1. Border Gateway Protocol 2008.3.27 DPNM Lab. Seongho Cho (nology@postech.ac.kr)

2. Contents  Introduction  BGP Operation  How Does BGP Work?  BGP Message Types  BGP Packet Formats  Open, Update, Notification, Keep-alive  Loopback Interface  EBGP Multihop & Load Balancing  Synchronization  BGP Filtering  Route Filtering, Path Filtering, Route Maps  BGP Attributes  Origin  Next hop  AS path  Local preference, ……  Path selection algorithm  Q&A

3. Introduction  Inter-Autonomous System routing protocol  Interior BGP (IBGP) - IBGP is NOT an IGP.  Exterior BGP (EBGP)  Basic activities  Determination of optimal routing paths  The transport of information  Request For Comments  RFC 1771-Describes BGP4  RFC 1654-Describes the first BGP4 specification  RFC 1105, 1163, 1267-Describes versions of BGP prior to BGP4

4. BGP Operations  BGP performs three types of routing  Inter-autonomous system routing  Between 2+ BGP routers in different AS.  To maintain a consistent view of the inter-network topology.  BGP neighbors must reside on the same physical network.  Intra-autonomous system routing  Between 2+ BGP routers located within the same AS.  To maintain a consistent view of the system topology.  To determine which router will serve as the connection point for specific external AS.  Pass-through autonomous system routing  Between 2+ BGP peer routers that exchange traffic across an AS that does not run BGP.  BGP must interact with whatever intra-AS routing protocol is being used to successfully transport BGP traffic through that AS.

5. How Does BGP Work?  BGP uses TCP as its transport protocol (port 179).  BGP maintains routing tables, transmits routing updates, and bases routing decision on routing metrics.  BGP routers exchange network reachability information including information about the list of AS paths.  BGP neighbors exchange full routing information when the TCP connection is first established.  BGP routers send to their neighbors only those routes that have changed.  BGP routers do not send periodic routing updates.  BGP routing updates advertise only the optimal path to a destination network.

6. BGP Message Types  Open message  opens a BGP communications session  is the first message sent by each side after a transport-protocol connection is established.  is confirmed using a keep-alive message and must be confirmed before any other messages can be exchanged.  Update Message  is used to provide routing updates to other BGP systems.  is sent by TCP to ensure reliable delivery.  Notification Message  is sent when an error condition is detected.  Is used to close an active session and to inform any connected router of why the session is being close.  Keep-alive Message  notifies BGP peers that a device is active.  is sent often enough to keep the session from expiring.

7. BGP Packet Formats  Header Format  All BGP message types use the basic packet header and have additional fields, but keep-alive message use only the basic packet header.   BGP Packet-Header Fields  Maker – Contains an authentication value that the message receiver can predict.  Length – Indicates the total length of the message in bytes.  Type – Specifies the message type as one of the following:  Open;  Update;  Notification;  Keep-alive  Data – Contains upper-layer information in this optional field.

8. BGP Packet Formats  Open Message Format  BGP open messages are comprised of a BGP header and additional fields. 

9. BGP Packet Formats  Open Message Format  BGP Open Message Fields  These fields provide the exchange criteria for two BGP routers to establish a peer relationship.  Version - Provides the BGP version number so that the recipient can determine whether it is running the same version as the sender.  Autonomous System - Provides the autonomous system number of the sender.  Hold-Time - Indicates the maximum number of seconds that can elapse without receipt of a message before the transmitter is assumed to be nonfunctional.  BGP Identifier - Provides the BGP identifier of the sender (an IP address), which is determined at startup and is identical for all local interfaces and all BGP peers.  Optional Parameters Length - Indicates the length of the optional parameters field (if present).  Optional Parameters - Contains a list of optional parameters (if any). Only one optional parameter type is currently defined: authentication information. Authentication information consists of the following two fields:  Authentication code: Indicates the type of authentication being used.  Authentication data: Contains data used by the authentication mechanism (if used).

