egg 应用机器凌晨0点 cpu 飙升到100%
发布于 5 年前 作者 xiaozhongliu 4879 次浏览 来自 问答

问题简述: 今天凌晨0点3台应用机器cpu爆炸, 持续了几个小时(之后也不是应用自己恢复正常的). 猜想是不是跟日志有关? 前一天清空了应用日志目录. image.png 机器配置: 4cpu x 3台k8s节点机器. 应用规模: 3台机器跑了大约50个egg k8s pod, 每个应用跑2个pod. 日志写入: 如果一个应用的2个pod落在同一个机器上, 会写相同的日志文件. (单个pod实例里4个egg worker是不是也存在这样的情况?) 启动方式: pod里面用egg-scripts start --daemon运行起来的.

初步排查结果, cpu暴涨的node进程都是egg worker: image.png

1111 ?        Rl     0:13 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":35401}
1566 ?        Rl     0:06 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":39980}
1610 ?        Rl     0:10 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":42591}
1667 ?        Sl     0:07 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":46809}
2222 ?        Rl     0:06 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/agent_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":35401}
2349 ?        Rl     0:04 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":46809}
2529 ?        Rl     0:04 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/agent_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":42591}
9716 ?        Rl     3:28 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":39980}
18599 ?        Rl     5:36 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/agent_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":43367}
18768 ?        Rl     5:31 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":43367}
18781 ?        Rl     5:37 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":43367}
18783 ?        Rl     7:42 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/agent_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":37730}
18871 ?        Rl     7:04 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":37730}
18873 ?        Rl     7:11 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":37730}
18878 ?        Rl     7:08 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":37730}
18888 ?        Rl     6:58 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":37730}
19292 ?        Rl     6:04 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/agent_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":42728}
19324 ?        Rl     5:05 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":42728}
19325 ?        Rl     5:06 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":42728}
19331 ?        Rl     5:04 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":42728}
19344 ?        Rl     5:01 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":42728}
20639 ?        Rl     4:58 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":39980}
21635 ?        Rl     2:50 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":39980}
26413 ?        Rl     5:29 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/agent_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":39980}
31557 ?        Rl     6:56 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/agent_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":45792}
31580 ?        Rl     6:27 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":45792}
31582 ?        Rl     6:24 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":45792}
31584 ?        Rl     6:22 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":45792}
31596 ?        Rl     6:29 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":45792}
32705 ?        Rl     0:18 /usr/local/bin/node --require /usr/local/lib/node_modules/egg-scripts/node_modules/source-map-support/register.js /app/node_modules/egg-cluster/lib/app_worker.js {"framework":"/app/node_modules/egg","baseDir":"/app","workers":4,"plugins":null,"https":false,"title":"*************","clusterPort":35401}
9 回复

这个只能你找一天再次复现下,然后当场用 alinode 抓下 Profile 看看了。 从给出的信息无法分析。

另外,在 docker 里面 egg-scripts start --daemon 的话,不就挂了?不应该加 daemon 后台运行的。

@atian25 感谢回复, 对的没错启动命令是egg-scripts start --title=$PROJECT, 年代久远忘了. 请问下egg如何关闭多进程模式, 本来就起了多个pod, 也开了自动扩容. 不需要内部多进程了.

有个 https://github.com/eggjs/egg/issues/3180 但还在 beta 阶段。 先开着多进程也没啥大问题。

@atian25 好吧暂时1个worker过渡

          env:
            - name: EGG_WORKERS
              value: 1

最好 2 个,一个的话,worker 挂了,但 master 还活着(所以 k8s 那边应该不会重启 pod 吧),那在新 worker 启动前,你的这个服务是挂的。

mark,期待问题分享

蛋疼了吧,哈哈

回到顶部