V1 [OBSERVE] WAL size and growth rate metrics

[austin-barrington]

Context

Operators have no visibility into WAL growth. The WAL grows unbounded until the flush interval or next flush cycle. There is no metric to:

• Alert when the WAL is growing faster than it can be flushed
• Plan capacity for WAL storage
• Detect a stalled flusher

Deliverables

Add Prometheus gauges:

Metric	Type	Labels	Description
hyperbytedb_wal_bytes	Gauge	db	Current WAL size in bytes
hyperbytedb_wal_entry_count	Gauge	db	Number of entries in WAL
hyperbytedb_wal_oldest_entry_seconds	Gauge	db	Age in seconds of the oldest unflushed entry
hyperbytedb_wal_flush_lag_seconds	Gauge	db	Time since last successful flush

Update docs/user-guide/monitoring.md.

References

• WAL implementation — src/adapters/wal/
• docs/user-guide/monitoring.md — existing metrics docs

Priority

P2 — Observability, operations