - see page 9-12
 set bgp igp-lockstep on|off  - see page 9-12
 set bgp keepalive-timer Seconds  - see page 9-13
 set bgp peer  Ipaddress(src) Ipaddress(dest) ASN
[assume-default [Number]] [confederation-member]
[route-reflector-client] [normal] [always-next-hop]
{easy-multihome|[accept-policy Policyname|all]
[inject-policy Policyname|all]
[advertise-policy Policyname|all]} 		 - see page 9-13
 set bgp policy Policyname [before] RuleNumber
permit|deny|include Policyname
[if
[prefix [exactly] Prefix/NM]
[prefix-longer-than NM]
[as-path String|empty] [community Tag]]
[then
[input-multi-exit-disc Number|strip]
[ degree-of-preference Number]
[local-pref Number]
[output-multi-exit-disc Number|strip]
[next-hop Ipaddress]
[community add|replace|strip Tag]
[ignore-community-restrictions]] 		 - see page 9-17,
page 9-21,
page 9-24
 set bgp policy Policyname blank  - see page 9-28
 set bgp summarization Prefix/NM [as ASN]
[cms ASN] [multi-exit-disc Number]
[local-pref Number] [community Tag] [all] 		 - see page 9-29
 set debug bgp on|off  - see page 14-2
 show bgp memory  - see page 9-31
 show bgp next-hop  - see page 9-32
 show bgp paths [Prefix/NM [verbose]]  - see page 9-33
 show bgp peers [verbose|packets]  - see page 9-36
 show bgp policy [Policyname]  - see page 9-40
 show bgp summarization  - see page 9-41
 show routes [String|Prefix/NM]  - see page 9-42
 

       BGP Commands

  These commands are used for configuring the BGP routing protocol on the PortMaster 4.

  Note ¯ BGP is a complex protocol to configure. Consult the instructions and examples in the PortMaster Routing Guide  before configuring BGP on a PortMaster 4.

 
 

       add bgp peer

  This command creates entries on the PortMaster for BGP peers.

  add bgp peer Ipaddress(src) Ipaddress(dest) ASN

 
 Ipaddress(src)  Local address of the PortMaster put in outgoing packets, specified in dotted decimal notation.
 Ipaddress(dest)  Destination address of the peer, specified in dotted decimal notation.

  Adding or Changing Peer Parameters.  The set bgp peer  command permits you to specify the parameters for an existing BGP peer without deleting that peer. However, the command assumes a "clean slate" for all parameters, and requires that you reenter them completely. For example, supposing you want to change your configuration of a peer 192.168.1.5 configured with the following command:
  add bgp peer 192.168.1.1 192.168.1.5 105 route-reflector-client
always-next-hop accept all inject all
  If you now want to add advertise all  as a policy statement to the command, you must specify all the original parameters together with the new parameter in the set bgp peer  command, as follows:
  set bgp peer 192.168.1.1 192.168.1.5 105 route-reflector-client
always-next-hop accept all inject all advertise all
  set bgp peer - page 9-13
  set bgp policy (acceptance) - page 9-17
set bgp policy (injection) - page 9-21
set bgp policy (advertisement) - page 9-24
 

       add bgp policy

  This command creates a BGP policy for route acceptance, injection, or advertisement.

  add bgp policy Policyname

 
 Policyname  Name of the policy to be created or deleted--a string of up to 16 nonspace characters.

  Use the delete bgp policy  command to delete a BGP policy. Define BGP policies with the set bgp policy  commands.
  delete bgp policy - page 9-6
  set bgp policy (acceptance) - page 9-17
set bgp policy (injection) - page 9-21
set bgp policy (advertisement) - page 9-24
 

       add bgp summarization

  This command creates a BGP summarization entry.

  add bgp summarization Prefix/NM

 
 Prefix  Address prefix that you want to advertise to the BGP peers in dotted decimal notation.
 /NM  Netmask that indicates the number of high-order bits in the address prefix. This is a number from 1 to 32, preceded by a slash (/)--for example, /24.

  set bgp policy - page 9-17
 

       delete bgp peer

  This command deletes existing BGP peer entries on the PortMaster.

  delete bgp peer Ipaddress(dest)

 
 Ipaddress(src)  Local address of the PortMaster put in outgoing packets, specified in dotted decimal notation.
 Ipaddress(dest)  Destination address of the peer, specified in dotted decimal notation.

  When a peer deletion is in process, the message and countdown timer "Deletion in Progress. Countdown 216" are displayed in the Accept, Inject, and Advertise columns of the show bgp peers  command. Deletion is complete when the countdown drops to zero.
  add bgp peer - page 9-4

  set bgp peer - page 9-13

 

       delete bgp policy

  This command deletes a BGP policy.

  Caution ¯ Be careful when deleting BGP policy statements. Make sure that they are no longer needed for BGP route selection.

  delete bgp policy Policyname|all

 
 Policyname  Name of the policy to be created or deleted--a string of up to 16 nonspace characters.
 all  Predefined policy that you can use to permit all routes to be accepted, injected, or advertised.

  Use the add bgp policy  command to create a BGP policy. Define BGP policies with the set bgp policy  commands.
  add bgp policy - page 9-5
  set bgp policy (acceptance) - page 9-21
set bgp policy (injection) - page 9-21
set bgp policy (advertisement) - page 9-24
 

       delete bgp summarization

  This command deletes a BGP summarization entry.

  delete bgp summarization Prefix/NM

 
 delete  Deletes an existing BGP summarization entry.
 Prefix  Address prefix that you want to advertise to the BGP peers. Specified in dotted decimal notation.
 /NM  Netmask that indicates the number of high-order bits in the address prefix. This is a number from 1 to 32, preceded by a slash (/)--for example, /24.

