Btukfyl

I am experiencing problems reloading the application using the ecosystem.config.js file. When the application is started for the first time, it starts correctly, but when I refresh/reload the application using the ecosystem.config.js file, the application restarts several times causing an error.

My SO is Ubuntu Xenial, PM2 version is 3.2.2 and Node v10.13.0. The application uses the latest version from Express module (4.16.4).
If I reload the application with "pm2 reload app_name", this problem doesn't occur.

The ecosystem.config.js content:

module.exports = {

  apps: [{

     script: "./index.js",

     instances: "max",

     exec_mode: "cluster",

     kill_timeout: "2000",

     env: {

       NODE_ENV: "development",

     },

     env_production: {

       NODE_ENV: "production",

     }

  }]

}

When I run the first time:

$ pm2 reload ecosystem.config.js

[PM2][WARN] Applications index not running, starting...

[PM2] App [index] launched (2 instances)

node@ubuntu:/data/$ pm2 logs

[TAILING] Tailing last 15 lines for [all] processes (change the value with >--lines option)

/home/node/.pm2/pm2.log last 15 lines:

PM2 | 2018-11-23T13:14:30: PM2 log: App [index:0] starting in -cluster >mode-

PM2 | 2018-11-23T13:14:31: PM2 log: App [index:0] online

PM2 | 2018-11-23T13:14:31: PM2 log: App [index:1] starting in -cluster >mode-

PM2 | 2018-11-23T13:14:31: PM2 log: App [index:1] online

When I reload the application by name (ex: pm2 reload app_name), the application contiue runnig, but I see some timeouts to kill process:

PM2 | 2018-11-23T14:01:02: PM2 log: pid=11296 msg=failed to kill - retrying in 100ms
PM2 | 2018-11-23T14:01:02: PM2 log: Process with pid 11289 still alive after 6000ms, sending it SIGKILL now...
PM2 | 2018-11-23T14:01:02: PM2 log: pid=11296 msg=failed to kill - retrying in 100ms
PM2 | 2018-11-23T14:01:02: PM2 log: Process with pid 11296 still alive after 6000ms, sending it SIGKILL now...
PM2 | 2018-11-23T14:01:02: PM2 log: App name:index id:_old_0 disconnected
PM2 | 2018-11-23T14:01:02: PM2 log: App [index:_old_0] exited with code [0] via signal [SIGKILL]
PM2 | 2018-11-23T14:01:02: PM2 log: App name:index id:_old_1 disconnected
PM2 | 2018-11-23T14:01:02: PM2 log: App [index:_old_1] exited with code [0] via signal [SIGKILL]
PM2 | 2018-11-23T14:01:02: PM2 log: pid=11289 msg=process killed
PM2 | 2018-11-23T14:01:02: PM2 log: pid=11296 msg=process killed

But, even though timeouts occur the application is running.

When I execute "pm2 reload ecosystem.config.js", the PM2 restart the application several times and one instance fail:

0|index | at Module.load (internal/modules/cjs/loader.js:598:32)

0|index | at tryModuleLoad (internal/modules/cjs/loader.js:537:12)

0|index | at Function.Module._load (internal/modules/cjs/loader.js:529:3)

0|index | at Object. (/usr/lib/node_modules/pm2/lib/ProcessContainerFork.js:48:21)

0|index | Error: listen EADDRINUSE :::3001

I believe the problem is related to some timeout to properly terminate the http connection of the Express module, but I'm still investigating this.

It has been fixed on the lastest PM2 version, please update:

npm install pm2@latest -g
pm2 update

Make sure you pm2 delete all and then start back your application again, it will work then when doing reload or restart