10. BGP Packet Formats  Update Message Format  BGP update messages are comprised of a BGP header and additional fields. 

11. BGP Packet Formats  Update Message Format  BGP Update Message Fields  Upon receiving an update message packet, routers will be able to add or delete specific entries from their routing tables.  Unfeasible Routes Length - Indicates the total length of the withdrawn routes field or that the field is not present.  Withdrawn Routes - Contains a list of IP address prefixes for routes being withdrawn from service.  Total Path Attribute Length - Indicates the total length of the path attributes field or that the field is not present.  Path Attributes - Describes the characteristics of the advertised path. The following are possible attributes for a path:  Origin: Mandatory attribute that defines the origin of the path information  AS Path: Mandatory attribute composed of a sequence of autonomous system path segments  Next Hop: Mandatory attribute that defines the IP address of the border router that should be used as the next hop to destinations listed in the network layer reachability information field  Multi Exit Disc: Optional attribute used to discriminate between multiple exit points to a neighboring autonomous system  Local Preference: Discretionary attribute used to specify the degree of preference for an advertised route  Atomic Aggregate: Discretionary attribute used to disclose information about route selections  Aggregator: Optional attribute that contains information about aggregate routes  Network Layer Reachability Information (NLRI) - Contains a list of IP address prefixes for the advertised routes

12. BGP Packet Formats  Notification Message Format  BGP notification messages are comprised of a BGP header and additional fields 

13. BGP Packet Formats  Notification Message Format  BGP Notification Message Fields  This packet is used to indicate some sort of error condition to the peers of the originating router.  Error Code - Indicates the type of error that occurred. The following are the error types defined by the field:  Message Header Error: Indicates a problem with a message header, such as unacceptable message length, unacceptable marker field value, or unacceptable message type.  Open Message Error: Indicates a problem with an open message, such as unsupported version number, unacceptable autonomous system number or IP address, or unsupported authentication code.  Update Message Error: Indicates a problem with an update message, such as a malformed attribute list, attribute list error, or invalid next-hop attribute.  Hold Time Expired: Indicates that the hold-time has expired, after which time a BGP node will be considered nonfunctional.  Finite State Machine Error: Indicates an unexpected event.  Cease: Closes a BGP connection at the request of a BGP device in the absence of any fatal errors.  Error Subcode - Provides more specific information about the nature of the reported error.  Error Data - Contains data based on the error code and error subcode fields. This field is used to diagnose the reason for the notification message.

14. BGP Filtering  Sending and receiving BGP updates can be controlled by using a number of different filtering methods based:  on route information;  on path information;  on community.  Route Filtering  Path Filtering  Route Maps  Regular Expression

15. BGP Filtering Route Filtering  Based on the network address information  Syntax  ip prefix-list [list-name] [seq seq-num] [permit/deny] netwrok/len [le/ge] length  Example (in R1)  ip prefix-list Block-128 permit 100.1.0.0/16 le 24  ip prefix-list Block-128 permit 100.1.2.0/24 le 25

16. BGP Filtering Path Filtering  Based on the BGP AS path information.  Syntax  neighbor [peer-address] filter-list [as-path access-list number] [in/out]  ip as-path access-list [number] permit [regexp]  Example (in R1)  neighbor 10.3.3.2 filter-list 10 out  ip as-path access-list 10 permit ^$  ip as-path access-list 10 deny.* this AS any path

17. BGP Filtering Route Maps  Syntax  route-map [map-name] [permit/deny] [sequence-number]  match [condition]  set [statement]  neighbor [peer-address] route-map [map-name] [in/out]  Example (in R1)  neighbor 10.3.3.2 route-map My-AS out  route-map My-AS permit 10  match as-path 10  ip as-path access-list 10 permit ^$  Ip as-path access-list 10 deny.*

18. BGP Filtering Regular Expression  A regular expression is a pattern to match against an input string.  Range  A range is a sequence of characters contained within left and right square brackets. For example: [abcd]  Atom  An atom is a single character, such as the following: . (Matches any single character)  ^ (Matches the beginning of the input string)  $ (Matches the end of the input string)  \ (Matches the character)  − (Matches a comma (,), left brace ({), right brace (}), the beginning of the input string, the end of the inputstring, or a space.)  Piece  A piece is an atom followed by one of the following symbols:  * (Matches 0 or more sequences of the atom)  + (Matches 1 or more sequences of the atom)  ? (Matches the atom or the null string)

19. BGP Filtering Regular Expression  Examples of regular expressions follow:  a*  Any occurrence of the letter "a", including none  a+  At least one occurrence of the letter "a" should be present  ab?a  This matches "aa" or "aba"  _100_  Via AS100  ^100$  Origin AS100  ^100.*  Coming from AS100  ^$  Originated from this AS

20. Loopback Interface  Using a loopback interface to define neighbors is common with IBGP, but not with EBGP.  In the case of EBGP, peer routers are frequently directly connected and loopback does not apply.  Normally the loopback interface is used to make sure the IP address of the neighbor stays up and is independent of hardware functioning properly.

