Linpeng Huang

Predicting Multi-step Citywide Passenger Demands Using Attention-based Neural Networks

Predicting passenger pickup/dropoff demands based on historical mobility trips has been of great importance towards better vehicle distribution for the emerging mobility-on-demand (MOD) services. Prior works focused on predicting next-step passenger demands at selected locations or hotspots. However, we argue that multi-step citywide passenger demands encapsulate both time-varying demand trends and global statuses, and hence are more beneficial to avoiding demand-service mismatching and developing effective vehicle distribution/scheduling strategies. In this paper, we propose an end-to-end deep neural network solution to the prediction task. We employ the encoder-decoder framework based on convolutional and ConvLSTM units to identify complex features that capture spatiotemporal influences and pickup-dropoff interactions on citywide passenger demands. A novel attention model is incorporated to emphasize the effects of latent citywide mobility regularities. We evaluate our proposed method using real-word mobility trips (taxis and bikes) and the experimental results show that our method achieves higher prediction accuracy than the adaptations of the state-of-the-art approaches.

NVHT: An efficient key–value storage library for non-volatile memory

Abstract Non-Volatile Memory (NVM) promises persistence, byte-addressability and DRAM-like read/write latency. These properties indicate that NVM has the potential to be incorporated with key–value stores to achieve high performance and durability simultaneously. Specifically, data can be stored in NVM inherently without DRAM buffering, which eliminates expensive disk I/Os and data format transformation cost. However, several challenges such as data inconsistency and write endurance arise along with the benefits. We propose a library named NVHT to provide APIs for NVM-based key–value store operations. In NVHT, we introduce non-volatile pointer to solve the dynamic address mapping problem and design a wear-out-aware memory allocator for NVM. The core of NVHT is a novel NVM-friendly hash table structure. NVHT guarantees consistency using a log-based mechanism. The experimental results show that compared with LevelDB and BerkeleyDB running on a DRAM-based file system, NVHT achieves more than 2x and 4x speedup for insert and search operation respectively. Compared with in-memory key–value store system Redis, NVHT still achieves higher transaction performance in terms of random update throughput (up to 1.5x and 2.5x for RDB scheme and AOF scheme, respectively).

A consistency mechanism for NVM-Based in-memory file systems

Non-Volatile Memory (NVM) has evolved to achieve non-volatility and byte-addressability with latency comparable to DRAM. This inspires the development of a new generation of file systems, namely NVM-based in-memory file systems, which include NVM on memory bus and allow in-NVM data to be directly accessed like DRAM. Meanwhile, an important issue, the consistency problem, arises as a new challenge. That is, the direct modification to the in-NVM data can be interrupted by arbitrary crashes in the system, which results in part of the modification being durable and others being lost. Traditional consistency mechanisms assume the existence of DRAM buffering and hence cannot be applied to this hybrid memory architecture. While several consistency methods have been proposed for NVM-based in-memory file systems, most of them have side-effects including unfriendliness to DRAM and penalties on concurrency control, which degrade the system performance. In this paper, we propose a novel mechanism to guarantee the consistency of NVM-based in-memory file systems. We abstract the storage area as a layered structure and employ a lazy-validated snapshot strategy to achieve a high consistency level. Since every consistency method comes with a cost, we introduce several algorithms to efficiently deal with block-sharing and reduce the overhead of consistency mechanism. The experimental results show that our mechanism incurs negligible consistency overhead and outperforms a state-of-the-art snapshot file system by reducing the latency of snapshot taking and removal by 95% and 60% respectively.

DELF: A Dual-Embedding based Deep Latent Factor Model for Recommendation

Among various recommendation methods, latent factor models are usually considered to be state-ofthe-art techniques, which aim to learn user and item embeddings for predicting user-item preferences. When applying latent factor models to recommendation with implicit feedback, the quality of embeddings always suffers from inadequate positive feedback and noisy negative feedback. Inspired by the idea of NSVD that represents users based on their interacted items, this paper proposes a dualembedding based deep latent factor model named DELF for recommendation with implicit feedback. In addition to learning a single embedding for a user (resp. item), we represent each user (resp. item) with an additional embedding from the perspective of the interacted items (resp. users). We employ an attentive neural method to discriminate the importance of interacted users/items for dualembedding learning. We further introduce a neural network architecture to incorporate dual embeddings for recommendation. A novel attempt of DELF is to model each user-item interaction with four deep representations that are subtly fused for preference prediction. We conducted extensive experiments on real-world datasets. The results verify the effectiveness of user/item dual embeddings and the superior performance of DELF on item recommendation.

An Adaptive Eviction Framework for Anti-caching Based In-Memory Databases

Current in-memory DBMSs suffer from the performance bottleneck when data cannot fit in memory. To solve such a problem, anti-caching system is proposed and with proper configuration, it can achieve better performance than state-of-the-art counterpart. However, in current anti-caching eviction procedure, all the eviction parameters are fixed while real workloads keep changing from time to time. Therefore, the performance of anti-caching system can hardly stay in the best state. We propose an adaptive eviction framework for anti-caching system and implement four tuning techniques to automatically tune the eviction parameters. In particular, we design a novel tuning technique called window-size adaption specialized for anti-caching system and embed it into the adaptive eviction framework. The experimental results show that with adaptive eviction, anti-caching based database system can outperform the traditional prototype by 1.2x–1.8x and 1.7x–4.5x under TPC-C benchmark and YCSB benchmark, respectively.

