# Fine-Grain Checkpointing with In-Cache-Line Logging

## TL;DR

This paper introduces Fine-Grain Checkpointing and In-Cache-Line Logging techniques to improve performance and recoverability of data structures in non-volatile memory, demonstrated on Masstree with low overhead.

## Contribution

The paper presents novel techniques for persistent data structures that reduce recovery time and overhead in non-volatile memory systems.

## Key findings

- Achieved quick recovery with minimal overhead (5.9-15.4%)
- Successfully integrated techniques into Masstree, a highly optimized data structure
- Demonstrated low-cost, consistent persistence in NVM systems

## Abstract

Non-Volatile Memory offers the possibility of implementing high-performance, durable data structures. However, achieving performance comparable to well-designed data structures in non-persistent (transient) memory is difficult, primarily because of the cost of ensuring the order in which memory writes reach NVM. Often, this requires flushing data to NVM and waiting a full memory round-trip time. In this paper, we introduce two new techniques: Fine-Grained Checkpointing, which ensures a consistent, quickly recoverable data structure in NVM after a system failure, and In-Cache-Line Logging, an undo-logging technique that enables recovery of earlier state without requiring cache-line flushes in the normal case. We implemented these techniques in the Masstree data structure, making it persistent and demonstrating the ease of applying them to a highly optimized system and their low (5.9-15.4\%) runtime overhead cost.