BGP Flashcards
BGP Network statemen
Ensure that the prefix in the network statement matches what is in the routing table. (under connected routes).. the network mask must be the same as in subnet seen in the routing table.
Eg 1 - if i configure a loopback interface with 10.1.1.1/32
then my bgp network statement should be -
network 10.1.1.1 mask 255.255.255.255
the routing table entry will look like this;
10.0.0.0/32 is subnetted, 1 subnets
C 10.1.1.1 is directly connected, Loopback1
Router#
================================================
Eg 2 - if i configure a loopback interface with 70.0.0.1/30
then my bgp network statement should be -
network 70.0.0.0 mask 255.255.255.252
the routing table entry will look like this;
70.0.0.0/8 is variably subnetted, 2 subnets, 2 masks
C 70.0.0.0/30 is directly connected, Loopback2
L 70.0.0.1/32 is directly connected, Loopback2
additional command that is needed to form ebgp neighbours
neighbor 1.1.1.1 ebgp-multihop 2
1. By default ios sends BGP messages with the TTL value in IP packet headers as 1, which means, routers must be directly connected. If we’re using Loopbacks to form peerings then we will have to increase the TTL value.
- When ebgp sends an update to another ebgp neighbor it changes the next hop. when ibgp sends an update to another ibgp neighbor, the next hop remains the same.
Basic BGP Peering
require neighbor statement and network statement, no auto summary and no synchronization
Split horizon rule
An update from one ibgp neighbour shall not be sent to another ibgp neighbour.
-Because of the split horizon rule, some routers may not receive bgp routes. The solution will be to configure full mesh or use route reflectors.
iBGP Peering using loopback interfaces
This solution provides redundancy.
- Loopbacks must be reachable to send unicast bgp open messages and also to advertise loopback interface inside IGP.
- Problem with source.
Default source will be the exit interface.
Change source address to loopback address.
neighbor 3.3.3.3 update-source loopback0
BGP Peer Groups
Used to apply same polices to multiple neighbors Step 1. define peer group neighbor CCIE peer-group Step2. Define common polices neighbor CCIE remote-as 500 neighbor CCIE update-source loopback1 neighbor CCIE version 4 neighbor CCIE password cisco123 Step3. Configure neighbors in Peer Group neighbor 12.0.0.1 peer-group CCIE
next hop self troubleshooting
routes may be seen in “sh ip bgp” but not in the routing table. this is because the router does not have a route for the next hop. use next-hop-self as a solution
reading sh ip bgp table.
AS Path - route with the shortest AS Path list is more desirable.
Origin - “i” = advertised in bgp | “?” = advertised from other routing protocols. | “e” = learning routes from older version of bgp.
Next hop with 0.0.0.0 means directly connected routes
Weight Attribute
The WEIGHT attribute is local to the router and does not get exchanged between routers; therefore it is only effective on in-bound route maps
BGP Path Vector protocol
BGP advertisements will contain the AS path information in them.
BGP Features
- It sends updates to manually defined neighbors as unicast
-It is an application layer protocol and use TCP port 179
-Metric is attributes
Administrative distance - Internal [200] External [20]
Types of ISP Connections - Single Homed
Single Homed
- A site with a single ISP connection
- Either use static routes, or advertise the site routes to the ISP and receive a default route from the ISP.
Types of ISP Connections - Dual-Homed Site
- has two connections to the same ISP. From either one router or two routers.
- one link is primary and the other secondary or traffic is load balanced across both links.
- dynamic or static routing would work.
Types of ISP Connections - Multi-Homing
- Means connecting to more than one ISP at the same time.
- Used for redundancy
- BGP is typically used with multi-homing environments.
Types of ISP Connections - Dual-MultiHomed
- Means connecting to more than one ISP at the same time.
- You have atleast two links going back to each ISP
