ByteDance Unveils USDB to Optimize Recommender Storage
ByteDance introduces USDB, a unified sequence storage system addressing the 'small updates, large reads' asymmetry in recommender models. By leveraging cascade LSM compaction, hybrid columnar formats, and aggregation pushdown, USDB reduces infrastructure costs by 42.1% and mean write latency by 85.8%, while significantly boosting throughput for long-history reads.
Full Report
Abstract models consume increasingly long, attribute-rich user histories, making small continuous updates and repeated large reads a central storage asymmetry. We present USDB, ByteDance's unified sequence storage practice for streaming and batch workloads. CascadeLSM time-based compaction stages materialize application-specific histories; differentiated replicas share persistent files; partial dumps reconstruct inference inputs from captured recent state and versioned history; and a hybrid columnar format accelerates projected reads. These designs reuse maintenance results across ingestion, inference, and training. Production migration records show 85.8% lower mean write latency, while normalized deployment accounting shows 42.1% lower infrastructure cost. Controlled comparisons with an Optimized RocksDB sequence baseline show faster long-history reads across projections and client loads. Aggregation pushdown sustains 8.5x the throughput of ordinary reads under a constrained receive budget. Production reconstruction and direct-file training results demonstrate how shared sequence maintenance supports fresh serving and repeated historical access.