Client stops working (Event { type: 'error', message: undefined })?

[DanijelDomazet]

My smee client is started as a system process, and works OK for some time, a few hours, and then it seems like the process is "ready", but notifications aren't received.

% > sudo systemctl status smee
smee.service - smee.io webhook delivery from GitHub
   Loaded: loaded (/etc/systemd/system/smee.service; enabled; vendor preset: enabled)
   Active: active (running) since Fri 2021-09-03 01:27:01 CEST; 13h ago
 Main PID: 9587 (node)
    Tasks: 11 (limit: 4915)
   Memory: 14.8M
   CGroup: /system.slice/smee.service
           └─9587 node /usr/bin/smee -u https://smee.io/1234abcde1234abcde --path /jenkins/github-webhook/ --port 8080

Sep 03 01:27:01 systemd[1]: Started smee.io webhook delivery from GitHub.
Sep 03 01:27:01 smee[9587]: Forwarding https://smee.io/1234abcde1234abcde to http://127.0.0.1:8080/jenkins/github-webhook/
Sep 03 01:27:04 smee[9587]: Connected https://smee.io/1234abcde1234abcde
Sep 03 03:09:02 smee[9587]: Event { type: 'error', message: undefined }
Sep 03 03:09:03 smee[9587]: Connected https://smee.io/1234abcde1234abcde

Could it be that Sep 03 03:09:02 smee[9587]: Event { type: 'error', message: undefined } is the problem?

After I restart the service with sudo /etc/init.d/smee restart, it starts working again, and the status is very similar:

Sep 03 14:31:19 systemd[1]: Started smee.io webhook delivery from GitHub.
Sep 03 14:31:20 smee[9622]: Forwarding https://smee.io/1234abcde1234abcde to http://127.0.0.1:8080/jenkins/github-webhook/
Sep 03 14:31:22 smee[9622]: Connected https://smee.io/1234abcde1234abcde
Sep 03 14:31:53 smee[9622]: POST http://127.0.0.1:8080/jenkins/github-webhook/ - 200

It is just that there is no Event { type: 'error', message: undefined } any more.

Thanks.

[DanijelDomazet]

Please note that this repeatedly happens: few hours after smee restart, the event Event { type: 'error', message: undefined } happens, and the client stops working. What could confuse it?

[DanijelDomazet]

Can I somehow configure the client to log more info?

[gr2m]

I don't think we have any additional debug logs that could be enabled. I'd recommend to clone the repository to your machine, run

npm run build

and

npm link .

then the smee should point to the local repository. You can add more console.log() statements directly into the code, just remember to run npm run build to transpile the changes from TypeScript to JavaScript, otherwise they will get ignored. Or edit the transpiled code directly

[DanijelDomazet]

Here is the smee.service file:

[Unit]
Description=smee.io webhook delivery from GitHub
After=network.target
StartLimitIntervalSec=0

[Service]
Type=simple
Restart=always
RestartSec=1
User=jenkins
ExecStart=/usr/bin/smee -u https://smee.io/1234abcde1234abcde --path /jenkins/github-webhook/ --port 8080

[Install]
WantedBy=multi-user.target

Could something be changed here to help resolve the problem?

[santiagob]

This might be related to issue #179.

[tcbyrd]

This might be related to issue #179.

Agreed. In general because of how smee.io is hosted, the server might restart and drop the connection. We intended smee-client to be used more as a development tool, but I think it's really interesting that people are setting it up as a systemd service. I'll follow up in #179 because I think that comment is a solid path forward.