Orchestration of configuration

How are folks managing configuration of ES and Kibana from source code?

For example, we have a number of Spaces, Users, Dashboards, Alerts, Pipeline configurations etc that we want to version and test through different environments up to production. We would like an Infrastructure as Code approach to manage this process against errors missed steps, etc.

How are people approaching this?

Thanks, Dan.

I don't have a good tutorial to point you to, but the saved object api is a good resource to look into. It allows you to load spaces/users/dashboards/... into Kibana via HTTP calls.

It probably makes sense to cross-post this question in the Elasticsearch forum as well.

Dan, you can read through the discussion here https://github.com/elastic/cloud-on-k8s/issues/3598 but it seems like the work on this has been put on hold. It's probably worth commenting on the issue with your use case.

For Kibana specifically we've got this issue open https://github.com/elastic/kibana/issues/2310. It's similar to Joe's suggestion to using the HTTP API, just a little bit more convenient.

Joe linked to the top-level saved object API docs, but I would suggest you look specifically at the import/export API https://www.elastic.co/guide/en/kibana/current/saved-objects-api-import.html

Thanks Joe and Rudolf,
We've previously tried the Kibana SavedObjects APIs and had difficulty getting them to work in the way we would like. Collective memory is hazy, but we didn't manage to get it to work for us at the time. It doesn't include some objects we would like to manage, such as Users and Spaces (though there are separate APIs for this of course). If more feedback occurs to us, we'll contribute to the issues you've mentioned.


This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.