求助：nginx日志报错no live upstreams while connecting to upstream

[北良辰-攻城狮]

我的服务配置是这样的：
upstream app_emp_server{
ip_hash;
server 172.16.70.23:10250;
server 172.16.70.24:10250;
server 172.16.70.25:10250;
}
server {
listen 5250;
server_name localhost;
location ~ / {
proxy_pass http://app_emp_server;
proxy_read_timeout 180; #连接成功后，后端服务器响应时间(代理接收超时)
}
}
环境是 前端服务器请求nginx代理5250端口，nginx收到请求发送后端应用服务器
每天错误日志里面有十几条这样的：
2017/06/12 23:44:12 [error] 189890#0: *20445565 no live upstreams while connecting to upstream, client: 172.16.70.3, server: localhost, request: "POST / HTTP/1.1", upstream: "http://app_emp_server/", host: "172.16.70.25:5250"
2017/06/13 10:03:11 [error] 189888#0: *20525784 no live upstreams while connecting to upstream, client: 172.16.70.3, server: localhost, request: "POST / HTTP/1.1", upstream: "http://app_emp_server/", host: "172.16.70.25:5250"
2017/06/13 10:13:53 [error] 189889#0: *20534209 no live upstreams while connecting to upstream, client: 172.16.70.3, server: localhost, request: "POST / HTTP/1.1", upstream: "http://app_emp_server/", host: "172.16.70.25:5250"
百度了好久 一些文档说报 “no live upstreams while connecting to upstream” 意思是upstream下的server全都挂了
这样解释不应该啊
希望大神给个处理方法！！！

[variationspider]

请问楼主的问题解决没，我的也是报这个错误，导致502，但是upstream下的server全部正常

[cqyy725]

你做负载均衡的情况下，是要做健康检查的。172.16.70.25:5250 这台机器有问题了。ip_hash 固定到了这台机器，所以502