21. EBGP Multihop & Load Balancing  In some cases, a router can run EBGP with a third party router that doesn't allow the two external peers to be directly connected.  By using EBGP multihop, the neighbor connection to be established between two non-directly-connected external peers.  To achieve load balancing between two EBGP speakers over parallel serial lines.  In normal situations, BGP picks one of the lines to send packets on, and load balancing wouldn't happen.  By introducing loopback interfaces, the next hop for EBGP is the loopback interface.  RTA has two choices to reach next hop 160.10.1.1: one via 1.1.1.2 and the other one via 2.2.2.2, and the same for RTB.

22. Synchronization  If the AS is passing traffic from another AS to a third AS, BGP should not advertise a route before all routers in the AS have learned about the route via IGP.  BGP will wait until IGP has propagated the route within the AS and then will advertise it to external peers.  This is called synchronization.  Do not need synchronization  If all routers in the AS will be running IBGP,  If the AS will not be passing traffic from a different AS through the AS,

23. BGP Attributes  Origin Attribute  The origin attribute is a mandatory attribute that defines the origin of the path information.  The origin attribute indicates how BGP learned about a particular route.  IGP - The route is interior to the originating AS. This value is set when the network router configuration command is used to inject the route into BGP.  EGP - The route is learned via the Exterior Border Gateway Protocol (EBGP).  Incomplete - The origin of the route is unknown or learned in some other way. An origin of incomplete occurs when a route is redistributed into BGP.

24. BGP Attributes  Origin Attribute  RTA will reach 170.10.0.0 via: 300 i (which means the next AS path is 300 and the origin of the route is IGP).  RTA will also reach 190.10.50.0 via: i (which means, the entry is in the same AS and the origin is IGP).  RTE will reach 150.10.0.0 via: 100 i (the next AS is 100 and the origin is IGP).  RTE will also reach 190.10.0.0 via: 100 ? (the next AS is 100 and the origin is incomplete "?", coming from a static route).

25. BGP Attribute  Next-Hop Attribute  The EBGP next hop attribute is the next hop IP address that is going to be used to reach the advertising router.  For EBGP peers, the next-hop address is the IP address of the connection between the peers.  For IBGP, the EBGP next-hop address is carried into the local AS.  Exchanging Next-Hop  Static Route  Dynamic Routing Protocol like OSPF  next-hop-self

26. BGP Attribute  Next-Hop Attribute  Static Route: ip route 10.2.2.0 255.255.255.0 10.1.1.3 (in R1)  next-hop-self: neighbor 10.1.1.2 next-hop-self (in R2)  Example (in R1)  50.1.1.0/2410.2.2.2  50.1.1.0/2410.1.1.3

27. BGP Attribute  AS_path Attribute  Whenever a route update passes through an AS, the AS number is prepended to that update.  The AS_PATH attribute is actually the list of AS numbers that a route has traversed in order to reach a destination.  Detecting routing loops  when its own AS number is detected in the route advertisement.

28. BGP Attribute  AS_Path Attribute  Network 190.10.0.0 is advertised by RTB in AS200, when that route traverses AS300 and RTC will append its own AS number to it.  So when 190.10.0.0 reaches RTA it will have two AS numbers attached to it: first 200 then 300.  So as far as RTA is concerned the path to reach 190.10.0.0 is (300,200).

29. BGP Attribute  Local Preference Attribute  The local preference attribute is used to prefer an exit point from the local AS.  The local preference attribute is propagated throughout the local AS.  If there are multiple exit points from the AS, the local preference attribute is used to select the exit point for a specific route.

30. BGP Attribute  Local Preference Attribute  AS 100 is receiving two advertisements for network 172.16.1.0 from AS 200.  When Router A receives the advertisement for network 172.16.1.0, the corresponding local preference is set to 50.  When Router B receives the advertisement for network 172.16.1.0, the corresponding local preference is set to 100.  These local preference values will be exchanged between routers A and B.  Because Router B has a higher local preference than Router A, Router B will be used as the exit point from AS 100 to reach network 172.16.1.0 in AS 200.

31. BGP Attribute  Community Attribute  The community attribute provides a way of grouping destinations, called communities, to which routing decisions can be applied.  Route maps are used to set the community attribute.  Predefined community attributes are listed here:  no-export - Do not advertise this route to EBGP peers.  no-advertise - Do not advertise this route to any peer.  internet - Advertise this route to the Internet community; all routers in the network belong to it.

