iBGP session between pe1 and pe2 shut down¶
Case 20 of the Network Change Verdict Corpus. Category: fat-finger. Expected verdict: unsafe (an experienced operator would reject it).
Why it is expected unsafe¶
The two halves of the network stop exchanging routes.
The config change¶
Against the FRR configs of the base lab:
--- a/pe1.conf
+++ b/pe1.conf
@@ -60,3 +60,6 @@
neighbor 10.0.3.2 send-community
exit-address-family
!
+router bgp 65000
+ neighbor 10.0.3.2 shutdown
+!
What each strategy said¶
The expected verdict was committed to git before any run. All three strategies tested the same running lab (method).
| Strategy | Verdict | Result | Checks that failed |
|---|---|---|---|
| Ping only | unsafe | ✅ | customer reaches peer |
| All BGP sessions up | unsafe | ✅ | pe1-pe2 up |
| Netdryrun intent checks | unsafe | ✅ | I1 pe1-pe2 iBGP up, I2 peer has customer route, I2 peer has our aggregate, I3 customer has peer route, I3 customer reaches peer, I5 peer receives exactly our 2 prefixes |
What Netdryrun flagged¶
These checks from the intent spec failed:
- I1 pe1-pe2 iBGP up: the BGP session from pe1 to 10.0.3.2 is established
- I2 peer has customer route: 10.100.0.0/24 must be in peer1's routing table as a bgp route
- I2 peer has our aggregate: 192.0.2.0/24 must be in peer1's routing table as a bgp route
- I3 customer has peer route: 198.51.100.0/24 must be in ce1's routing table as a bgp route
- I3 customer reaches peer: ce1 can ping 198.51.100.1 from 10.100.0.1
- I5 peer receives exactly our 2 prefixes: peer1 receives exactly 2 prefixes from 10.0.4.1
Reproduce it¶
On Linux or WSL2 with Docker and containerlab, from a checkout of the repository:
Case file: corpus/cases/20-ibgp-shutdown.yml. Published result: netdryrun 0.2.0, containerlab 0.81.1, FRR quay.io/frrouting/frr:10.4.1, commit ec23379, run on 2026-10-26.