 

       reset bgp

  This command recreates startup conditions for BGP.

  reset bgp [peer Ipaddress]

 
 peer  Resets only the session with the specified peer.
 Ipaddress  IP address of the peer to be reset, specified in dotted decimal notation.

  When used with no parameters, this command causes the PortMaster to lose all currently known BGP information except for configuration information. The PortMaster then rereads configuration information for BGP and re-establishes sessions with peers. This process is not instantaneous, but takes some time to finish.
  After you use this command, BGP is in a transient state, during which the show  commands are inoperative.
  Using the command set console  before entering this command allows you to see the message "BGP Reset Complete" on the console when the reset process is complete. Otherwise, the command provides no response.
  When you use the command with the optional peer  Ipaddress, only the configuration session with the specified peer is reset.
 

       save bgp

  This command writes any changes in the BGP tables to the nonvolatile memory of the PortMaster.

  save bgp

  Note ¯ To save all configuration information, including BGP and global parameters such as the local system and local BGP router ID, use the save all  command instead.

 

       set bgp as

  This command sets the number of the autonomous system that the PortMaster is a member of.

  set bgp as ASN

 
 ASN  Unique number that identifies the autonomous system--a 16-bit number ranging from 1 to 65535.

  Autonomous system identifiers are supplied by the Internet Network Information Center (InterNIC). If autonomous system confederations are in use, this number identifies your BGP confederation's autonomous system to BGP peers outside the confederation.
 

       set bgp cluster-id

  This command identifies the PortMaster as a BGP route reflector in a cluster.

  set bgp cluster-id Ipaddress

 
 Ipaddress  IP address in dotted decimal notation. It can be any IP address, but is typically the BGP ID of one of the route reflectors. Setting the cluster ID to 0.0.0.0  removes it, and disables the ability of this PortMaster to be a route reflector.  Route reflection is disabled by default.

  An autonomous system can be divided into many clusters. Each cluster contains one or more internal peers configured as route reflectors, with the remaining peers in the cluster called route reflector clients. Peers configured as route reflectors in an autonomous system are fully meshed with each other, but the clients are configured as peers only with route reflectors in their cluster.
  The same cluster ID must be set on each route reflector in a cluster, but cluster IDs are not set on the reflector clients.
  Advantages of Clustering.  The use of clusters reduces the traffic and CPU overhead compared with a fully meshed system. When compared to confederations, route reflector clusters are simpler to configure, but do not allow the degree of policy control that is possible across confederation boundaries. The primary advantage of route reflector clusters is that they allow the PortMaster to interoperate with BGP peers that are third-party routers without the ability to be configured into confederations.
  For information about the effects of route reflection on BGP policies, see page 9-16.
 

       set bgp cma

  This command sets the number of the BGP confederation member autonomous system (CMAS) that the PortMaster is in.

  set bgp cma ASN

 
 ASN  The CMAS identifier--a 16-bit number ranging from 0 to 65535. A value of 0 disables the CMAS configuration. Confederations are disabled by default.

  You can divide an autonomous system into multiple autonomous systems and group them into a single confederation. To external autonomous systems, the confederation appears as a single autonomous system. When confederations are in use, the PortMaster advertises this autonomous system identifier to BGP peers that are marked as confederation members in its configuration.
  Choosing a value of zero disables use of confederations on this PortMaster. Confederations are disabled by default.
 

       set bgp connect-retry-interval

  This command sets the BGP connection retry interval for the PortMaster.

  set bgp connect-retry-interval Seconds

 
 Seconds  Connection retry interval in seconds. The valid range is from 30 to 1000 seconds. The default is 120 seconds.

  This command sets the interval at which the PortMaster attempts to open sessions to peers that are not fully established.
 

       set bgp enable|disable

  This command enables or disables the use of BGP on the PortMaster.

  Note ¯ You must issue the save all  and reboot  commands immediately after issuing the set bgp enable  command, before you can continue with any other BGP configuration.

  set bgp enable|disable

 
 enable  Loads the BGP software upon the next PortMaster reboot.
 disable  Disables the use of BGP upon the next reboot of the PortMaster, and frees the system memory used by BGP. This is the default.

  You must enable BGP and reboot the PortMaster before configuring or using BGP.
The save all  and reboot  commands must be issued after you use this command with either the enable  or disable  options.
 

       set bgp hold-time

  This command sets the BGP hold time interval for the PortMaster.

  set bgp hold-time Seconds

 
 Seconds  Hold time interval in seconds. The valid range is from 30 to 1000 seconds. The default is 90 seconds.

  This command sets the interval that the PortMaster waits between keepalive, update, or notification messages from a peer, before identifying the peer as no longer operational and dropping all information learned from that peer.
 

       set bgp id

  This command identifies the PortMaster as a BGP router.

  set bgp id Ipaddress

 
 Ipaddress  PortMaster IP address, specified in dotted decimal notation.

  The BGP identifier must be an IP address on the PortMaster. A setting of 0.0.0.0 removes the BGP ID.
 

       set bgp igp-lockstep

  This command enables or disables a feature that forces the PortMaster to match a route learned from internal BGP peers with a route learned from OSPF, RIP, static routing, or RADIUS before advertising the route to external peers.

  set bgp igp-lockstep on|off

 
 on  Enables the matching feature.
 off  Disables the matching feature.

  Normally, when the PortMaster learns a route from internal peers, it forwards the information to any external peers as soon as possible. Enabling the lockstep feature forces the PortMaster to wait until it finds a suitable Interior Gateway Protocol (IGP) route--an OSPF, RIP, or static route, or a static route via RADIUS--that supports the route before advertising it. An IGP route supports a BGP route if it has the same IP address and prefix as the BGP route.

