This topic is a chance to discuss more about about the most widely deployed Routing Protocols in today’s Enterprise Networks, such as EIGRP, OSPF and BGP. The use of dynamic Routing Protocols is necessary in order to keep the network operational, provide scalability and reduce complexity. Routing Protocols are easy to set up, however, inner workings and predicting the network reconverge might be a challenge specially in failover scenarios.
To participate in this event, please use the 
button below to ask your questions
Ask questions from Monday, August 8 to August 19th, 2016
Featured Expert
Hector Gustavo Serrano is a Customer Support Engineer in the Global Cisco Technical Assistance Center. He is part of the Routing Protocols team, specialized in troubleshooting Routing issues, additionally he educates about best practices of Routing Protocols deployed on customer’s networks. For the last 6 six years he has worked with Cisco platforms for Enterprise Networks, specially focused on US customers. Hector Gustavo also holds a CCIE in Routing & Switching (49102) and a Bachelor’s degree in Telecommunication engineering.
Hector Gustavo might not be able to answer each question due to the volume expected during this event. Remember that you can continue the conversation on the Contact Center Community
Find other https://supportforums.cisco.com/expert-corner/knowledge-sharing.
**Ratings Encourage Participation! **
Please be sure to rate the Answers to Questions
The information can indeed be found in the Summary section of Chapter 10 in the book:
Moving forward, as per RFC4271:
Adj-RIB-In
The Adj-RIBs-In contains unprocessed routing information that has
been advertised to the local BGP speaker by its peers.
As you mention, the behaviour differs depending on the BGP command:
- neighbor Neighbor-IP soft-reconfiguration inbound
exists on the configuration or not.
Let's see both scenarios:
1) The command is missing.
There are 2 commands that can help us to see what networks our Router is receiving from the BGP peer.
- show ip bgp neighbor Neighbor-IP routes
- show ip bgp neighbor Neighbor-IP received-routes
The 1st command will tell you the routes that are received from the specified peer AND are passing the filter inspection that may be configured inbound for that peer in the BGP section (like a Route-Map, Prefix-list, etc.)
Being that said, the command will only show the unfiltered inbound routes.
The 2nd command will not work without the soft-reconfiguration inbound command in place, as the Adj-RIB-in data structure would be missing (It is gone after the processing of creating the NLRIs in the BGP Loc-RIB).
Router2#show ip bgp neighbors 10.1.2.1 received-routes
% Inbound soft reconfiguration not enabled on 10.1.2.1
Router2#
Furthermore, the command 'clear ip bgp 10.1.2.1 soft in' will trigger a Route-REFRESH Message (compliant with RFC2918) that will make the peer to re-send a BGP UPDATE message destined to us and containing the networks (aka NLRIs) it should be advertising.
2) Using the 'soft-reconfiguration inbound' command.
Once the command is in place, the Adj-RIB-in data structure persists.
It is important to remark that the command 'show ip bgp neighbor Neighbor-IP received-routes' now shows an output.
For instance:
R2#show ip bgp neighbors 10.1.2.1 received-routes
BGP table version is 4, local router ID is 2.2.2.2
Status codes: s suppressed, d damped, h history, * valid, > best, i - internal,
r RIB-failure, S Stale, m multipath, b backup-path, f RT-Filter,
x best-external, a additional-path, c RIB-compressed,
Origin codes: i - IGP, e - EGP, ? - incomplete
RPKI validation codes: V valid, I invalid, N Not found
Network Next Hop Metric LocPrf Weight Path
*> 1.1.1.1/32 10.1.2.1 0 0 1 i
* 192.168.1.0 10.1.2.1 0 0 1 i
Total number of prefixes 2
R2#
As the RFC4271 mentions, this is "unprocessed" routing information. These are the NLRIs advertised by the peer, filtered or not on its way to the BGP Loc-RIB.
Keeping the Adj-RIB-in data structure means more RAM will be required of course.
Upon a new execution of the 'clear ip bgp 10.1.2.1 soft in' command (let's say to commit the configuration of a new filter in the BGP section) the Router will now look at the Adj-RIB-in for the routes, there is no need now to request the information to the BGP peer since we keep that unfiltered information on our Adj-RIB-in.
I hope this information has been useful for you.
Best Regards.
The use of Route tagging and filters are suggested to avoid sub-optimal paths and/or Routing loops in scenarios with Multiple 'mutual' Redistribution points.
Let's take a quick look at the below example:
- R3 is receiving the 192.168.1.0/24 as EIGRP External (AD 170).
- The route is forwarded to R1 and R2 via EIGRP.
