Increased DELETE API Latency on GKE 1.31.5 When Using kubevpn connect

[felipeneuwald]

When running kubevpn connect on a GKE cluster version 1.31.5-gke.1169000, DELETE API calls experience increased latency.

Environment:

$ kubevpn version
KubeVPN: CLI
    Version: v2.3.13
    Daemon: v2.3.13
    Image: docker.io/naison/kubevpn:v2.3.13
    Branch: master
    Git commit: brew
    Built time: 2025-02-23 13:42:21
    Built OS/Arch: darwin/arm64
    Built Go version: go1.24.0

Reproduction Steps:
1. Connect to a GKE cluster (1.31.5-gke.1169000).
2. Run kubevpn connect.
3. Execute DELETE API operations (e.g., from a Kubernetes Operator or via kubectl).
4. Observe significantly increased latency in DELETE operations.
5. Run kubevpn uninstall and note that DELETE operations become significantly faster.

This behavior has been observed both within a Kubernetes Operator running inside the cluster and when executing kubectl delete commands.

[wencaiwulue]

@felipeneuwald

1. Can you provide the log file ~/.kubevpn/daemon/daemon.log to help diagnose problem?
2. Look like api-server receive more request. maybe you can increase deployment api-server resource (cpu/memory)?