  Note ¯ Exact matches only are allowed because simple default routes to support BGP routes can lead to network instability or lost packets.

 

       set bgp keepalive-timer

  This command sets the BGP keepalive timer interval.

  set bgp keepalive-timer Seconds

 
 Seconds  Keepalive timer interval in seconds. The valid range is from 30 to 1000 seconds. The default is 30 seconds.

  This command sets the interval at which the PortMaster sends keepalive messages to its peers, to let them know it is still reachable.
 

       set bgp peer

  This command modifies entries on the PortMaster for BGP peers, and provide options that control how policies are implemented for route selection.

  set bgp peer Ipaddress(src) Ipaddress(dest) ASN
[assume-default [Number] ] [confederation-member]
[route-reflector-client] [normal] [always-next-hop]
{easy-multihome|[accept-policy Policyname|all]
[inject-policy Policyname|all] [advertise-policy Policyname|all]}

 
 Ipaddress(src)  Local address of the PortMaster put in outgoing packets, specified in dotted decimal notation.
 Ipaddress(dest)  Destination address of the peer, specified in dotted decimal notation.
 ASN  Autonomous system number of the peer. If this autonomous system is the same as that of the PortMaster, the peer is an internal peer; if it is different, the peer is an external peer. The autonomous system number is a 16-bit number ranging from 1 to 65535.
 assume-default  Indicates that a default route to this external peer is created if the peer is up. You must assign a hop-count value to the default routes of different peers to specify a preferred peer.
 Number  Hop count to advertise this default route. When multiple peers are configured with assume-default , the one with the lowest hop count is the preferred router for default-route forwarding. Number is a value from 1 to 15.
 confederation-member  When specified, identifies a peer that is a member of the same confederation as the PortMaster. By default this keyword is not specified.
 route-reflector-client  When specified, identifies a peer as a route reflector client  that the PortMaster forwards internal routes to. For the peer to be enabled as a route-reflector client, you must have configured the PortMaster with a cluster ID using the set bgp cluster-id  command.
 normal  When specified, identifies a peer that is neither a confederation member nor a route-reflector client. By default normal  is specified.
 always-next-hop  When specified, identifies the PortMaster as the next hop  in any update packet sent to it from the peer, even if the PortMaster determines that it is not always the best next hop choice for this peer.  This option is useful when you know that this peer has connectivity to the PortMaster, but possibly not to the same devices that you would choose as a next hop--for example, in a partially meshed Frame Relay network.  By default always-next-hop  is disabled.

  Note ¯ Standard BGP speaker behavior is to forward next hop  information to internal peers without modification. The always-next-hop  parameter enables this behavior to be changed. Therefore, when using the always-next-hop  parameter, you must take care to ensure that inconsistent routing information is not propagated from multiple external peers to the autonomous system.

 easy-multihome  Enables an alternative method to policies for handling multihome paths from the PortMaster. The easy-multihome  keyword restricts the BGP routing table to accept only paths through the remote autonomous system, and optionally through one additional autonomous system. Otherwise, the PortMaster uses the assume-default  keyword to determine how to route packets.
 accept-policy  Enables a BGP policy Policyname whose criteria must be met for the PortMaster to accept any IP prefix from this peer as a viable BGP route. If a then degree-of-preference  parameter is specified in the policy (see set bgp policy (acceptance)  on page 9-17), it is used in place of any information learned from the path for path preference calculation purposes only. Advertisement filters indicate what the other peers are told.  If not specified, and easy-multihome  is not enabled for this peer, then nothing is accepted from this peer.
 all  Predefined policy that you can use to permit all routes to be accepted, injected, or advertised.
 Policyname  Name of a BGP policy statement defined by the set bgp policy  command.
 inject-policy  Enables a BGP policy Policyname whose criteria must be met for the PortMaster to place any IP address prefix received from this peer in the routing table. No then  parameters are used in this policy.  If not specified, and easy-multihome  is not enabled for this peer, then nothing is injected from this peer into the routing table.
 advertise-policy  Enables a BGP policy Policyname whose criteria must be met for the PortMaster to advertise any IP address prefix to this peer. The advertisement you set with the set bgp policy  command indicates the metrics and any community information to advertise with the prefix.  If not specified, and easy-multihome  is not enabled for this peer, then nothing is advertised to this peer.

  If no policy is defined, then the default behavior is not  to accept, advertise, or inject any BGP routes. Therefore, when you define a peer you must do one of the following:
  Adding or Changing Peer Parameters.  The set bgp peer  command permits you to specify the parameters for an existing BGP peer without deleting that peer. However, the command assumes a "clean slate" for all parameters, and requires that you reenter them completely. For example, supposing you want to change your configuration of a peer 192.168.1.5 configured with the following command:
  add bgp peer 192.168.1.1 192.168.1.5 105 route-reflector-client
always-next-hop accept all inject all
  If you now want to add advertise all  as a policy statement to the command, you must specify all the original parameters together with the new parameter in the set bgp peer  command, as follows:
  set bgp peer 192.168.1.1 192.168.1.5 105 route-reflector-client
always-next-hop accept all inject all advertise all
  Requirement for Internal Peers to Be Fully Meshed.  Unless route reflection is used, BGP requires that all BGP peers within an autonomous system or within a confederation member autonomous system (CMAS) be linked to each other. In this way, when one BGP peer learns an external route--path attributes and destination--it forwards this information to all its internal peers. Because they are fully meshed, each peer has the same information as its internal peers in the autonomous system and does not need to forward it again to them. If route reflector clusters are used, only the route reflectors--but not the route reflection clients--need to be fully meshed.
  Length of Time Information Is Held Before Forwarding.  When information is first learned from a peer, that information is held for at least 30 seconds before being forwarded to other peers as trustworthy and stable.
  Peer Deletion.  When a peer deletion is in process, the message and countdown timer "Deletion in Progress. Countdown 216" are displayed in the Accept, Inject, and Advertise columns of the show bgp peers  command. Deletion is complete when the countdown drops to zero.
  Effects of Route Reflection on BGP Policies.  When a route reflector reflects an internal route  that it learned from other internal peers either from or to a reflector client, the BGP policies for the cluster changes as follows:
  This modified behavior applies only  to reflected internal routes learned from other internal peers, and not  to routes originating from the route reflector itself. The route reflector can generate routes from locally configured summarizations, or from routing information learned via external peers attached to the route reflector.
  You can use policy statements to permit or deny certain routes from being reflected.
  set bgp policy (acceptance) - page 9-17
set bgp policy (injection) - page 9-21
set bgp policy (advertisement) - page 9-24
 
