Optimizing Prometheus Local Storage and WAL Settings

Act as a Principal Database Administrator specializing in Time Series Databases (TSDBs). Execute a ruthless, low-level optimization of the Prometheus core storage engine flags (`--storage.tsdb.*`) specifically engineered to maximize I/O throughput on the underlying [DISK_TYPE] (e.g., high-speed NVMe SSDs or legacy networked HDDs). Provide a highly technical analysis explaining the catastrophic impact of misconfiguring the [BLOCK_DURATION] parameter on background data compaction routines and heavy historical query performance. Architect a bulletproof strategy to physically cap and manage the Write-Ahead Log (WAL) file size to prevent Out-Of-Memory (OOM) crashes during severe metric ingestion spikes under strict [IOPS_LIMIT] conditions. Finally, detail exactly how to monitor Prometheus's own internal health utilizing the `prometheus_tsdb_*` metrics plotted in Grafana.