Cisco IOS BGP (eBGP / iBGP)

Cisco IOS / IOS-XEAdvanced

eBGP and iBGP configuration with route reflectors, communities, prefix-lists, and route-maps for Cisco IOS/IOS-XE.

What is it?

BGP (Border Gateway Protocol) is the routing protocol that runs the internet. It's an Exterior Gateway Protocol (EGP) designed to exchange routing information between different organizations (Autonomous Systems). Every AS on the internet has a unique AS number, and BGP is how they announce which IP prefixes they own. Unlike OSPF which finds the 'shortest' path, BGP uses a complex set of attributes (AS_PATH, LOCAL_PREF, MED, communities) to implement routing policies.

✓ When to use

·Connecting to one or more ISPs (internet peering) — always BGP
·Multi-homed networks that need traffic engineering across multiple ISPs
·Large enterprise networks or data centers that need policy-based routing between different internal domains
·SD-WAN overlay routing where BGP carries VPN prefixes between sites
·Any time you need to exchange routes with an organization you don't control

✗ When NOT to use

·Routing inside your own organization — use OSPF or EIGRP instead (they converge faster)
·Small networks with a single ISP — use a default static route (0.0.0.0/0) pointing to the ISP
·Networks where you don't need routing policy control — BGP complexity is not worth it

How it compares

vs OSPFOSPF is an IGP (Interior Gateway Protocol) for routing inside one organization. BGP is an EGP for routing between organizations. They are complementary, not competitors — large networks run OSPF internally and BGP externally.

vs EIGRPEIGRP is Cisco-proprietary and only suitable for internal routing. BGP is the only option for inter-domain (internet-facing) routing. Not comparable use cases.

vs Static routesStatic routes are simple but don't adapt to failures. BGP detects when a neighbor goes down and withdraws routes automatically. For ISP connectivity with redundancy, BGP is required.

Prerequisites

→Loopback interfaces configured on all routers (used as BGP router IDs)
→IGP (OSPF or EIGRP) must be running between iBGP neighbors for reachability
→AS numbers allocated — public AS from ARIN/RIPE if connecting to internet, private AS (64512-65534) for internal BGP
→For eBGP: physical connectivity to the peer (or single-hop reachability)
→Prefix-lists and route-maps should be prepared before enabling BGP neighbors

Config Generator

Hostname *

Local AS *

Router ID *

Log neighbor changes?

BGP Neighbors

Networks to advertise

Prefix Lists

Route Maps

Redistribute OSPF?

OSPF process ID

Redistribute static?

Default-information originate?

Verification commands

show bgp summary

Show all BGP neighbors and their state

Expected: Neighbors should show 'Established' state and a non-zero 'Up/Down' time. 'Active' or 'Idle' means not connected.

show bgp neighbors <ip>

Show detailed BGP session info for a specific neighbor

Expected: Shows hold time, capabilities, messages sent/received, and the reason for last reset

show ip route bgp

Show routes learned via BGP in the routing table

Expected: Routes marked 'B' (eBGP) or 'b' (iBGP) with the correct next-hop

show bgp ipv4 unicast

Show the BGP table with all paths and best-path selection

Expected: The '>' symbol marks the best path. '* i' means a valid iBGP path not in the routing table.

Debug commands

debug ip bgp <ip> events

Debug BGP session events with a specific neighbor

⚠ Generates significant output. Use 'undebug all' after. Use in maintenance windows only.

show bgp ipv4 unicast neighbors <ip> advertised-routes

Show which routes you are advertising to a neighbor

Common mistakes & fixes

⚡ BGP neighbor stuck in Active state

Cause: TCP port 179 is blocked, wrong peer IP, or wrong remote-as configured

Fix: Verify: 'telnet <peer-ip> 179' from the router. Check ACLs blocking TCP 179. Verify remote-as matches the peer's actual AS number.

⚡ iBGP neighbor established but routes not in routing table

Cause: BGP synchronization rule or next-hop not reachable via IGP

Fix: Add 'no synchronization' (IOS default), and verify the next-hop IP is reachable. Use 'next-hop-self' on route reflectors.

⚡ Routes not being advertised to neighbor

Cause: Route not in the routing table, prefix-list filtering it out, or network statement with wrong mask

Fix: Verify 'show ip route <network>' exists. Check 'show bgp neighbors <ip> advertised-routes' vs 'show bgp neighbors <ip> routes'.

⚡ BGP session flapping frequently

Cause: Hold timer expiry due to high CPU or interface instability

Fix: Check CPU: 'show processes cpu sorted'. Increase hold timers: 'neighbor <ip> timers 10 30'. Check interface stability.