I decided to test our [[ https://docs.victoriametrics.com/victorialogs/ | VictoriaLogs ]] to inject logs from various sources, one of which is VyOS.
It listens on port 9428
I put in the following config:
```
[edit service monitoring telegraf loki]
tim@ferrari# show
metric-name-label VyOS
port 9428
url http://192.168.0.5/insert
```
But this is what's generated in the telegraf.conf file:
```
### Loki ###
[[outputs.loki]]
## The domain of Loki
domain = "http://192.168.0.5/insert:9428"
metric_name_label = "VyOS"
### End Loki ###
```
The port is being inserted in the incorrect place.
If I delete the port command and try to manually set that URL:
```
[edit service monitoring telegraf loki]
tim@ferrari# show
metric-name-label VyOS
url http://192.168.0.5:9428/insert
```
We still end up with the wrong config as the config generator is still adding the default 1300 to the URL:
```
### Loki ###
[[outputs.loki]]
## The domain of Loki
domain = "http://192.168.0.5:9428/insert:3100"
metric_name_label = "VyOS"
### End Loki ###
```
It seems to me the url command isn't expecting/looking for anything after the hostname.
The only way to make this work is to manually edit the telegraf.conf file and HUP the telegraf process.