BGP Concept There are two types of protocols: Interior Gateway Protocol (IGP) Exterior Gateway Protocol (EGP) An IGP is a routing protocol that operates within the same Autonomous System (AS) such as OSFP and EIGRP. Usually routers running IGP are under the same administration or company. An EGP is a routing protocol that operates between different AS. BGP is the only EGP used nowadays Border Gateway Protocol (BGP) BGPv4 is a standardized protocol and is considered as a Path Vector protocol. It is used to route traffic between different AS and it maintains a separate routing table based on the shortest AS Path to a destination. BGP is the routing protocol used on the internet or between different ISPs and it is identified using a 16-bits number, known as Autonomous System Number (ASN) which can be a value from 1-65535. BGP ASN 64512-65535 for private use. One of the differences that BGP has compared to other IGPs is that BGP does not require neighbors to be attached to the same subnets. Instead, BGP routers use a TCP connection (Port 179) between routers to pass BGP messaged, allowing neighboring routers to be on the same subnet or to be separated by several routers. Some of the features of BGP are: • It forms a neighbor relationship before sending routing information (same as OSFP and EIGRP) • Neighbor IP addresses are explicitly configured and may not be on common subnet. • Advertise prefix/length, called Network Layer Reachability Information (NLRI). • Advertises a variety of Path Attributes (PA) that BGP uses instead of a metric to choose the best path. • Focus on fast scalability and might not always choose the most efficient route. PATH VECTOR PROTOCOL Path Vector protocol does not rely on bandwidth of the link (like OSPF) or hopcount (like RIP) or a group of parameters (like EIGRP). It relies on the number of autonomous system it has to go through before reaching its destination. In other words, it chooses the path with the least number of autonomous system (shortest AS path) to reach a destination, provided that the path is loop free. The path can be changed easily based on our requirements. BGP uses the AS path to perform two key functions: • Choose the best route for a prefix based on the fewest ASNs it has to go through • Prevent routing loops Let’s take an example: If AS1 needs to communicate with AS3, BGP will go through AS1>AS2>AS. So BGP will go through AS2 as it only has to go through only 1 AS which is AS. One reason to use BGP is that it can handle very big routing tables, which IGP cannot handle. Currently the global Internet routing table contains over 500,000 routes and BGP handles such large routing table. BGP Message Types For BGP to function, BGP routers (known as speakers) must form a neighbor relationship (known as peers). There are two types of BGP neighbor relationships: • iBGP peers – BGP neighbors within the same AS • eBGP peers – BGP neighbors in different AS Let’s take an example: In the example above, Router B, Router C and Router D are in AS 200 and in order for them to communicate, an iBGP neighbor relationship would be form. For Router A which is in AS 100 to communicate with Router E which is in AS 300, an eBGP peering would be formed. Once all the BGP peers have full neighbor relationship, they will send their full routing table to each other. When fully converged, only updates are then exchanged. By default, BGP assumes that eBGP peers are a maximum of one hop away from each other. But this can be changed using a command. iBGP does not have any hop restriction. By default, all iBGP peers must be directly connected with each other within the same AS number. The administrative distance for routes learned by eBGP is 20 and iBGP is 200. Administrative distance is the feature that routers use in order to select the best path when there are two or more different routes to the same destination from two different routing protocols. BGP PEER MESSAGES BGP will form its peer relationship using a series of messages: • OPEN • KEEPALIVE • UPDATE • NOTIFICATION OPEN MESSAGE First an OPEN message is sent between peers to initiate the session. The OPEN message contains several parameters: • BGP version – must be the same between BGP peers • Local AS Number • BGP Router ID KEEPALIVE MESSAGE Keepalive messages are sent by default every 60 seconds periodically to ensure that the remote peer is still alive. If a router does not receive keepalive message for a hold-time period of 180 seconds ( by default) the router declares that peer dead. UPDATE MESSAGE Update messages are used to exchange routes between peers. NOTIFICATION MESSAGE Notification messages are sent there is an error condition. If that message is sent, the BGP peer session is torn down and reset. Finite