Hey there! If you’re running into the "failed to get external IP" error with your LiveKit deployment on Google GKE, no worries—let’s walk through some steps to get it sorted out: 1. RTCM Configuration: Double-check that

use_external_ip: true

is set in your LiveKit config. This helps LiveKit find the external IP using STUN, which is super important in cloud setups. 2. Load Balancer Configuration: Make sure you’re using a GKE-friendly load balancer in your Helm chart. If you’re on a VPC-native GKE cluster, set the load balancer type to

gke-vpc-native

. 3. Network Requirements: Verify that the necessary ports are open—this includes

tcp_port

(default 7881),

udp_port

(default 3478) for TURN, and the RTP/RTCP port range (50000-60000). Also, check that GKE’s firewall rules allow traffic on these ports. 4. STUN Servers: LiveKit relies on STUN servers to find your external IP. Make sure your deployment can reach the default STUN servers, or you can specify your own custom STUN servers if needed. 5. Verify Service Exposure: Take a look at your Kubernetes service setup to ensure LiveKit is exposed properly. You can check the service status with: `kubectl get svc -n <namespace>`If you’re using an Ingress, confirm it’s routing traffic correctly to LiveKit. 6. Check Logs: Still stuck? Peek at the LiveKit logs to get more clues about what’s going on with the external IP lookup:

kubectl logs <livekit-pod-name> -n <namespace>

Give these a try, and you should be back on track in no time. If you hit any bumps or want me to walk through anything else, just let me know—I’m here to help! Sources: Deploying LiveKit | LiveKit Docs | gclb-sample.yaml | Deploy to Kubernetes | LiveKit Docs | pkg/rtcconfig/ip.go | Self-hosting the Ingress Service | LiveKit Docs