- R1 and R2 are mutual redistributing EIGRP & OSPF, so now the 192.168.1.0/24 exists in the OSPF domain.
- One of the Routers will prefer the 192.168.1.0/24 via OSPF due to the fact that OSPF AD is 110 while EIGRP External is AD 170 (knowing which Router will have the OSPF entry might be unpredictable as it would be a 'race condition').
# Sub-optimal path example.
R1#show ip route
<Snippet>
D EX 192.168.1.0/24 [170/3072] via 10.1.23.3 (to R3), 00:01:33, GigabitEthernet0/0
R1#
R2#show ip route
<Snippet>
O E2 192.168.1.0/24 [110/20] via 10.1.2.1 (to R1), 00:01:34, GigabitEthernet0/1
R2#
Another possibility is that, for some reason, R1 or R2 calculates the path towards the 192.168.1.0/24 network via one of the Routers doing the redistribution and not via R3.
The outcome would be similar to:
# Routing loop example:
R1#show ip route
<Snippet>
D EX 192.168.1.0/24 [170/28416] via 10.1.23.2 (to R2), 00:01:33, GigabitEthernet0/0
R1#
R2#show ip route
<Snippet>
O E2 192.168.1.0/24 [110/20] via 10.1.2.1 (to R1), 00:01:34, GigabitEthernet0/0
R2#
R1#trace 192.168.1.0
Type escape sequence to abort.
Tracing the route to 192.168.1.0
VRF info: (vrf in name/id, vrf out name/id)
1 10.1.23.2 40 msec 48 msec 44 msec
2 10.1.2.1 32 msec 52 msec 52 msec
3 10.1.23.2 88 msec 88 msec 92 msec
4 10.1.2.1 92 msec 44 msec 120 msec
5 10.1.23.2 176 msec 148 msec 180 msec
6 10.1.2.1 160 msec 200 msec 176 msec
7 10.1.23.2 208 msec 244 msec 240 msec
8 10.1.2.1 236 msec 240 msec 240 msec
9 10.1.23.2 308 msec 160 msec 52 msec
10 10.1.2.1 64 msec 56 msec 56 msec
11 10.1.23.2 68 msec 72 msec 68 msec
12 10.1.2.1 72 msec 72 msec 72 msec
13 10.1.23.2 80 msec 80 msec 84 msec
14 10.1.2.1 112 msec 156 msec 92 msec
15 10.1.23.2 92 msec 92 msec 96 msec
16 10.1.2.1 92 msec 104 msec 96 msec
17 10.1.23.2 104 msec 100 msec 156 msec
18 10.1.2.1 108 msec 124 msec 112 msec
19 10.1.23.2 140 msec 116 msec 120 msec
20 10.1.2.1 120 msec 120 msec 116 msec
21 10.1.23.2 132 msec 132 msec 128 msec
22 10.1.2.1 132 msec 128 msec 128 msec
23 10.1.23.2 140 msec 140 msec 156 msec
24 10.1.2.1 260 msec 152 msec 140 msec
25 10.1.23.2 256 msec 148 msec 152 msec
26 10.1.2.1 176 msec 152 msec 156 msec
27 10.1.23.2 188 msec 196 msec 168 msec
28 10.1.2.1 224 msec 164 msec 292 msec
29 10.1.23.2 348 msec 240 msec 308 msec
30 10.1.2.1 296 msec 212 msec 212 msec
R1#
The solution would be to use Route-Tagging or Filters (i.e. Distribute-lists).
A Router/L3 Switch chooses the Routing decision based on:
- Prefix length.
- Administrative Distance
- Metrics
Lets get into more details:
1. Prefix length. The longest subnet mask is preferred.
For instance:
- 192.168.10.0/24 entry will be preferred over
- 192.168.0.0/16 entry. Due to the prefix length.
2. Administrative Distance (AD). If the Router learns about a destination from more than one routing protocols, the AD is compared and the preference is given to the route with the lower AD.
For instance, with default AD values, EIGRP (internal) has precedence over OSPF which has precedence over EIGRP (External)
EIGRP (internal) 90
OSPF 110
EIGRP (external) 170
3. Metrics. Once both previously mentioned conditions are tied and there are multiple paths to the same destination from a single routing protocol, then the multiple paths would have the same administrative distance and the best path is selected based on the metrics.
The path with the lower metric is elected. If there are multiple pats with equal lowest metrics, those can be elected for ECMP (Equal Cost Multi-Path) which is Load Balancing.
- Route Selection in Cisco Routers
I hope this information has been useful for you.
Best Regards.