State Machine (FSM) When BGP session is forming between peers, it will go through several state known as BGP Finite-Sate Machine (FSM). The states are: • IDLE This is the first state of BGP where it will try to initiate a TCP connection to the BGP peer and also listens for a new peer router. • CONNECT In this state, BGP initiates the TCP connection. If the 3-way TCP handshake completes, the established BGP session BGP process resets the ConnectRetry timer and sends the open message to the neighbor, and then changes to the OPENSENT state. If the connectivity timer is over before this stage is completed, a new TCP connection is attempted, the connectretry timer is reset and the state is moved to Active. If any other input is received, the state is changed to Idle. During this state, the neighbor with the highest IP address manages the connection. The router initiating the request uses a dynamic source port, but the destination port is always 179. ACTIVE In this state, BGP starts a new 3-way TCP handshake. If a connection is established, an Open message is sent, the Hold Timer is set to 4 minutes and the state moves to OPENSENT. If this attempt for the TCP connection fails, the state moves back to the connect state and resets the connnectretry timer. OPENSENT In this state, an OPEN message has been sent from the originating router and is waiting an OPEN message form the other router. When the source router receives the OPEN message form the other router, both OPEN messages are check for errors. The following items are compared: • BGP versions must match. • The source IP address of the Open message must match the IP address that is configured for the neighbor. • The AS number in the Open message must match what is configured for the neighbor. • BGP identifiers (RID) must be unique. If RID does not exists, this condition is not met. • Security parameters like password, TTL, etc…., are also compared. If the OPEN messagedo not have any errors, the connection state is then moved to the OPENCONFIRM. If an error is found, a NOTIFICATION message is sent and the state is moved back to IDLE. OPENCONFIRM In this state, BGP waits for a keepalive message. Upon receipt of a neighbor's keepalive message, the state is moved to established. If the hold timer expires, the state is moved to Idle. ESTABLISHED In this state, BGP session is established. BGP neighbors exchange routes using Update messages. As Update and Keepalive messages are received, the Hold Timer is reset. If the Hold Timer expires, an error is detected and BGP moves the neighbor back to the Idle state. BGP General Concept • TYPES OF CONNECTION TO THE ISP I. Single – homed (1 link per ISP, 1 ISP) II. Dual – homed (2+ links per ISP, 1 ISP) III. Single – multihomed ( 1 link per ISP, 2+ ISPs) IV. Dual – multihomed (2+ links per ISP, 2+ ISPs) Single – homed It uses a single link between the enterprise and the ISP. Only one possible next hop router exist for any or all routes for destinations in the internet. An example is shown below: Dual – homed It has two (or more) links to the Internet, but with all links connecting to a single ISP. Comparing the dual – homed to single – homed, the second link gives the enterprise a choice. The routers could choose between one of the two links to send data. PARTIAL AND FULL BGP UPDATES Exchanging BGP information for large number of routes consumes bandwidth. It also consumes memory in the routers and requires some processing to choose the best routes. To help reduce the memory requirements of receiving full BGP updates (BGP updates that include all routes), some ISPs give you 3 basic options for what routes the ISP advertises: • Default route only – The ISP advertises a default route with BGP, but no other routes. • Full updates – The ISP sends you the entire BGP table. • Partial updates – The ISP sends you routes for prefixes that might be better reached through that ISP, but not all routes, plus a default route (to use as needed instead of the purposefully omitted routes). BGP partial updates give you the benefits of choosing the best route for some destination, while limiting the bandwidth and memory consumption. With partial updates, the ISP advertises routes for prefixes that are truly better reached through a particular link. Single – multihomed A single- multihomed topology means a single link per ISP, but multiple (at least two) ISPs. An example is shown below: Dual – multihomed Two or more ISPs are used, with two or more connections to each. Some examples are shown below: At least two ISPs exist with at least two connections per ISP, much redundancy exists. That redundancy can be used for backup, but most often, BGP used it to make some decision about the best path to reach various destinations. Configuring