cilium lb status - BGP Peers Check Degraded with 2+ BGP Peers

[camrossi]

Bug report

I am testing thew new Cilium LoadBalancer and I got it to work in BGP mode, in my testing I am peering with 2 Upstream Routers so I have 2 BGP peers:

(default) ➜  cilium git:(master) ✗ cilium bgp peers
Node   Local AS   Peer AS   Peer Address   Session State   Uptime   Family         Received   Advertised
t1-1   65101      65001     192.168.1.27   established     6m24s    ipv4/unicast   0          2    
       65101      65001     192.168.1.28   established     6m24s    ipv4/unicast   0          2    

But the cilium lb status command is not happy with this, it expect only 1 Peer per T1 nodes and returns a DEG [2/1] message.

In production I would always have 2 or more BGP peers per T1 Node.

(default) ➜  cilium git:(master) ✗ cilium lb status
=========
Summary
=========

T1 Nodes:   1
T2 Nodes:   2
Services:   1
VIPs:       1

=========
Services
=========

Namespace   Name   VIP          Port    Type         D-Mode   BGP Peers   BGP Routes   T1          HC T1->[T2|B]   T2          HC T2->B    Backendpools   Status
---------   ----   ---          ----    ----         ------   ---------   ----------   --          -------------   --          --------    ------------   ------
default     http   100.64.0.0   40000   HTTP Proxy   T1-T2    DEG [2/1]   OK  [1/1]    OK  [1/1]   OK  [2/2]       OK  [2/2]   OK  [2/2]   OK  [1/1]      ONLINE

General Information

cilium version
cilium-cli: enterprise/cilium-cli/v0.18.4-cee.1 compiled with go1.24.4 on linux/amd64
cilium image (default): v1.17.4
cilium image (stable): v1.17.5
cilium image (running): 1.18.9999-dev

(default) ➜  cilium git:(master) ✗ kubectl version
Client Version: v1.32.0
Kustomize Version: v5.5.0
Server Version: v1.33.2

Platform: Talos

[github-actions]

This issue has been automatically marked as stale because it has not
had recent activity. It will be closed if no further activity occurs.