Webhook: PM2 problem

[koniecznytomek]

I'm using PM2 to automatically start the webhook on my VPS server (UBUNTU 18) and everything works fine. The problem is when I restart the server. When I run the process again, i have an error and I must to execute command "pkill webhook" and "pm2 restart process_name" and then its work again.

0|start | [webhook] 2020/05/24 19:31:09 listen tcp my_ip:9000: bind: address already in use
0|start | [webhook] 2020/05/24 19:31:09 version 2.5.0 starting
0|start | [webhook] 2020/05/24 19:31:09 setting up os signal watcher
0|start | [webhook] 2020/05/24 19:31:09 attempting to load hooks from hooks.json
0|start | [webhook] 2020/05/24 19:31:09 found 1 hook(s) in file
0|start | [webhook] 2020/05/24 19:31:09 loaded: build-webhook
0|start | [webhook] 2020/05/24 19:31:09 serving hooks on http://my_ip:9000/hooks/{id}

Is there any way to kill the previous address after server reboot?
Sorry I'm a layman in the subject :(

Tom.

[moorereason]

I'm not familiar with PM2. Are you saying that pm2 restart webhook doesn't work?

[koniecznytomek]

PM2 is only for maintenance hooks with command "webhook -hooks hooks.json -verbose", and it's working ok. Problem is when i try to restart or reboot my linux server. After reboot, when i try to run same process witch PM2 or only just witch command above i have error:

[webhook] 2020/05/25 08:42:44 serving hooks on http://0.0.0.0:9000/hooks/{id}
[webhook] 2020/05/25 08:42:44 listen tcp 0.0.0.0:9000: bind: address already in use

Every time, after reboot, i must kill "old" process before running webhook with PM2 or just with command webhook -hooks hooks.json -verbose. I don't know why.

[moorereason]

It sounds like PM2 is starting the webhook service upon reboot and then you're trying to start it again manually. In my mind, it's all working as designed. What am I missing?

[koniecznytomek]

Thanks for your patience, Cameron :). Let's forget about PM2 for a while. The same problem occurs without using it. Maybe step by step.

1. I start the webhook from the webhook -hooks hooks.json -verbose command. Everything works, the hook is available under the configured endpoint.

2. I am restarting/reboot my ubuntu server.

3. I start the webhook again as in the first point, and this appears in the terminal:

[webhook] 2020/05/25 08:42:44 serving hooks on http://0.0.0.0:9000/hooks/{id}
[webhook] 2020/05/25 08:42:44 listen tcp 0.0.0.0:9000: bind: address already in use
...

1. When i check webhook endpoint link in browser i have "Hook not found."

[koniecznytomek]

This is exactly the same case as in https://github.com/adnanh/webhook/issues/246 issue, but for me it happens after server reset. I have to do it after reboot (kill this connection) every time to be able to run webhook again.

[moorereason]

After rebooting the server (step 2 above), webhook is already running. What is starting webhook (before step 3)? Whatever it is, share with us how that's configured.

[koniecznytomek]

I have 3 files:

hooks.json with:

[
  {
    "id": "build-webhook",
    "execute-command": "/var/www/webhook/build.sh",
    "command-working-directory": "/var/www/mydomain.com/public_html/sandbox"
  }
]

build.sh with:

#!/bin/bash
gatsby build

start.sh with:

webhook -hooks hooks.json -ip "my server ip" -verbose

I used PM2 as process manager for the webhook to work without interruption. (PM2 start start.sh --name webhook)

and everything working fine. I have endpoint link with working webhook that realised gatsby build command everytime when i use this endpoint.

After reboot, webhook endpoint not working (Hook not found). As i wrote in above comments, even if I restart the PM2 process again, i must close/kill old network connection with 9000 port because i have the same error like in https://github.com/adnanh/webhook/issues/246

I want the webhook process to start automatically after a reset without this error and without manual killing old connection. I`m sorry, I don't know how to describe this problem more simply :( Thanks!

[moorereason]

It sounds like webhook is being started by systemd before PM2 starts. See what systemctl status webhook shows.

[koniecznytomek]

After reboot:

● webhook.service - Small server for creating HTTP endpoints (hooks)
   Loaded: loaded (/lib/systemd/system/webhook.service; enabled; vendor preset: enabled)
   Active: active (running) since Wed 2020-05-27 20:15:14 CEST; 2min 16s ago
     Docs: https://github.com/adnanh/webhook/
 Main PID: 784 (webhook)
    Tasks: 4 (limit: 2361)
   CGroup: /system.slice/webhook.service
           └─784 /usr/bin/webhook -nopanic -hooks /etc/webhook.conf

Warning: Journal has been rotated since unit was started. Log output is incomplete or unavailable.

seems that webhook is being started but endpoint link not working, i have "Hook not found."

After this, i put in terminal sudo pkill webhook and i run PM2 process again. I have:

● webhook.service - Small server for creating HTTP endpoints (hooks)
   Loaded: loaded (/lib/systemd/system/webhook.service; enabled; vendor preset: enabled)
   Active: inactive (dead) since Wed 2020-05-27 20:29:02 CEST; 1min 8s ago
     Docs: https://github.com/adnanh/webhook/
  Process: 805 ExecStart=/usr/bin/webhook -nopanic -hooks /etc/webhook.conf (code=killed, signal=TERM)
 Main PID: 805 (code=killed, signal=TERM)

... inactive(dead), but now webhook working fine :) its crazy.

[moorereason]

I think the Ubuntu webhook package uses /etc/webhook.conf as the default hooks configuration file. That's why you're hook isn't found. (We don't manage the Ubuntu packages, BTW)

Try disabling the systemd webhook service:

systemctl disable webhook

[koniecznytomek]

Yeah! Now its work exactly what i want :) Thank You so much for your help and time, Cameron, You are Great :)