BGP Neighbors The first step in configuring BGP is to enable the BGP process, and specify the router’s Autonomous System (AS): RouterB(config)# router bgp 100 RouterB is now a member of AS 100. Next, neighbor relationships must be established. To configure a neighbor relationship with a router in the same AS (iBGP Peer): RouterB(config)# router bgp 100 RouterB(config-router)# neighbor 10.1.1.1 remote-as 100 To configure a neighbor relationship with a router in a separate AS (eBGP Peer): RouterB(config)# router bgp 100 RouterB(config-router)# neighbor 172.16.1.2 remote-as 900 Notice that the syntax is the same, and that the remote-as argument is always used, regardless if the peering is iBGP or eBGP. For stability purposes, the source interface used to generate updates to a particular neighbor can be specified: RouterB(config)# router bgp 100 RouterB(config-router)# neighbor 172.16.1.2 update-source lo0 RouterC must then point to RouterB’s loopback (assume the address is 1.1.1.1/24) in its neighbor statement: RouterC(config)# router bgp 900 RouterC(config-router)# neighbor 1.1.1.1 remote-as 100 RouterC must have a route to RouterB’s loopback in its routing table. Configuring BGP Neighbors (continued) Remember though: by default, BGP assumes that external peers are exactly one hop away. Using the loopback as a source interface puts RouterB two hops away from RouterC. Thus, the ebgp-multihop feature must be enabled: RouterC(config)# router bgp 900 RouterC(config-router)# neighbor 1.1.1.1 ebgp-multihop 2 The 2 indicates the number of hops to the eBGP peer. If left blank, the default is 255. To authenticate updates between two BGP peers: RouterB(config)# router bgp 100 RouterB(config-router)# neighbor 172.16.1.2 password CISCO Configuring BGP Timers To globally adjust the Keepalive and Hold-time timers for all neighbors: RouterB(config)# router bgp 100 RouterB(config-router)# timers bgp 30 90 The above command sets the Keepalive timer to 30 seconds, and the Hold- time timer to 90 seconds. If the configured Hold-time timers between two peers are different, the peer session will still be established, and the smallest timer value will be used. To adjust the timers for a specific neighbor (which overrides the global timer configuration): RouterB(config)# router bgp 100 RouterB(config-router)# neighbor 172.16.1.2 timers 30 90 Viewing BGP Neighbors To view the status of all BGP neighbors: RouterB# show ip bgp neighbors BGP neighbor is 172.16.1.2, remote AS 900, external link Index 1, Offset 0, Mask 0x2 Inbound soft reconfiguration allowed BGP version 4, remote router ID 172.16.1.2 BGP state = Established, table version = 27, up for 00:03:45 Last read 00:00:19, hold time is 180, keepalive interval is 60 seconds Minimum time between advertisement runs is 30 seconds Received 25 messages, 0 notifications, 0 in queue Sent 20 messages, 0 notifications, 0 in queue Inbound path policy configured Route map for incoming advertisements is testing Connections established 2; dropped 1 Connection state is ESTAB, I/O status: 1, unread input bytes: 0 Local host: 172.16.1.1, Local port: 12342 Foreign host: 172.16.1.2, Foreign port: 179 Enqueued packets for retransmit: 0, input: 0, saved: 0 Event Timers (current time is 0x530C294): Timer Starts Wakeups Retrans 15 0 TimeWait 0 0 AckHold 15 13 SendWnd 0 0 KeepAlive 0 0 GiveUp 0 0 PmtuAger 0 0 <snip> To view the status of a specific BGP neighbor: RouterB# show ip bgp neighbors 172.16.1.2 Next 0x0 0x0 0x0 0x0 0x0 0x0 0x0 The BGP Routing Table Recall that BGP maintains its own separate routing table. This table contains a list of routes that can be advertised to BGP peers. To view the BGP routing table on RouterB: RouterB# show ip bgp BGP table version is 426532, local router ID is 2.2.2.2 Status codes: s suppressed, * valid, > best, i - internal Origin codes: i - IGP, e - EGP, ? - incomplete Network *> 10.5.0.0 Next Hop 0.0.0.0 Metric LocPrf Weight Path 0 0 32768 i The route has been injected into BGP using the network command. The Next Hop of 0.0.0.0 indicates that the route was locally originated into BGP. The Path is empty, as the route originated in the Autonomous Systems. Notice the Status Codes of “*>”. The * indicates that this route is valid (i.e. in the routing table). The > indicates that this is the best route to the destination. BGP will never advertise a route to an eBGP peer unless it is both valid and the best route to that destination. BGP routes that are both valid and best will also added the IP routing table as well. To view the BGP routing table on RouterC: RouterC# show ip bgp Network *> 10.5.0.0 Next Hop 172.16.1.1 Metric LocPrf Weight Path 0 100 0 100 i Notice that AS 100 has been added to the path, and that the Next Hop is now RouterB.
