BGP Routing for Cloud Management with IOS XE

What is iBGP?

Internal Border Gateway Protocol or iBGP is a BGP version used by routers in the same autonomous system (AS). Without using an IGP (Interior Gateway Protocol), iBGP enables routers to share routing data about external networks, such as the Internet. To avoid routing loops, iBGP peers must create a full mesh topology. iBGP bases its routing choices on network regulations and rulesets by using BGP features like local preference.

Characteristics of iBGP

• iBGP operates within the same Autonomous System (AS).

• The next-hop addresses remain unchanged when routes are share
• It employs Split-Horizon Rules to prevent loops.

Advantages of iBGP

• Route-Reflection: Route reflection helps eliminate the need for a large number of iBGP peers
• Easy Configuration: iBGP is simpler to configure than eBGP, with Route-Reflectors and Peer-Groups simplifying configuration in bulk.
• Redundancy: In the case of failures, iBGP offers multiple routes for data path

Disadvantages of iBGP

• Full Mesh Requirement: iBGP requires a Full Mesh of iBGP peers. This results in numerous connections and can impact scale/performance
• Route Oscillation: Although iBGP is very effective, it is exposed to route oscillation, which in turn makes the network unstable.
• Slow Convergence: It takes relatively more time for the network to attain its stable state in case of a failure or a change.

What is eBGP?

An external Border Gateway(eBGP) allows communication between several autonomous systems (AS). Network connections via the Internet or between various organizations are made possible by eBGP. eBGP operates in the opposite manner of iBGP, i.e., inside the same AS. eBGP routes don’t need a full mesh topology and have an administrative distance of 20.

Characteristics of eBGP

• It can operate between different Autonomous Systems.
• Whenever a route is advertised, the next-hop address is updated.
• It employs AS Path Attribute to prevent loops.

Advantages of eBGP

• Inter-AS Connectivity: With eBGP, one can easily communicate between separate networks in different AS
• Flexibility in Route Advertisement: eBGP gives greater flexibility on which routes are advertised to neighboring AS.
• Improved Network Resilience: eBGP assists in improving network resilience by offering multiple paths for data transmission. 
• Easy Troubleshooting: Easier to troubleshoot issues between each AS since each AS can be isolated and can be debugged independently.

Disadvantages of eBGP

• Higher Configuration Overhead: Although eBGP is easier than iBGP, it requires more configuration effort.  
• Route Leaking: Poorly configured eBGP is prone to route leaking, which in turn causes network instability and security issues.
• Distance Vector Limitations: eBGP uses distance vector routing, which in most cases can result in the formation of routing loops if not well controlled.
• Higher Network Resource Requirements: eBGP is more network resource intensive so hardware and scale must be considered for large deployments.

Peer-Groups

1. To create a Peer-Group, select “Add Peer Group” from inside the BGP router instance.

    

    

2. Create your Peer-Group, this contains common configuration elements for peers.     

    

    

3. Be aware that some settings are only available for iBGP or eBGP respectively.

• iBGP Only Features 

  • Route-reflector-client 

• eBGP Only Features 

  • next-hop-unchanged 

  • Disable-connected-check 

  • ebgp-multihop 

  • remove-private-as 

Peers and Ranges

Within the Peer Group you can create multiple 'peer' or 'range' entries in each Peer Group. When configuring peers & ranges, be aware that ranges are passive constructs which cannot connect to other ranges. 
Example: Use of "Range" on Core/Distribution level switches can simplify config across multiple IDFs/remote switches. Each remote switch would have a single peer configuration pointing to the Core switch, removing the need from building 2-way peer configurations on each core/IDF pair.
 

Connectivity Compatibility	Peer	Range (Passive)
Peer	✓	✓
Range (Passive)	✓	X

Peer: Neighbor IP of the remote iBGP/eBGP peer. Peers inherit peer group configurations by default but allow for overrides per peer.
Range: When you create a 'range', you specify a subnet range for the remote peers. Configuration is inherited from the peer-group and can be overriden with a peer config.
For example, a range of "192.168.0.0/24" would allow any remote peer in that subnet to connect to the local router.