 

       set bgp policy (acceptance)

  This command creates a policy rule for admitting an IP prefix learned from a peer into a BGP database on the PortMaster for further consideration as a route.

  Caution ¯ The creation of long, complex lists of policy rules can adversely affect PortMaster CPU performance.

  set bgp policy Policyname [before] RuleNumber
permit|deny|include Policyname
[if
[prefix [exactly] Prefix/NM]
[prefix-longer-than NM]
[as-path String|empty]
[community Tag]]
[then
[input-multi-exit-disc Number|strip]
[degree-of-preference Number]]

 
 Policyname  Name of an acceptance policy already created.
 before  Optionally inserts this BGP rule before an existing rule in the policy.
 RuleNumber  Number of a rule in the policy.

  · Use the RuleNumber of an existing rule to replace that rule.

  · Add this rule to the end of the list of rules by using a RuleNumber value that is 1 greater than the current largest rule number.

  · A maximum of 160 rules is permitted in a policy. If more rules are needed, they can be added with the include  Policyname option.

 permit  Allows the IP prefix into the BGP database if the criteria in the rule are met.
 deny  Prohibits the IP prefix from the BGP database if the criteria in the rule are met.
 include Policyname  Inserts an existing policy Policyname into the current policy. Included policies can themselves include other policies, up to a maximum level of 10 nested included policies.
 if  Compares the prospective IP prefix against corresponding elements specified after if  in this rule. Specifying no if  elements causes all prefixes to match the current rule.

  · If all elements of the IP prefix match these if  criteria, this rule is applied to the prefix and the prefix is either permitted or denied.

  · If the elements do not match, the list of policy rules is further scanned for a matching rule.

  · If no matches are found, the IP prefix is denied from the BGP database.

 prefix Prefix/NM  IP prefix Prefix and netmask NM to compare the prospective IP prefix against. The netmask indicates the number of high-order bits in the IP prefix.

  · Specify Prefix in dotted decimal notation.

  · Specify NM as number from 1 to 32, preceded by a slash (/)--for example, /24.

 By default, any prefix that matches the netmask in the rule prefix in the leftmost--most significant--bits, matches the rule prefix.
 exactly  Requires the entire prospective IP prefix and netmask to exactly match the IP prefix and netmask specified in the rule.
 prefix-longer-than NM  When used with the deny  keyword, prohibits from the BGP database any prospective IP address with a prefix containing more high-order bits than are specified by the netmask NM.
 as-path String  Autonomous system path String to compare the prospective IP prefix against.  String is a list of autonomous system numbers, separated by periods (.)--for example, AS1.AS2.AS3. or AS2.AS1.
   When String is compared to an autonomous system path sequence , the order of the sequence must match the order of String. When String is compared to an autonomous system path set , the set  is put in ascending numerical order, and then matched against String. Multiple sequences or sets in a single autonomous system path are concatenated before being compared to String.  The following special characters have the following meaning in the expression:

  · An asterisk (*) matches one or more entries in the autonomous system sequence.

  · A question mark (?) matches any single item in the autonomous system sequence.

 empty  Value for String that matches only paths containing no autonomous system path information.  Use as-path empty  only to permit or deny routes originating from an internal or confederation member peer within the autonomous system of the PortMaster.
 community  Identifier Tag that categorizes a group of destinations to compare the prospective IP prefix against.  See RFC 1997 for more information on a BGP community.
 Tag  32-bit number that indicates a destination category in one of the following forms:

  · One 32-bit value identifying the autonomous system of the destination

  · Two 16-bit values: one containing the autonomous system number of the destination, and the other containing additional information about the autonomous system. If only the first 16-bit word is considered significant in matching the community Tag, replace the second 16-bit value with the keyword any .

 

  · One of the following reserved community keywords that restrict route advertisement for peers receiving the route information:

   no-export  Destinations only within a confederation.  Advertise the route only to BGP peers within your confederation or autonomous system.
   no-advertise  No destinations.  Do not advertise this route.
   no-export-subconfed  Internal destinations only.  Advertise this route only to internal BGP peers.
   The restrictions imposed by these reserved community keywords do not apply to the PortMaster originating this information.
 then  Assigns the following metric or metrics to any IP prefix selected for acceptance by the rule.
 input-multi-exit-disc Number|strip  Assigns an arbitrary Number for the learned multiexit discriminator, overriding any that is learned from the peer. Number is a 32-bit integer. The strip  keyword causes any multiexit discriminator information learned from a peer to be ignored.  input-multi-exit-disc  can be abbreviated as imed  in this command.  Lower  numbers indicate an increased preference for a specific route. Use this metric to discriminate among multiple exit or entry points between the same pair of neighboring autonomous systems.
 degree-of-preference Number  Assigns a degree-of-preference Number to a route. Number is a 32-bit integer.  degree-of-preference  can be abbreviated as dop  in this command  Higher  numbers indicate an increased preference for a specific route when more than one route exists. Use this metric to screen a particular autonomous system from your map of routes, for example.  If you do not assign a degree of preference to the IP prefix, one of the following values is assigned by default:

  · If the route comes from an internal peer, the learned local preference number is assigned.

