How to fix BGP peer down in Kubernetes

Verify node readiness and check BIRD BGP status:

kubectl get nodes
kubectl logs -n calico-system ds/calico-node

Look for "BIRD" related errors indicating BGP connectivity issues.

List configured BGP peers to verify correct configuration:

calicoctl get bgppeers -o yaml
calicoctl get bgpconfig -o yaml

Verify peer IPs, AS numbers, and node selectors are correct.

From the Calico node, test reachability to peer on port 179:

kubectl exec -it -n calico-system ds/calico-node -- \
  nc -zv <peer-ip> 179

If connection fails, check firewall rules.

Ensure network policies and host firewalls allow BGP:

# Check if port 179 is open
netstat -tlnp | grep 179

# Verify firewall rules (varies by OS)
iptables -L -n | grep 179
sudo ufw allow 179

BGP keepalive requires synchronized clocks:

date
ntpstat
chronyc tracking

If clocks are out of sync, synchronize using NTP or chrony.

Ensure the BGP peer IP is reachable from the node:

ping -c 3 <peer-ip>
traceroute <peer-ip>

If unreachable, verify network routes and connectivity.

Verify BGP configuration is not conflicting:

calicoctl get node <node-name> -o yaml | grep -A 5 asNumber

Ensure each node has unique router ID and correct AS numbers.

If configuration is correct but BGP still down, restart node:

kubectl rollout restart daemonset/calico-node -n calico-system

Monitor logs during restart to confirm BGP comes up.