Summary
In-memory databases can store much more data using new hybrid memory setups that combine fast local DDR memory with slower CXL memory, but this mix creates challenges in how data and indexes are stored and accessed. The authors studied five different ways to organize indexes and data across these two memory types, finding trade-offs in speed, memory use, and system complexity. They designed a smart adaptive method that tracks how often different data parts are used and moves them between memory types to optimize performance. Their tests showed that keeping indexes and data managed separately with their adaptive method improves overall database throughput by up to 59%.
What this means in practice
- •For database engineers: Improve performance and memory efficiency of in-memory databases running on systems with mixed DDR and CXL memory.
- •For cloud infrastructure teams: Optimize memory usage and data placement in single-node servers that integrate slower CXL memory expansions to host larger databases.
Authors
Huijie Cao, Shun Yang, Yinan Zhang, Huiqi Hu, Xuan Zhou, Weining Qian
Abstract
CXL memory expands single-node memory capacity for in-memory databases but has higher latency and lower bandwidth than local DDR. Hybrid DDR--CXL databases require joint indexing and data distribution design. Indexing designs differ in access and migration paths and constrain the placement of indexes and tuples, creating trade-offs in runtime performance, DDR memory efficiency, and system integration complexity. Their performance impact is thus difficult to determine. Indexes and tuples may differ in access characteristics: placing them in the same tier may reduce DDR efficiency, while separate management adds tracking and decision overhead. We propose five indexing designs based on two approaches: treating the two tiers as a unified data space or managing them separately. We analyze their trade-offs and underlying causes and develop adaptive data distribution optimization for single-index designs in which the database explicitly manages distribution. The method manages index objects and tuples separately, reduces metadata overhead through hierarchical hotness tracking and an index object residency policy, and extends benefit-aware admission and randomized eviction to determine and adjust their placement according to each type's expected benefits and memory footprints. We implemented the designs in a hybrid DDR--CXL database prototype and evaluated them using YCSB. The single-index design with direct addressing achieves the best performance under most workloads and system configurations. The dual-index design performs better mainly when DDR capacity is limited or accesses are highly concentrated, and has relatively low integration complexity. Adaptive distribution optimization improves DDR memory efficiency and system performance, enabling single-index designs to achieve up to $1.59\times$ the throughput of their counterparts with non-separated management.