Abstract:High-performance flash-based key-value stores in data-centers utilize large amounts of DRAM to cache hot data. However, motivated by the high cost and power consumption of DRAM, server designs with lower DRAM-per-compute ratio are becoming popular. These low-cost servers enable scale-out services by reducing server workload densities. This results in improvements to overall service reliability, leading to a decrease in the total cost of ownership (TCO) for scalable workloads. Nevertheless, for key-value stores with large memory footprints, these reduced DRAM servers degrade performance due to an increase in both IO utilization and data access latency. In this scenario, a standard practice to improve performance for sharded databases is to reduce the number of shards per machine, which degrades the TCO benefits of reduced DRAM low-cost servers. In this work, we explore a practical solution to improve performance and reduce the costs and power consumption of key-value stores running on DRAM-constrained servers by using Storage Class Memories (SCM). SCMs in a DIMM form factor, although slower than DRAM, are sufficiently faster than flash when serving as a large extension to DRAM. With new technologies like Compute Express Link, we can expand the memory capacity of servers with high bandwidth and low latency connectivity with SCM. In this article, we use Intel Optane PMem 100 Series SCMs (DCPMM) in AppDirect mode to extend the available memory of our existing single-socket platform deployment of RocksDB (one of the largest key-value stores at Meta). We first designed a hybrid cache in RocksDB to harness both DRAM and SCM hierarchically. We then characterized the performance of the hybrid cache for three of the largest RocksDB use cases at Meta (ChatApp, BLOB Metadata, and Hive Cache). Our results demonstrate that we can achieve up to 80% improvement in throughput and 20% improvement in P95 latency over the existing small DRAM single-socket platform, while maintaining a 43–48% cost improvement over our large DRAM dual-socket platform. To the best of our knowledge, this is the first study of the DCPMM platform in a commercial data center.

HeterMM: applying in-DRAM index to heterogeneous memory-based key-value stores

Hotspot-Aware Hybrid Memory Management for In-Memory Key-Value Stores

SwapKV: A Hotness Aware In-memory Key-Value Store for Hybrid Memory Systems

HMM-V: Heterogeneous Memory Management for Virtualization

A Multi-hashing Index for hybrid DRAM-NVM memory systems

Online Application Guidance for Heterogeneous Memory Systems

HPDK: A Hybrid PM-DRAM Key-Value Store for High I/O Throughput

Hardware Memory Management for Future Mobile Hybrid Memory Systems

Demystifying the Performance of HPC Scientific Applications on NVM-based Memory Systems

A Performance Evaluation of DRAM Access for In-Memory Databases

A Distributed In-Memory Key-Value Store System on Heterogeneous CPU–GPU Cluster

TrieKV: A High-Performance Key-Value Store Design With Memory as Its First-Class Citizen

HeteroKV: A Scalable Line-rate Key-Value Store on Heterogeneous CPU-FPGA Platforms.

BonsaiKV: Towards Fast, Scalable, and Persistent Key-Value Stores with Tiered, Heterogeneous Memory System

A hybrid memory architecture supporting fine-grained data migration

HiStore: Rethinking Hybrid Index in RDMA-based Key-Value Store

Exploring the Performance Benefit of Hybrid Memory System on HPC Environments

Power-optimized Deployment of Key-value Stores Using Storage Class Memory

Observations on Porting In-memory KV stores to Persistent Memory

Mega-KV

MigrantStore: Leveraging Virtual Memory in DRAM-PCM Memory Architecture