我的worker是部署在內(nèi)網(wǎng)用的,內(nèi)網(wǎng)有3臺(tái)服務(wù)器,11.2是應(yīng)用服務(wù),thinkphp框架,11.3是thinkphp+worker,負(fù)責(zé)推送及幾個(gè)簡(jiǎn)單的頁(yè)面;
11.3用的centos8+nginx來(lái)跑tp和wk的,自己做的證書,就是為了跑ssl。
跑了1個(gè)多月一切正常。前幾天機(jī)房ups故障,服務(wù)器被直接斷電,故障處理后開機(jī)就出現(xiàn)問題了!登錄時(shí)大幾率出現(xiàn)ERR_CONNECTION_REFUSED,還會(huì)有客戶端掉線,也就是說(shuō),會(huì)突然掉20幾個(gè)客戶端。然后客戶端除非刷新,然后在碰運(yùn)氣不出現(xiàn)ERR_CONNECTION_REFUSED錯(cuò)誤,才能重新連接上來(lái)。
客戶端打頂最高不會(huì)超過400,同時(shí)在線也不會(huì)高于300。
客戶端連接ws = new WebSocket("wss://10.xxx.xxx.xxx/wss");
在11.3的防火墻開了1236/2348/2000/2001~2009/80/443
nginx.conf配置文件
worker的狀態(tài)
ERR_CONNECTION_REFUSED,說(shuō)明連接被拒絕了,感覺是nginx層的問題。
看看是不是域名配置了多個(gè)ip,個(gè)別ip沒有啟動(dòng)服務(wù),客戶端連到有問題的ip導(dǎo)致連接被拒絕。
還有一種可能是nginx掛了。
另外看到status里gateway和businessworker進(jìn)程有業(yè)務(wù)異常,突然20幾個(gè)客戶端掉線是因?yàn)間ateway異常進(jìn)程退出導(dǎo)致的。
謝謝你熱心的分析,我們公司內(nèi)網(wǎng)沒有域名,簡(jiǎn)單粗暴的直接上IP,所有網(wǎng)站都是IP訪問。
如果是nginx掛了,應(yīng)該不會(huì)存在刷新一下又會(huì)連上的情況??!所以這幾天都找不出原因
剛看了一下nginx的日志:
2021/09/26 09:10:12 [error] 38812#0: 3 recv() failed (104: Connection reset by peer) while reading response header from upstream, client: 10.xxx.xxx.146, server: _, request: "GET /wss HTTP/1.1", upstream: "http://127.0.0.1:2348/wss", host: "10.xxx.xxx.3"
2021/09/26 09:10:12 [error] 38812#0: 3 open() "/var/www/wss/public/50x.html" failed (2: No such file or directory), client: 10.xxx.xxx.146, server: _, request: "GET /wss HTTP/1.1", upstream: "http://127.0.0.1:2348/wss", host: "10.xxx.xxx.3"
2021/09/26 09:10:12 [error] 38812#0: *3 recv() failed (104: Connection reset by peer) while reading response header from upstream, client: 10.xxx.xxx.146, server: _, request: "GET /wss HTTP/1.1", upstream: "http://127.0.0.1:2348/wss", host: "10.xxx.xxx.3"
2021/09/26 09:10:12 [error] 38812#0: *3 open() "/var/www/wss/public/50x.html" failed (2: No such file or directory), client: 10.xxx.xxx.146, server: _, request: "GET /wss HTTP/1.1", upstream: "http://127.0.0.1:2348/wss", host: "10.xxx.xxx.3"