arXiv subjects

Huijun Wu

Publications and source records attributed to Huijun Wu.

14 recordsLinked to original sources

arXiv2026

MARS: Multi-stage Accelerated Read Stack for Large-buffer Buffered Reads

Large-buffer reads increasingly connect data-intensive applications to high-speed storage. They amortize system-call overhead and create a larger in-kernel window for organizing page-cache work and submitting I/O. However, Linux buffered read primarily exploits only the former benefit. Within a large read, its conventional interleaved path repeatedly switches among fine-grained page-cache operations, amplifying metadata and serial orchestration overheads and failing to consistently expose enough in-flight requests to modern parallel SSDs. We present MARS, a multi-stage accelerated read stack for synchronous large-buffer buffered reads. MARS treats each large-range read as one unit of work and stages page-cache operations by data structure and dependency. During I/O waits, it handles user-buffer page faults and performs reorderable data copies early. Opportunistic kernel workers then copy remaining data in parallel and, when the backend provides sufficient parallelism, optionally submit I/O in parallel. We implement MARS in Linux 6.6.58. For MiB-scale fio reads, MARS improves bandwidth by up to 6.56 times over Linux. On five NVMe SSDs in RAID0, it reaches 36.87 GiB/s for 128 MiB random reads, 4.44 times Linux. MARS also accelerates DuckDB/Parquet queries by 1.80--2.15 times and ExecuTorch model loading by 3.17--3.61 times.

cs.OS↗

arXiv2026

Rethinking Burst Buffer Optimization: Enabling Layout Heterogeneity via Hybrid Analysis and LLM Guidance

Burst buffers (BBs) are essential for mitigating I/O bottlenecks in modern HPC systems. However, existing BB file systems often suffer from structural performance degradation due to fixed data layouts that fail to align with diverse application behaviors. While current machine-learning-based optimizations focus primarily on tuning storage stack parameters for a given layout, they offer diminishing returns when a fundamental mismatch exists between I/O patterns and the underlying data organization. Furthermore, these approaches typically incur prohibitive costs due to extensive training or intrusive profiling. To bridge this gap, we present Proteus, a semantic-aware BB system that treats data layout as a first-class optimization dimension. The core insight of Proteus is that application I/O intent can be reconstructed by synergetically combining static code structures with lightweight runtime signals. Through a hybrid pipeline and a single execution probe, Proteus extracts latent semantic cues to determine the optimal layout prior to production runs-eliminating the need for prior training or exhaustive profiling. Evaluation with representative HPC workloads shows that Proteus achieves 91.30\% decision accuracy, delivering up to 3.24$\times$ and 2.9$\times$ speedups for write-intensive and metadata-intensive workloads, respectively.

cs.DC↗

arXiv2024

Perseus: Leveraging Common Data Patterns with Curriculum Learning for More Robust Graph Neural Networks

Graph Neural Networks (GNNs) excel at handling graph data but remain vulnerable to adversarial attacks. Existing defense methods typically rely on assumptions like graph sparsity and homophily to either preprocess the graph or guide structure learning. However, preprocessing methods often struggle to accurately distinguish between normal edges and adversarial perturbations, leading to suboptimal results due to the loss of valuable edge information. Robust graph neural network models train directly on graph data affected by adversarial perturbations, without preprocessing. This can cause the model to get stuck in poor local optima, negatively affecting its performance. To address these challenges, we propose Perseus, a novel adversarial defense method based on curriculum learning. Perseus assesses edge difficulty using global homophily and applies a curriculum learning strategy to adjust the learning order, guiding the model to learn the full graph structure while adaptively focusing on common data patterns. This approach mitigates the impact of adversarial perturbations. Experiments show that models trained with Perseus achieve superior performance and are significantly more robust to adversarial attacks.

cs.LG↗

arXiv2024

Talos: A More Effective and Efficient Adversarial Defense for GNN Models Based on the Global Homophily of Graphs

Graph neural network (GNN) models play a pivotal role in numerous tasks involving graph-related data analysis. Despite their efficacy, similar to other deep learning models, GNNs are susceptible to adversarial attacks. Even minor perturbations in graph data can induce substantial alterations in model predictions. While existing research has explored various adversarial defense techniques for GNNs, the challenge of defending against adversarial attacks on real-world scale graph data remains largely unresolved. On one hand, methods reliant on graph purification and preprocessing tend to excessively emphasize local graph information, leading to sub-optimal defensive outcomes. On the other hand, approaches rooted in graph structure learning entail significant time overheads, rendering them impractical for large-scale graphs. In this paper, we propose a new defense method named Talos, which enhances the global, rather than local, homophily of graphs as a defense. Experiments show that the proposed approach notably outperforms state-of-the-art defense approaches, while imposing little computational overhead.

