Sequential vs. Random Access¶

Sequential access exploits spatial locality and read-ahead optimization, achieving throughput orders of magnitude higher than random access, which pays seek/latency costs for each operation.

Tension Addressed¶

Sequential vs. Random Access Optimization

Recognizable As¶

Log-structured systems, append-only data structures, batch processing pipelines, database index design choices, any system that "compacts" or "reorganizes" data

Canonical Sources¶

Rosenblum & Ousterhout, "The Design and Implementation of a Log-Structured File System" (1991); O'Neil et al., "The Log-Structured Merge-Tree" (1996)

Transfers To¶

Database storage engines, message queues, time-series databases, event sourcing, backup systems

This tension explains why databases offer multiple index types, why log-structured merge trees exist, and why batch processing systems outperform transactional systems for large-scale analytics.