  · If the route comes from an external peer, Number is based on the autonomous system path length, with a shorter path being preferred.

  A BGP policy  is a list of rules that restrict the BGP routes your PortMaster accepts from its peers, uses, and advertises to its peers. You can use the easy-multihome  alternative to policies--or accept-policy all  to accept all routes--when you add each BGP peer to your peer group, or you can define your own policies.
  A PortMaster uses an acceptance policy  to determine whether to admit an IP prefix received in a update from a BGP peer into its BGP database for further consideration as a route. If the PortMaster accepts the IP prefix, it uses an injection policy  to determine whether to use the route to forward packets, and an advertisement policy  to determine whether to advertise the route to its BGP peers.
  You can create any number of acceptance, injection and advertisement policies.
  Performing Three Functions in One Policy.  You can create separate policies for each function, or create one policy to perform all three functions.
  Permitting or Denying All Prefixes.  If you define a rule that contains no if  or then  clauses, the rule universally permits or denies all prefixes, with no modification.
  Applying and Saving a Rule.  After adding or changing a rule in a BGP policy, use one of the following commands to apply and save the modified policy:
  Removing a Rule.  Specifying only the rule number RuleNumber in the command, as in set bgp policy policyname 1 , removes that rule from the BGP policy.
  Creating a Common Policy.  You can create a common BGP policy for inclusion in other BGP policies. For example:

  1. Create and define a common BGP policy as follows:

  add bgp policy permit1011

  set bgp policy permit1011 1 permit if prefix 10.0.0.0/8

  set bgp policy permit1011 2 permit if prefix 11.0.0.0/8

  2. Include this policy by reference in another policy as follows:

  set bgp policy otherone 5 include permit1011

  This command inserts the statements of the permit1011  policy at line 5 of the otherone  policy.

  Policy inclusions can be nested to a maximum depth of 10 levels. Any inclusions beyond the 10th level are ignored.
  Reducing the Number of Advertised Routes. Some BGP routes  received by your PortMaster might not be summarized. Unsummarized routes can include IP prefixes containing as many as 32 high-order bits--many specific addresses rather than fewer route summaries. If your BGP policy rules accept such routes into your BGP database, you can propagate extremely large numbers of routes to your BGP peers and possibly overwhelm them. To avoid this problem, use the prefix-longer-than  keyword in a BGP acceptance policy to deny IP prefixes with a netmask longer than a particular NM value. Specifying prefix-longer-than  16, for example, would be highly effective for this purpose.
  For more information about the effects of route reflection on BGP policies, see  page 9-16.
 

       set bgp policy (injection)

  This command creates a policy rule for injecting IP prefixes into the routing table--displayed by the show route  command--that the PortMaster uses to forward packets it receives to their ultimate destination.

  Caution ¯ The creation of long, complex lists of policy rules can adversely affect PortMaster CPU performance.

  set bgp policy Policyname [before] RuleNumber
permit|deny|include Policyname
[if
[prefix [exactly] Prefix/NM]
[as-path String|empty]
[community Tag]]

 
 Policyname  Name of an injection policy already created.
 before  Optionally inserts this BGP rule before an existing rule in the policy.
 RuleNumber  Number of a rule in the policy.  Use the RuleNumber of an existing rule to replace that rule.  Add this rule to the end of the list of rules by using a RuleNumber value that is 1 greater than the current largest rule number.
 permit  Allows the IP prefix into the PortMaster routing table if the criteria in the rule are met.
 deny  Prohibits the IP prefix from the PortMaster routing table if the criteria in the rule are met.
 include Policyname  Inserts an existing policy Policyname into the current policy. Included policies can themselves include other policies, up to a maximum level of 10 nested included policies.
 if  Compares the prospective IP prefix against corresponding elements specified after if  in this rule. Specifying no if  elements causes all prefixes to match the current rule.

  · If all elements of the IP prefix match these if  criteria, this rule is applied to the prefix and the prefix is either added or not added to the PortMaster routing table.

  · If the elements do not match, the list of policy rules is further scanned for a matching rule.

  · If no matches are found, the IP prefix is prohibited from the routing table.

 prefix Prefix/NM  IP prefix Prefix and netmask NM to compare the prospective IP prefix against. The netmask indicates the number of high-order bits in the IP prefix.

  · Specify Prefix in dotted decimal notation.

  · Specify NM as number from 1 to 32, preceded by a slash (/)--for example, /24.

 By default, any prefix that matches the netmask in the rule prefix in the leftmost--most significant--bits, matches the rule prefix.
 exactly  Requires the entire prospective IP prefix and netmask to exactly match the IP prefix and netmask specified in the rule.
 as-path String  Autonomous system path String to compare the prospective IP prefix against.  String is a list of autonomous system numbers, separated by periods (.)--for example, AS1.AS2.AS3. or AS2.AS1.  When String is compared to an autonomous system path sequence , the order of the sequence must match the order of String.
   When String is compared to an autonomous system path set , the set  is put in ascending numerical order, and then matched against String. Multiple sequences or sets in a single autonomous system path are concatenated before being compared to String.  The following special characters have the following meaning in the expression:

  · An asterisk (*) matches one or more entries in the autonomous system sequence.