32. BGP Attribute  Community Attribute  AS 1 advertises 172.16.1.0 to AS 2 with the community attribute no- export. AS 2 will propagate the route throughout AS 2 but will not send this route to AS 3 or any other external AS.  AS 1 advertises 172.16.1.0 to AS 2 with the community attribute no- advertise. Router B in AS 2 will not advertise this route to any other router.  There are no limitations to the scope of the route advertisement from AS 1.

33. BGP Attribute  Weight Attribute  The weight attribute is a Cisco defined attribute.  The weight attribute is not advertised to neighboring routers.  If the router learns about more than one route to the same destination, the route with the highest weight will be preferred.

34. BGP Attribute  Weight Attribute  Router A is receiving an advertisement for network 172.16.1.0 from routers B and C.  When Router A receives the advertisement from Router B, the associated weight is set to 50.  When Router A receives the advertisement from Router C, the associated weight is set to 100.  Both paths for network 172.16.1.0 will be in the BGP routing table, with their respective weights.  The route with the highest weight will be installed in the IP routing table.

35. BGP Attribute  Multi-Exit Discriminator (MED) Attribute  The multi-exit discriminator (MED) or metric attribute is used as a suggestion to an external AS regarding the preferred route into the AS that is advertising the metric.  The term suggestion is used because the external AS that is receiving the MEDs may be using other BGP attributes for route selection.

36. BGP Attribute  Multi-Exit Discriminator (MED) Attribute  Router C is advertising the route 172.16.1.0 with a metric of 10, while Route D is advertising 172.16.1.0 with a metric of 5.  The lower value of the metric is preferred, so AS 100 will select the route to router D for network 172.16.1.0 in AS 200.  MEDs are advertised throughout the local AS.

37. BGP Attribute  Atomic aggregate & aggregator Attribute  One of the main enhancements of BGP4 over BGP3 is Classless Inter Domain Routing (CIDR).  CIDR or supernetting is a new way of looking at IP addresses and of reducing the routing table.

38. BGP Attribute  Atomic aggregate & aggregator Attribute

39. Path Selection Algorithm  BGP could possibly receive multiple advertisements for the same route from multiple sources.  BGP selects only one path as the best path.  When the path is selected, BGP puts the selected path in the IP routing table and propagates the path to its neighbors.

40. Path Selection Algorithm  BGP uses the following criteria, in the order to select a path for a destination:  If the path specifies a next hop that is inaccessible, drop the update.  Prefer the path with the largest weight.  If the weights are the same, prefer the path with the largest local preference.  If the local preferences are the same, prefer the path that was originated by BGP running on this router.  If no route was originated, prefer the route that has the shortest AS_path.  If all paths have the same AS_path length, prefer the path with the lowest origin type (where IGP is lower than EGP, and EGP is lower than incomplete).  If the origin codes are the same, prefer the path with the lowest MED attribute.  If the paths have the same MED, prefer the external path over the internal path.  If the paths are still the same, prefer the path through the closest IGP neighbor.  Prefer the path with the lowest IP address, as specified by the BGP router ID.

41. References  CISCO Documentation – Border Gateway Protocol (BGP) http://www.cisco.com/univercd/cc/td/doc/cisintwk/ito_doc/bgp.htm http://www.cisco.com/univercd/cc/td/doc/cisintwk/ito_doc/bgp.htm  Border Gateway Protocol (BGP) http://www.pulsewan.com/data101/pdfs/bgp.pdf http://www.pulsewan.com/data101/pdfs/bgp.pdf  BGP Case Studies (Document ID:26634) http://www.cisco.com/warp/public/459/bgp-toc.html http://www.cisco.com/warp/public/459/bgp-toc.html  BGP Techniques for Internet Service Providers http://www.nanog.org/mtg-0606/pdf/philip-smith.pdf http://www.nanog.org/mtg-0606/pdf/philip-smith.pdf  BGP for Internet Service Providers http://www.ripe.net/ripe/meetings/ripe-40/tutorials/bgp-tutorial/index.html http://www.ripe.net/ripe/meetings/ripe-40/tutorials/bgp-tutorial/index.html  Introduction to the Border Gateway Protocol (BGP) http://www.academ.com/nanog/feb1997/BGPTutorial/index.htm http://www.academ.com/nanog/feb1997/BGPTutorial/index.htm  BGP4.AS http://www.bgp4.as/ http://www.bgp4.as/  CISCO Routers Help, CCIE lab. http://www.routergod.com/ http://www.routergod.com/  Advanced CISCO Networking, 성안당  http://www.enclue.com/library/protocol_bgp.html http://www.enclue.com/library/protocol_bgp.html

42. Questions ?