fix(otbr): drop hostNetwork — wrong abstraction for this pod
OTBR's only outbound need is the spinel TCP connection to slzb-thread.renner.uno:6638. HA and matter-server reach OTBR's REST API via the ClusterIP Service. None of that requires the host's network namespace. The wiki's hostnetwork-app-workload-antipattern applies here. If border-router coordination on the IoT VLAN ever becomes a real requirement, the right answer is a Cilium-L2 LoadBalancer pinned IP (same shape as home-assistant-iot / mosquitto-iot), not hostNetwork. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
@@ -16,11 +16,12 @@ spec:
|
||||
labels:
|
||||
app.kubernetes.io/name: otbr
|
||||
spec:
|
||||
# OTBR needs a real network namespace it can attach the wpan0 Thread
|
||||
# interface to + IPv6 forwarding. hostNetwork on edge keeps it on the
|
||||
# same L2 as the network-attached SLZB Thread RCP.
|
||||
hostNetwork: true
|
||||
dnsPolicy: ClusterFirstWithHostNet
|
||||
# Regular pod network. The only outbound this pod needs is the
|
||||
# spinel TCP connection to slzb-thread.renner.uno:6638; everything
|
||||
# else (HA, matter-server) talks to OTBR via the ClusterIP
|
||||
# Service. If border-router coordination on the IoT VLAN is ever
|
||||
# needed it gets its own Cilium-L2 LoadBalancer IP (analogous to
|
||||
# home-assistant-iot / mosquitto-iot), not hostNetwork.
|
||||
nodeSelector:
|
||||
topology.kubernetes.io/zone: edge
|
||||
tolerations:
|
||||
|
||||
Reference in New Issue
Block a user