Users can configure overlapping peers and ranges if desired. Peer config will take priority over range config
For example, user has a range of "192.168.0.0/24" and a single peer config of "192.168.0.100". The single peer config will take priority over the common peer config.
 

1. Create a Peer or Range. 

    

    

2. To create a “Peer” select "Add Peer" button

• Peers will inherit all the settings from the Peer-group by default.

• (OPTIONAL) Peers can override certain configurations if needed. Any field that isn't overridden will preserve that config from the peer-group.

   

   

• You can then override the following fields from the Peer Group.

  Weight, default-originate, soft-reconfiguration, Prefix-list Inbound, AS-Path List  Inbound, description, password, disable-connected-check, ebgp-multihop, update-source.

3. To create a “Range” click on "Add Range". This allows ANY router from a specific subnet to peer with the local bgp router.

    

    

   • The local BGP router would be configured for a Range. Remote BGP peers in that subnet range would have a single peer config pointing back to that local router.

     • (ex. Local BGP Router is the MDF switch, and all the remote IDF switches peer back to the MDF. In this example the MDF is configured with a listen-range and all the IDFs are configured with a single peer pointing to the MDF) 

        

        

Range-to-Peers (1:Many) Example:

Peer-to-Peer Example:

AS-Path Filters and Regex

1. Access-lists can be created with multiple entries per list
   
   
2. They're reference via "number" and can be applied to filter inbound or outbound exchanges

AS Regular Expression

This section explains the creation of a regular expression.

A regular expression is a pattern to match against an input string. When you build a regular expression, you specify a string that input must match. In the case of BGP, you specify a string that consists of path information that an input must match.

In the example in the section  Path Filter , you specified the string^200$. You wanted path information that comes inside updates to match the string in order to decide.

A regular expression comprises:

Range
A range is a sequence of characters within left and right square brackets. An example is[abcd].

Atom

An atom is a single character. Here are some examples:
.    The . matches any single character.
^    The ^ matches the start of the input string.
$   The $ matches the end of the input string.
\    The \ matches the character.
-    The_matches a comma (,), left brace ({), right brace (}), the start of the input string, the end of the input string, or a space.

Piece

A piece is one of these symbols, which comes after an atom:
*    The * matches 0 or more sequences of the atom.
+    The + matches 1 or more sequences of the atom.
?    The ? matches the atom or the null string.

Branch
A branch is 0 or more concatenated pieces.

Here are some examples of regular expressions:

a*        This expression indicates any occurrence of the letter "a", which includes none.

a+        This expression indicates that at least one occurrence of the letter "a" must be present.

ab?a     This expression matches "aa" or "aba".

_100_    This expression means via AS100.

_100$    This expression indicates an origin of AS100.

^100 .*    This expression indicates transmission from AS100.

^$        This expression indicates origination from this AS.

Prefix-List Filters 

1. Multiple Permit/Deny Rules can be created in sequence and modified as needed.

 

2. (OPTIONAL) GE/LE Fields allow you filter prefixes by subnet size.

These fields allow you specify the desired subnet size filters for inclusion or exclusion.

• GE (Minimum Prefix Length)
  • "Match only routes with a prefix length greater than or equal to this value. Must be more specific than the base prefix."
  • example: GE = 30,  will include prefixes larger or equal to /30 (so /30, /24, /16 etc)
• LE (Maximum Prefix Length)
  • "Match only routes with a prefix length less than or equal to this value. Must be more specific than the base prefix."
  • example: LE = 16, will include prefixes that are /16 or smaller (ex /18, /24, /32)
• Combining both GE and LE
  • example: GE = 24, LE = 20, will include only prefixes between /20 and /24 (ex /24, /22) but would exclude /30 and /16 routes.