HMFS: A hybrid in-memory file system with version consistency

Abstract Emerging non-volatile memory (NVM) such as PCM and STT-RAM has memory-like byte-addressability as well as disk-like persistent storage capability. It offers an opportunity to bring NVM into the existing computer architecture for constructing an efficient and high-performance in-memory file system. Several NVM-optimized file systems have been designed. However, most of them fail to exploit all important features of NVM, and can only guarantee the file system consistency to the data consistency level. In this paper, we present HMFS, a hybrid in-memory full-versioning file system. HMFS manages DRAM and NVM in a unified address space and adopts different updating mechanisms to them. Besides, HMFS achieves version consistency with a simple and efficient multi-version approach. Experimental results show that HMFS achieves significant throughput improvement comparing with the state-of-the-art NVM-optimized file systems, such as PMFS and NOVA, and 3 . 1 × to 13 . 5 × higher versioning efficiency compared to some other multi-versioned file system such as BTRFS and NILFS2.

Adaptive Prefetching for Accelerating Read and Write in NVM-Based File Systems

The byte-addressable Non-Volatile Memory (NVM) offers fast, fine-grained access to persistent storage. While DRAM and NVM have similar read performance, the write operations of existing NVM materials incur longer latency and lower bandwidth than DRAM. This read-write asymmetry nature of NVM causes two bottlenecks for accessing read-and write-intensive file data: expensive data block lookups via file inner structure and high-latency direct writes to data blocks in NVM. However, existing NVM-based file systems fail to address both bottlenecks well. This paper presents WARP, an adaptive prefetching module designed for NVM-based file systems, which aims to deal with two bottlenecks effectively. WARP employs two acceleration approaches: 1) mapping data blocks into kernel virtual address space to bypass the indirection of file inner structure for read-intensive file data; and 2) allocating DRAM buffer to absorb frequent writes for write-intensive file data. We design a WARP benefit model to identify read-and write-intensive access patterns for file data, and use a successor prediction model to predict future data access based on historical file access traces. With WARP, we are able to prefetch file data according to both file access patterns and traces with consistency guarantee. WARP can be implemented on various NVM-based file systems, and we choose HMVFS for the experiments. The evaluation results show that HMVFS with WARP provides high prefetching accuracy and up to 32%-83% improvement compared with the state-of-the-art NVM-based file systems.

HMVFS: A Versioning File System on DRAM/NVM Hybrid Memory

Abstract The byte-addressable Non-Volatile Memory (NVM) offers fast, fine-grained access to persistent storage, and a large volume of recent researches are conducted on developing NVM-based in-memory file systems. However, existing approaches focus on low-overhead access to the memory and only guarantee the consistency between data and metadata. In this paper, we address the problem of maintaining consistency among continuous snapshots for NVM-based in-memory file systems. We propose an efficient versioning mechanism and implement it in Hybrid Memory Versioning File System (HMVFS), which achieves fault tolerance efficiently and has low impact on I/O performance. Our results show that HMVFS provides better performance on snapshotting and recovering compared with the traditional versioning file systems for many workloads. Specifically, HMVFS has lower snapshotting overhead than BTRFS and NILFS2, improving by a factor of 9.7 and 6.6, respectively. Furthermore, HMVFS imposes minor performance overhead compared with the state-of-the-art in-memory file systems like PMFS.

LibreKV: A Persistent In-Memory Key-Value Store

Emerging Non-Volatile Memory (NVM) possesses unique features including byte-addressability and high density which bring huge opportunities to combine DRAM and NVM in a unified main memory space. And key-value store (KVS) systems play an important role in many applications, such as large-scale websites. Several existing KVS have been proposed for NVMs. However, they have drawbacks such as extra write overhead leading, lower memory utilization and worse endurance. In this paper, we design and develop an NVM-based key-value store system named LibreKV. It specifically targets the hybrid DRAM and NVM memory architecture, leveraging the NVM as the eventual persistent storage medium. It uses both static hash table and dynamic hash tables to achieve a balance between system performance and memory utilization. It adopts a checksum based consistency mechanism to guarantee data consistency and persistent storage on NVM. LibreKV works independently without relying on an underlying file system and simplified the IO stack comparing to the traditional KVS system. Experimental results show that LibreKV outperforms the state-of-the-art KVS systems and achieves better scalability and consistency while with low overhead.

NVHT: an efficient key-value storage library for non-volatile memory

Modern Non-Volatile Memory (NVM) promises persistence, byte-addressability and DRAM-like read and write latency, which offers great opportunities for big data storage architecture. These excellent properties indicate that NVM has the potential to be incorporated with key-value stores to achieve high performance and durability simultaneously. In this paper, we propose an efficient key-value storage library named NVHT for NVM. NVHT provides APIs to facilitate the development of NVM-based key-value stores where data can be stored and processed in NVM inherently. The core structure of NVHT is a hash table with NVM-optimized metadata design. In NVHT, we introduce non-volatile pointer to solve the dynamic address mapping problem and develop a simple yet efficient memory allocator called pointer-free buddy allocator. NVHT guarantees consistency using a log-based mechanism. The experimental results show that NVHT achieves 1.3x-4x speedup for insert operation and 4x-5x speedup for search operation, compared with two advanced key-value stores, LevelDB and BerkeleyDB running on DRAM-based file system.