cs.LG↗

arXiv2024

GraphLearner: Graph Node Clustering with Fully Learnable Augmentation

Contrastive deep graph clustering (CDGC) leverages the power of contrastive learning to group nodes into different clusters. The quality of contrastive samples is crucial for achieving better performance, making augmentation techniques a key factor in the process. However, the augmentation samples in existing methods are always predefined by human experiences, and agnostic from the downstream task clustering, thus leading to high human resource costs and poor performance. To overcome these limitations, we propose a Graph Node Clustering with Fully Learnable Augmentation, termed GraphLearner. It introduces learnable augmentors to generate high-quality and task-specific augmented samples for CDGC. GraphLearner incorporates two learnable augmentors specifically designed for capturing attribute and structural information. Moreover, we introduce two refinement matrices, including the high-confidence pseudo-label matrix and the cross-view sample similarity matrix, to enhance the reliability of the learned affinity matrix. During the training procedure, we notice the distinct optimization goals for training learnable augmentors and contrastive learning networks. In other words, we should both guarantee the consistency of the embeddings as well as the diversity of the augmented samples. To address this challenge, we propose an adversarial learning mechanism within our method. Besides, we leverage a two-stage training strategy to refine the high-confidence matrices. Extensive experimental results on six benchmark datasets validate the effectiveness of GraphLearner.The code and appendix of GraphLearner are available at https://github.com/xihongyang1999/GraphLearner on Github.

cs.LG↗

arXiv2022

Taming Hybrid-Cloud Fast and Scalable Graph Analytics at Twitter

We have witnessed a boosted demand for graph analytics at Twitter in recent years, and graph analytics has become one of the key parts of Twitter's large-scale data analytics and machine learning for driving engagement, serving the most relevant content, and promoting healthier conversations. However, infrastructure for graph analytics has historically not been an area of investment at Twitter, resulting in a long timeline and huge engineering effort for each project to deal with graphs at the Twitter scale. How do we build a unified graph analytics user experience to fulfill modern data analytics on various graph scales spanning from thousands to hundreds of billions of vertices and edges? To bring fast and scalable graph analytics capability into production, we investigate the challenges we are facing in large-scale graph analytics at Twitter and propose a unified graph analytics platform for efficient, scalable, and reliable graph analytics across on-premises and cloud, to fulfill the requirements of diverse graph use cases and challenging scales. We also conduct quantitative benchmarking on Twitter's production-level graph use cases between popular graph analytics frameworks to certify our solution.

cs.DB↗

arXiv2022

Serving Hybrid-Cloud SQL Interactive Queries at Twitter

The demand for data analytics has been consistently increasing in the past years at Twitter. In order to fulfill the requirements and provide a highly scalable and available query experience, a large-scale in-house SQL system is heavily relied on. Recently, we evolved the SQL system into a hybrid-cloud SQL federation system, compliant with Twitter's Partly Cloudy strategy. The hybrid-cloud SQL federation system is capable of processing queries across Twitter's data centers and the public cloud, interacting with around 10PB of data per day. In this paper, the design of the hybrid-cloud SQL federation system is presented, which consists of query, cluster, and storage federations. We identify challenges in a modern SQL system and demonstrate how our system addresses them with some important design decisions. We also conduct qualitative examinations and summarize instructive lessons learned from the development and operation of such a SQL system.

cs.DB↗

arXiv2022

CWcollab: A Context-Aware Web-Based Collaborative Multimedia System

Remote collaboration tools for conferencing and presentation are gaining significant popularity during the COVID-19 pandemic period. Most prior work has issues, such as a) limited support for media types, b) lack of interactivity, for example, an efficient replay mechanism, c) large bandwidth consumption for screen sharing tools. In this paper, we propose a general-purpose multimedia collaboration platform-CWcollab. It supports collaboration on general multimedia by using simple messages to represent media controls with an object-prioritized synchronization approach. Thus, CWcollab can not only support fine-grained accurate collaboration, but also rich functionalities such as replay of these collaboration events. The evaluation shows hundreds of kilobytes can be enough to store the events in a collaboration session for accurate replays, compared with hundreds of megabytes of Google Hangouts.

