[GH-ISSUE #446] Configuration not reload on systemctl restart. #342

Closed
opened 2026-05-07 00:23:15 +02:00 by BreizhHardware · 6 comments

Originally created by @Microwave-WYB on GitHub (Oct 21, 2022).
Original GitHub issue: https://github.com/binwiederhier/ntfy/issues/446

When received a link to an attachment, the port number is missing and must be manually added.

For example the topic is at "example.com:8000/ntfy"

The file link will appear to be "example.com/files/hash.png"

Originally created by @Microwave-WYB on GitHub (Oct 21, 2022). Original GitHub issue: https://github.com/binwiederhier/ntfy/issues/446 When received a link to an attachment, the port number is missing and must be manually added. For example the topic is at "example.com:8000/ntfy" The file link will appear to be "example.com/files/hash.png"
BreizhHardware 2026-05-07 00:23:15 +02:00
  • closed this issue
  • added the
    question
    label
Author
Owner

@binwiederhier commented on GitHub (Oct 21, 2022):

Can you be more specific where you see this? Android app, web app?

The generated attachment URL is based on the base-url configured in the server.yml file. If that does not have the correct port, then your generated attachment URL will be incorrect.

<!-- gh-comment-id:1287039376 --> @binwiederhier commented on GitHub (Oct 21, 2022): Can you be more specific where you see this? Android app, web app? The generated attachment URL is based on the `base-url` configured in the `server.yml` file. If that does not have the correct port, then your generated attachment URL will be incorrect.
Author
Owner

@Microwave-WYB commented on GitHub (Oct 21, 2022):

Can you be more specific where you see this? Android app, web app?

The generated attachment URL is based on the base-url configured in the server.yml file. If that does not have the correct port, then your generated attachment URL will be incorrect.

I have the port number in the base url. This is not a problem about android or web app, it's the back-end sending the wrong url. I restarted the server to make sure the configuration is reloaded.

<!-- gh-comment-id:1287415900 --> @Microwave-WYB commented on GitHub (Oct 21, 2022): > Can you be more specific where you see this? Android app, web app? > > The generated attachment URL is based on the `base-url` configured in the `server.yml` file. If that does not have the correct port, then your generated attachment URL will be incorrect. I have the port number in the base url. This is not a problem about android or web app, it's the back-end sending the wrong url. I restarted the server to make sure the configuration is reloaded.
Author
Owner

@Microwave-WYB commented on GitHub (Oct 21, 2022):

Screenshot_20221021-135448
From here it seems like the base url should be used, but I've tested multiple times that this is not including the port number

<!-- gh-comment-id:1287428667 --> @Microwave-WYB commented on GitHub (Oct 21, 2022): ![Screenshot_20221021-135448](https://user-images.githubusercontent.com/68882116/197287090-f31e023f-a126-415f-8c03-7caf9e950d4b.png) From here it seems like the base url should be used, but I've tested multiple times that this is not including the port number
Author
Owner

@Microwave-WYB commented on GitHub (Oct 21, 2022):

I found the problem here. systemctl restart ntfy does not reload configuration. I had to do systemctl stop ntfy; systemctl start ntfy

<!-- gh-comment-id:1287444604 --> @Microwave-WYB commented on GitHub (Oct 21, 2022): I found the problem here. `systemctl restart ntfy` does not reload configuration. I had to do `systemctl stop ntfy; systemctl start ntfy`
Author
Owner

@binwiederhier commented on GitHub (Oct 21, 2022):

I found the problem here. systemctl restart ntfy does not reload configuration. I had to do systemctl stop ntfy; systemctl start ntfy

This is the second instance of this that I'm hearing about. I use systemctl restart ntfy when I deploy a new version and it works nicely. What system are you on? What does journalctl say when you issue a restart? Are you using the .deb/.rpm or something custom?

<!-- gh-comment-id:1287460761 --> @binwiederhier commented on GitHub (Oct 21, 2022): > I found the problem here. `systemctl restart ntfy` does not reload configuration. I had to do `systemctl stop ntfy; systemctl start ntfy` This is the second instance of this that I'm hearing about. I use `systemctl restart ntfy` when I deploy a new version and it works nicely. What system are you on? What does journalctl say when you issue a restart? Are you using the .deb/.rpm or something custom?
Author
Owner

@binwiederhier commented on GitHub (Nov 19, 2022):

Closed due to inactivity

<!-- gh-comment-id:1320962475 --> @binwiederhier commented on GitHub (Nov 19, 2022): Closed due to inactivity
Sign in to join this conversation.
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
starred/ntfy#342
No description provided.