  · A question mark (?) matches any single item in the autonomous system sequence.

 empty  Value for String that matches only paths containing no autonomous system path information.  Use as-path empty  only to permit or deny routes originating from an internal or confederation member peer within the autonomous system of the PortMaster.
 community  Identifier Tag that categorizes a group of destinations to compare the prospective IP prefix against.  See RFC 1997 for more information on a BGP community.
 Tag  32-bit number that indicates a destination category in one of the following forms:

  · One 32-bit value identifying the autonomous system of the destination

  · Two 16-bit values: one containing the autonomous system number of the destination, and the other containing additional information about the autonomous system. If only the first 16-bit word is considered significant in matching the community Tag, replace the second 16-bit value with the keyword any .

  · One of the following reserved community keywords that restrict route advertisement for peers receiving the route information:

   no-export  Destinations only within a confederation.  Advertise the route only to BGP peers within your confederation or autonomous system.
   no-advertise  No destinations.  Do not advertise this route.
   no-export-subconfed  Internal destinations only.  Advertise this route only to internal BGP peers.
   The restrictions imposed by these reserved community keywords do not apply to the PortMaster originating this information.

  A BGP policy  is a list of rules that restrict the BGP routes your PortMaster accepts from its peers, uses, and advertises to its peers. You can use the easy-multihome  alternative to policies--or inject-policy all  to use all routes--when you add each BGP peer to your peer group, or you can define your own policies.
  A PortMaster uses an injection policy  to determine whether to add an IP prefix to its routing table, as shown in the output of the show route  command. The PortMaster has already accepted this IP prefix for consideration as a BGP route via an acceptance policy . If the PortMaster injects the route, it will use the route to forward packets. The PortMaster also subjects the IP prefix to an advertisement policy  to determine whether to share the route with its BGP peers.
  An injection policy allows the PortMaster to receive and forward BGP routing information, but to forward packets based on simpler criteria. For example, you might want to forward packets only on routes received from OSPF or on a configured default route.
  For more information about creating injection policies, see page 9-20.
 

       set bgp policy (advertisement)

  This command creates a policy rule for advertising an IP prefix that the PortMaster learned from another peer to a BGP internal or external peer.

  Caution ¯ The creation of long, complex lists of policy rules can adversely affect PortMaster CPU performance.

  set bgp policy Policyname [before] RuleNumber
permit|deny|include Policyname
[if
[prefix [exactly] Prefix/NM]
[as-path String|empty]
[community Tag]]
[then
[local-pref Number]
[output-multi-exit-disc Number|strip]
[next-hop Ipaddress]
[community add|replace|strip Tag]
[ignore-community-restrictions]]

 
 Policyname  Name of an advertisement policy already created.
 before  Optionally inserts this BGP rule before an existing rule in the policy.
 RuleNumber  Number of a rule in the policy.

  · Use the RuleNumber of an existing rule to replace that rule.

  · Add this rule to the end of the list of rules by using a RuleNumber value that is 1 greater than the current largest rule number.

 permit  Allows the IP prefix to be advertised if the criteria in the rule are met.
 deny  Prohibits the IP prefix from being advertised if the criteria in the rule are met.
 include Policyname  Inserts an existing policy Policyname into the current policy. Included policies can themselves include other policies, up to a maximum level of 10 nested included policies.
 if  Compares the prospective IP prefix against corresponding elements specified after if  in this rule. Specifying no if  elements causes all prefixes to match the current rule.

  · If all elements of the IP prefix match these if  criteria, this rule is applied to the prefix and the prefix is either advertised or not advertised.

  · If the elements do not match, the list of policy rules is further scanned for a matching rule.

  · If no matches are found, the IP prefix is not advertised.

 prefix Prefix/NM  IP prefix Prefix and netmask NM to compare the prospective IP prefix against. The netmask indicates the number of high-order bits in the IP prefix.

  · Specify Prefix in dotted decimal notation.

  · Specify NM as number from 1 to 32, preceded by a slash (/)--for example, /24.

 By default, any prefix that matches the netmask in the rule prefix in the leftmost--most significant--bits, matches the rule prefix.
 exactly  Requires the entire prospective IP prefix and netmask to exactly match the IP prefix and netmask specified in the rule.
 as-path String  Autonomous system path String to compare the prospective IP prefix against.  String is a list of autonomous system numbers, separated by periods (.)--for example, AS1.AS2.AS3. or AS2.AS1.  When String is compared to an autonomous system path sequence , the order of the sequence must match the order of String. When String is compared to an autonomous system path set , the set  is put in ascending numerical order, and then matched against String. Multiple sequences or sets in a single autonomous system path are concatenated before being compared to String.
   The following special characters have the following meaning in the expression:

  · An asterisk (*) matches one or more entries in the autonomous system sequence.

  · A question mark (?) matches any single item in the autonomous system sequence.

 empty  Value for String that matches only paths containing no autonomous system path information.  Use as-path empty  only to permit or deny routes originating from an internal or confederation member peer within the autonomous system of the PortMaster.
 community  Identifier Tag that categorizes a group of destinations to compare the prospective IP prefix against.  See RFC 1997 for more information on a BGP community.
 Tag  32-bit number that indicates a destination category in one of the following forms:

  · One 32-bit value identifying the autonomous system of the destination

  · Two 16-bit values: one containing the autonomous system number of the destination, and the other containing additional information about the autonomous system. If only the first 16-bit word is considered significant in matching the community Tag, replace the second 16-bit value with the keyword any .

