agent设置kubesphere-service代理本地测试服务,proxy panic

run agen
```bash
./bin/agent --kubeconfig ~/.kube/config --kubesphere-service http://0.0.0.0:8000 --token 13530761f40fc3161ed59316a6b3ed7b6ba44c5cca613f03d37db3fe9d903e87 --name alpha --v 4 --proxy-server http://192.168.1.31:30880
```

`http://0.0.0.0:8000`是本地的一个python文件服务 (`python3 -m http.server`)

集群pod内访问代理的80端口, proxy panic,

日志:

```
http: panic serving 10.233.124.47:41500: runtime error: invalid memory address or nil pointer dereference
goroutine 588 [running]:
net/http.(*conn).serve.func1(0xc000238280)
/usr/local/go/src/net/http/server.go:1795 +0x139
panic(0x13594c0, 0x208c260)
/usr/local/go/src/runtime/panic.go:679 +0x1b2
k8s.io/apimachinery/pkg/util/proxy.normalizeLocation(0xc000017b88, 0x20c5d60)
/go/pkg/mod/k8s.io/apimachinery@v0.18.6/pkg/util/proxy/upgradeaware.go:168 +0x48
k8s.io/apimachinery/pkg/util/proxy.NewUpgradeAwareHandler(...)
/go/pkg/mod/k8s.io/apimachinery@v0.18.6/pkg/util/proxy/upgradeaware.go:178
kubesphere.io/tower/pkg/proxy.(*Server).ServeHTTP(0xc000337b00, 0x16d7f80, 0xc00012a700, 0xc000260c00)
/workspace/pkg/proxy/proxy_server.go:184 +0x28b
net/http.serverHandler.ServeHTTP(0xc0002aa380, 0x16d7f80, 0xc00012a700, 0xc000260c00)
/usr/local/go/src/net/http/server.go:2831 +0xa4
net/http.(*conn).serve(0xc000238280, 0x16dbc40, 0xc0000a6dc0)
/usr/local/go/src/net/http/server.go:1919 +0x875
created by net/http.(*Server).Serve
/usr/local/go/src/net/http/server.go:2957 +0x384

```

Reproduce the proxy panic with the provided agent command and a request from a cluster pod, then inspect pkg/proxy/proxy_server.go around line 184 and the k8s.io/apimachinery proxy call shown in the stack trace. Done means the same request completes without a nil-pointer panic and the regression is covered by an appropriate test.

Written by the indexing model from the issue text.