A note for the community
- Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
- If you are interested in working on this issue or have submitted a pull request, please leave a comment
Idea is to have a file watcher inside vector that will automatically reload some file or files on filesystem when they are changed, and make content available to vrl script that is executed in remap transform or filter transform. In config it could look like a filepath and script that is run on content when file is changed to for example parse json, etc. And some key property that defines a identifier through which you can access this data in your transforms. The motivation for this very simple, there is a lot of state on the system that vector might not be aware of that is required to properly enrich collected metrics and logs before sending them to some centralized location. This information is usually easily acceptable on the node, and writing some script that would update a json file if let`s say network interfaces list change is not complicated.
Use Cases
In my case I want to have ability to resolve cilium network interfaces devices to pod ids when I collect network metrics.
Attempted Solutions
Right now vector vrl has option to send http request which in theory could be used to solve this, however I find it unsatisfactory because of the overhead, especially considering it would have to be run on each metric event.
Proposal
Here is incomplete config example to illustrate the idea:
state:
network_interface_to_pod_id:
filepath: /etc/somefile.json
load_script: |
. = parse_json!(.buffer)
transforms:
host_metrics:
type: remap
inputs: ["host_metrics-input"]
state:
ni_to_pod_id: network_interface_to_pod_id
source: |
.tags.pod_id = get(ni_to_pod_id, [.tags.device]) ?? nullReferences
No response
Version
vector 0.54.0 (x86_64-unknown-linux-gnu 2b8b875 2026-03-10 15:47:37.284215410)
A note for the community
Idea is to have a file watcher inside vector that will automatically reload some file or files on filesystem when they are changed, and make content available to vrl script that is executed in remap transform or filter transform. In config it could look like a filepath and script that is run on content when file is changed to for example parse json, etc. And some key property that defines a identifier through which you can access this data in your transforms. The motivation for this very simple, there is a lot of state on the system that vector might not be aware of that is required to properly enrich collected metrics and logs before sending them to some centralized location. This information is usually easily acceptable on the node, and writing some script that would update a json file if let`s say network interfaces list change is not complicated.
Use Cases
In my case I want to have ability to resolve cilium network interfaces devices to pod ids when I collect network metrics.
Attempted Solutions
Right now vector vrl has option to send http request which in theory could be used to solve this, however I find it unsatisfactory because of the overhead, especially considering it would have to be run on each metric event.
Proposal
Here is incomplete config example to illustrate the idea:
References
No response
Version
vector 0.54.0 (x86_64-unknown-linux-gnu 2b8b875 2026-03-10 15:47:37.284215410)