  · One of the following reserved community keywords that restrict route advertisement for peers receiving the route information:

   no-export  Destinations only within a confederation.  Advertise the route only to BGP peers within your confederation or autonomous system.
   no-advertise  No destinations.  Do not advertise this route.
   no-export-subconfed  Internal destinations only.  Advertise this route only to internal BGP peers.
   The restrictions imposed by these reserved community keywords do not apply to the PortMaster originating this information.
 then  Assigns the following metric or set of metrics to any IP prefix selected for advertisement before advertising it.
 local-pref Number  Assigns an arbitrary rating Number to an external route for advertisement to internal or confederation-member peers only. Number is a 32-bit integer.  local-pref  can be abbreviated as lp  in this command.  Higher  numbers indicate an increased preference for a specific route when more than one route exists. Use this metric to screen a particular autonomous system from your map of routes, for example.
   If you do not assign a local preference rating to the IP prefix, one of the following values is assigned by default:

  · If the route comes from an internal peer, the learned local preference number is assigned.

  · If the route comes from an external peer, Number is based on the autonomous system path length, with a shorter path being preferred.

 output-multi-exit-disc Number|strip  Assigns an arbitrary rating Number for the multiexit discriminator to an external route for advertisement to external or confederation member peers only. Number is a 32-bit integer.  A multiexit discriminator configured in a policy takes precedence over one configured in a route summarization.  output-multi-exit-disc  can be abbreviated as omed  in this command.  Lower  numbers indicate an increased preference for a specific route. Use this metric to discriminate among multiple exit or entry points between the same pair of neighboring autonomous systems.  If you do not assign a multiexit discriminator, no value is sent unless the PortMaster is advertising one of its own summarizations that specifies a multiexit discriminator. In this case, the value specified in the add bgp summarization  command is used if none is present in the policy.
   To avoid advertising any multiexit discriminator, use the strip  keyword.
 next-hop Ipaddress  Assigns the IP address to advertise as the next hop. If you do not assign a value, a value is computed automatically for the best possible next hop to reach this route. However, if this peer is configured with the set peer always-next-hop on  option, this router's local IP address is always used as the next hop.
 add  Adds the community categories identified in Tag to the IP prefix to be advertised.
 replace  Replaces the community categories identified in the community Tag of the IP prefix to be advertised with new Tag values.
 strip  Removes existing community categories from the IP prefix to be advertised.
 ignore-community-restrictions  Instructs the PortMaster to ignore the restrictive keywords no-advertise , no-export , and no-export-subconfed  when advertising this route to a peer. Use this keyword in the rule to override these restrictions received from other peers.

  A BGP policy  is a list of rules that restrict the BGP routes your PortMaster accepts from its peers, uses, and advertises to its peers. You can use the easy-multihome  alternative to policies--or advertise-policy all  to advertise all routes--when you add each BGP peer to your peer group, or you can define your own policies.
  A PortMaster uses an advertisement policy  to determine whether to share an IP prefix as a route with its internal and external BGP peers. The PortMaster has already accepted this IP prefix for consideration as a BGP route via an acceptance policy . The PortMaster also subjects the IP prefix to an injection policy  to determine whether to add an IP prefix to its routing table, as shown in the output of the show route  command. For more information about creating injection policies, see page 9-20.
 

       set bgp policy blank

  This command deletes all policy rules from a BGP policy list.

  set bgp policy Policyname blank

 
 Policyname  Name of the policy to be created or deleted--a string of up to 16 nonspace ASCII characters.

  Use the set bgp policy blank  command to remove all the policy rules from a BGP policy list.
  delete bgp policy - page 9-6
set bgp policy (acceptance) - page 9-17
set bgp policy (injection) - page 9-21
set bgp policy (advertisement) - page 9-24
 

       set bgp summarization

  This command modifies a BGP summarization entry that indicates how Interior Gateway Protocol (IGP) routing information from OSPF, RIP, or static routing is forwarded into BGP for advertisement to other BGP peers.

  set bgp summarization Prefix/NM
[as ASN] [cma ASN] [multi-exit-disc Number]
  [local-pref  Number] [community  Tag]

 
 set  Modifies an existing BGP summarization entry. All settings need to be respecified.
 delete  Deletes an existing BGP summarization entry.
 Prefix  Address prefix that you want to advertise to the BGP peers in dotted decimal notation.
 /NM  Netmask that indicates the number of high-order bits in the address prefix. This is a number from 1 to 32, preceded by a slash (/)--for example, /24.
 as  Autonomous system that receives this summarization. Include your local autonomous system number in this list to enable the summarization to go to local internal peers. You can list up to 14 autonomous systems.
 ASN  Autonomous system number.
 cma  Your confederation member autonomous system (CMAS) that receives this summarization. Include your CMAS number in this list to enable the summarization to go to internal peers in your CMAS.
 multi-exit-disc Number  Assigns an arbitrary rating Number to an external route for advertisement to external or confederation-member peers only. Number is a 32-bit integer.  multi-exit-disc  can be abbreviated as med  in this command.  Lower  numbers indicate an increased preference for a specific route. Use this metric to discriminate among multiple exit or entry points between the same pair of neighboring autonomous systems.  If you do not assign a multiexit discriminator, the value 1 is assigned by default.  A multiexit discriminator configured in a policy takes precedence over one configured in this route summarization.
   To explicitly prevent advertisement of a multiexit discriminator for IP prefixes matching this rule, set this keyword to zero (0). The PortMaster never forwards a 0 value of this metric to any peer, even if 0 was explicitly received from a peer.
 local-pref Number  Assigns an arbitrary rating Number to an external route for advertisement to internal or confederation-member peers only. Number is a 32-bit integer.  local-pref  can be abbreviated as lp  in this command.  Higher  numbers indicate an increased preference for a specific route when more than one route exists. Use this metric to screen a particular autonomous system from your map of routes, for example.
   If you do not assign a local preference rating to the IP prefix, one of the following values is assigned by default:

  · If the route comes from an internal peer, the learned local preference number is assigned.

