How to configure log retention with tsdb (gcs, azure) on SimpleScalable setup?

I run Loki in SimpleScalable mode. I tried to configure log retention from Log retention | Grafana Loki documentation .
I understood the documentation in the way that that compactor should be configured instead of relying on the retention policy of the gcs bucket or azure storage container (“If you have a lifecycle policy configured on the object store, please ensure that it is longer than the retention period.”).

After a failed deployment and reading through the helm chart it seems to me as if compactor configuration is effective only in distributed mode (compactor… is only read in statefulset-compactor.yaml)

Even though the documentation is probably not wrong it is ambiguous to me.

How is basic compactor configuration (/docs/loki/latest/configure/examples/configuration-examples/#6-compactor-snippetyaml) applied in SimpleScalable mode? According to /docs/loki/latest/get-started/deployment-modes/#simple-scalable is run in the backend statefulset, yet there is no compactor reference in statefulset-backend.yaml in the chart.

  1. Compactor is part of the backend target in simple scalable mode, and it does work.
  2. You can have both Loki retention and lifecycle policy from object storage.