Robert Gildein
Robert Gildein
I really like the idea of using `Jinja2` and I created an issue that is duplicated with this. I'll close it and put an original description here. > I think...
I can confirm that version 7> is working fine.
This is really helpful and I will use it, thanks. However, this does help to deploy the charm, but the `charm-build` still copies those directories.
Maybe we can use similar approach as SolEng team used [here](https://github.com/canonical/solutions-engineering-automation/tree/main/terraform-plans/modules/GitHub/settings). They are using Terraform to set repos.
Summary of my investigation around getting metrics from ServingRuntimes and ClusterServingRuntimes via **kserve-controller** charm. These metrics should include: request_preprocess_seconds | pre-processing request latency | Histogram -- | -- | --...
> I was wondering though, shouldn't we enable metrics by default by updating the apps' Dockerfiles? When I thought about it more, maybe you're right. My idea was not to...
@orfeas-k Should we update this issue?
I'm closing this issue because it's not updated and metrics have already been added everywhere, where upstream support it. More information can be found in epic KF-5558.
Original values: ```md ## kubeflow-profiles Alert | Description | Severity --- | --- | --- KfamDown | Kubeflow-kfam service is Down (instance {{ $labels.instance }}) | critical ProfilesDown | Kubeflow-profiles...