  · If the route comes from an external peer, Number is based on the autonomous system path length, with a shorter path being preferred.

 A local preference value configured in a policy takes precedence over one configured in this summarization.
 community  Advertises the 32-bit community attribute, defined by Tag, along with this summarization.
 Tag  Thirty-two-bit number that indicates a destination category in one of the following forms:

  · One 32-bit value identifying the autonomous system of the destination

  · Two 16-bit values: one containing the autonomous system number of the destination, and the other containing additional information about the autonomous system. If only the first 16-bit word is considered significant in matching the community Tag, replace the second 16-bit value with the keyword any .

 One of the following reserved community keywords that restrict route advertisement for peers receiving the route information:
   no-export  Destinations only within a confederation.  Advertise the route only to BGP peers within your confederation or autonomous system.
   no-advertise  No destinations.  Do not advertise this route.
   no-export-subconfed  Internal destinations only.  Advertise this route only to internal BGP peers.
   The restrictions imposed by these reserved community keywords do not apply to the PortMaster originating this information.

  BGP originates to peers only the routing information that is explicitly indicated by--and supported by--the interior routing protocols in use (OSPF, RIP, static routes, or directly attached routes). These special advertisements are called summarizations , and must be explicitly configured in most cases.
  The settings you configure for community, local preference, and multiexit discriminator in this summarization command interact with advertisement policy definitions as follows:
  To help provide stability in the Internet, summarizations are advertised only when supported by one or more specific routes that exist for at least 30 seconds before the advertisement.
  set bgp policy - page 9-17
 

       show bgp memory

  This command displays information on BGP memory usage.

  show bgp memory

  Memory usage is an important concern when you are running BGP because of the large number of routes that are stored in the BGP database.
 
 Destination-specific use: 3,296,384  This value depends on the total number of IP prefixes accepted in the network layer reachability information (NLRI) from all peers, whether or not multiple peers provide the same prefix. Destination-specific bytes of memory are normally consumed only once for each unique destination.
 Peer-specific use: 3,728,096 bytes  This value depends on the total amount of information accepted from all peers. Redundant information from multiple peers can increase this value.
 

       show bgp next-hop

  This command displays the known BGP next hop addresses and the gateways to them.

  show bgp next-hop

  Use this command to conveniently determine where packets go when forwarded. The information displayed is based on entries in the routing table that are used to forward BGP packets to their destinations.
 

       show bgp paths

  This command displays BGP path information learned by the PortMaster.

  show bgp paths [Prefix/NM [verbose]]

 
 Prefix  IP prefix address, specified in dotted decimal notation. If you do not include the verbose  keyword, the display shows only the NLRI for the best match to this specified prefix address.
 /NM  Netmask that indicates the number of high-order bits in the IP prefix. This value is a number from 0 to 32, preceded by a slash (/)--for example, /24.
 verbose  Displays all the NLRI associated with the paths that the specified prefix address is on.

 
 
 Command> show bgp paths  
 O: INC  AAS: 12345  AIP: 1.2.3.4  OID: 192.168.1.130
 Cluster List: 192.168.135.1  
 Sequence: 60149 1 2 3  
 NH: 172.16.96.76 LP: 99000  MED Learned/Used: 100/200
 Metrics to NH: 3/2/0/2/0  Gateway to NH: 192.168.10.1
 Communities info: 129/129/8454273  
 NLRI: +10.24.0.0/16/8/7  
This example shows a simple path, with few routes.
 

       show bgp peers

  This command displays a list of BGP peers and, optionally, a summary of packets sent to and received from the peers.

  show bgp peers [verbose|packets]

  show table bgp

 
 verbose  Provides detailed information about BGP peers.
 packets  Provides a summary of packets sent to and received from the peers.

  Using the command without either optional keyword provides summary information. This is the default.
  The command show table bgp  displays the same output as show bgp peers .
  This message appears because a peer is not fully deleted or idled until the peer has acknowledged the close of the TCP session.
 

       show bgp policy

  This command shows BGP policy names and definitions.

  show bgp policy [Policyname]

 
 Policyname  Name of an existing policy for which you want details displayed--a string of up to 16 nonspace ASCII characters. Without this option only the names of existing BGP policies are displayed.

 

       show bgp summarization

  This command shows the route summaries configured for advertisement to BGP peers.

  show bgp summarization [all

 
 all  Displays both manually configured summaries and those automatically built with the add propagation static bgp  command. The manually configured summaries are shown with /C after the prefix and netmask, and the automatically generated ones are shown with /A. The default is to display only manually configured summaries.
]

  The following example shows a summary configured for a route to an IP address with a prefix of 10.0.0.0, a netmask of /8, and a multiexit discriminator of 5. The summary is being forwarded to autonomous systems 1, 2, and 3.
 
 Command> show bgp summarization all
 10.0.0.0/8/C  Count of Supporting Routes: 53
 LP: 0  MED: 5  CAS: no-advertise
 Export to AS: 1 2 3  
 Export to CMA: 4
 

       show routes

  This command shows the IP routing table. For more information, see the explanation of routing tables in the PortMaster Routing Guide.

  show routes [String|Prefix/NM]

 
 String  Displays only routes that contain the matching String in their show routes  command output. For example, show routes bgp  shows only routes that contain the string bgp .
 Prefix/NM  Displays routes only to the destination indicated by this IP address prefix Prefix and netmask NM. The netmask indicates the number of high-order bits in the IP prefix.

  · Specify Prefix in dotted decimal notation.

  · Specify NM as number from 1 to 32, preceded by a slash (/)--for example, /24.

[Top] [Table Of Contents] [Prev] [Next] [Index]
12 out of 25 total pages
spider@livingston.com
Copyright © 1999, Lucent Technologies. All rights reserved.