用这个配置，提示：
Error: unknown variant aws_lambda, expected one of direct, fronted, direct_tcp, race, priority_race
我用的是 cargo 上最新的 v0.2.72 的 geph5-client。
注掉 aws_lambda 那一段之后，提示：
2025-07-01T13:23:00.856633Z DEBUG refresh_conn_token: geph5_client::broker::fronted_http: calling broker through http method="get_user_info" url="https://www.cdn77.com/" host=Some("1826209743.rsc.cdn77.org")
2025-07-01T13:23:01.356958Z DEBUG refresh_conn_token: geph5_client::broker::fronted_http: calling broker through http method="get_user_info" url="https://www.vuejs.org/" host=Some("svitania-naidallszei-3.netlify.app")
2025-07-01T13:23:01.739091Z DEBUG refresh_conn_token: geph5_client::broker::fronted_http: response received through http method="get_user_info" url="https://www.vuejs.org/" host=Some("svitania-naidallszei-3.netlify.app") resp_len=50 elapsed=382.080281ms
2025-07-01T13:23:10.858146Z WARN geph5_client::auth: failed to refresh conn token err=transport-level error: cannot send request to front
Caused by:
0: error sending request for url (https://www.cdn77.com/)
1: operation timed out
反复了几分钟之后连上了。因为 4.99 版和旧版的 geph5 也是时通时断，不确定这个是不是好一些。
