Feature request: Add Prometheus retention duration switch to dynamically respond to disk space's free storage

[phanama]

Proposal

Feature request:
Auto-configure prometheus' retention duration in case of not enough disk space left.
Make a "true"/"false" switch to make Prometheus delete oldest data that are still in its retention duration window (not stale) if it is set to true, until some disk space threshold is achieved to make Prometheus keep running even if its disk space is full (this requires the organization's/users' soft-policy that the oldest non-stale data can be deleted, though).

Use case. Why is this important?:

• Really helps under the inability to expand disk space due to strict company's/organization's capacity-cost policy.
• Keep Promeheus running under full disk-space condition without needing operators to notice it, reducing ops-work (while still alerting ops to expand disk size).

Sometimes metrics are quite dynamic (in some use cases, we can add or delete metrics from a running Prometheus), so is its storage space calculation result in https://prometheus.io/docs/prometheus/latest/storage/

[brian-brazil]

Dupe of #968.

[lock]

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.