Starburst EKS: Coordinator does not respect insights persistence config


I am having issues with a Starburst EKS deployment. The coordinator configuration is generated through
the official Helm Charts (obtained from ECR via AWS Marketplace offering). To be specific I want to disable
insights persistence for Coordinator.

The documentation states here: You must explicitly enable the persistence for Insights to use the persisted data in the backend service database. Defaults to false.

However my Coordinator container does not startup properly due to invalid configuration:
io.trino.server.Server Configuration is invalid

And the following log error entry: Invalid configuration property insights.jdbc.url: must not be null

According to the docs state here
Insights can be configured through the additionalProperties where I’ve set insights.persistence-enabled=false explicitly. This does also not prevent Coordinator from complaining about the missing insights.jdbc.url. However when I provide some JDBC URL coordinator validates its config but tries to actually connect via that URL.

Is there a way to disable the backend service persistence entirely?

Kind regards

Persistence is required for many enterprise features in Starburst. Basically we claim it is a required configuration, however we still support an option of having this disabled.

insights.persistence-enabled=false explicitly.

If you disabled it and then server still tries to use insights.jdbc.url then it sounds like a bug. Can you please share your

A little bit more context here… Customers are required to allow usage-metrics data to be collected for billing purposes (eg: You’re running X vCPU on your cluster, you used Y vCPU hours last month, etc.). This capability leverages persistence. So, while you theoretically can run without the feature turned on, we wouldn’t consider that a supported configuration.