cs.MM↗

arXiv2022

Forecasting SQL Query Cost at Twitter

With the advent of the Big Data era, it is usually computationally expensive to calculate the resource usages of a SQL query with traditional DBMS approaches. Can we estimate the cost of each query more efficiently without any computation in a SQL engine kernel? Can machine learning techniques help to estimate SQL query resource utilization? The answers are yes. We propose a SQL query cost predictor service, which employs machine learning techniques to train models from historical query request logs and rapidly forecasts the CPU and memory resource usages of online queries without any computation in a SQL engine. At Twitter, infrastructure engineers are maintaining a large-scale SQL federation system across on-premises and cloud data centers for serving ad-hoc queries. The proposed service can help to improve query scheduling by relieving the issue of imbalanced online analytical processing (OLAP) workloads in the SQL engine clusters. It can also assist in enabling preemptive scaling. Additionally, the proposed approach uses plain SQL statements for the model training and online prediction, indicating it is both hardware and software-agnostic. The method can be generalized to broader SQL systems and heterogeneous environments. The models can achieve 97.9\% accuracy for CPU usage prediction and 97\% accuracy for memory usage prediction.

cs.DB↗

arXiv2021

CoG: a Two-View Co-training Framework for Defending Adversarial Attacks on Graph

Graph neural networks exhibit remarkable performance in graph data analysis. However, the robustness of GNN models remains a challenge. As a result, they are not reliable enough to be deployed in critical applications. Recent studies demonstrate that GNNs could be easily fooled with adversarial perturbations, especially structural perturbations. Such vulnerability is attributed to the excessive dependence on the structure information to make predictions. To achieve better robustness, it is desirable to build the prediction of GNNs with more comprehensive features. Graph data, in most cases, has two views of information, namely structure information and feature information. In this paper, we propose CoG, a simple yet effective co-training framework to combine these two views for the purpose of robustness. CoG trains sub-models from the feature view and the structure view independently and allows them to distill knowledge from each other by adding their most confident unlabeled data into the training set. The orthogonality of these two views diversifies the sub-models, thus enhancing the robustness of their ensemble. We evaluate our framework on three popular datasets, and results show that CoG significantly improves the robustness of graph models against adversarial attacks without sacrificing their performance on clean data. We also show that CoG still achieves good robustness when both node features and graph structures are perturbed.

cs.LG↗

arXiv2019

Lightweight Container-based User Environment

Modern operating systems all support multi-users that users could share a computer simultaneously and not affect each other. However, there are some limitations. For example, privacy problem exists that users are visible to each other in terms of running processes and files. Moreover, users have little freedom to customize the system environment. Last, it is a burden for system administrator to safely manage and update system environment while satisfying multiple users. Facing the above problems, this paper proposes CUE, a Lightweight Container-based User Environment. CUE proposes a new notion that stands in between application container and operating system container:user container. CUE is able to give users more flexibility to customize their environment, achieve privacy isolation, and make system update easier and safer. Its goal is to optimize and enhance the multi-user notion of current operating system and being lightweight. Moreover, it is able to facilitate application deployment in high performance clusters. It is currently deployed in NUDT's Tianhe E prototype supercomputer. Experiment results show that it introduces negligible overhead.

cs.DC↗

arXiv2019

Adversarial Examples on Graph Data: Deep Insights into Attack and Defense

Graph deep learning models, such as graph convolutional networks (GCN) achieve remarkable performance for tasks on graph data. Similar to other types of deep models, graph deep learning models often suffer from adversarial attacks. However, compared with non-graph data, the discrete features, graph connections and different definitions of imperceptible perturbations bring unique challenges and opportunities for the adversarial attacks and defenses for graph data. In this paper, we propose both attack and defense techniques. For attack, we show that the discreteness problem could easily be resolved by introducing integrated gradients which could accurately reflect the effect of perturbing certain features or edges while still benefiting from the parallel computations. For defense, we observe that the adversarially manipulated graph for the targeted attack differs from normal graphs statistically. Based on this observation, we propose a defense approach which inspects the graph and recovers the potential adversarial perturbations. Our experiments on a number of datasets show the effectiveness of the proposed methods.

cs.LG↗

arXiv2017

Interpreting Shared Deep Learning Models via Explicable Boundary Trees

Despite outperforming the human in many tasks, deep neural network models are also criticized for the lack of transparency and interpretability in decision making. The opaqueness results in uncertainty and low confidence when deploying such a model in model sharing scenarios, when the model is developed by a third party. For a supervised machine learning model, sharing training process including training data provides an effective way to gain trust and to better understand model predictions. However, it is not always possible to share all training data due to privacy and policy constraints. In this paper, we propose a method to disclose a small set of training data that is just sufficient for users to get the insight of a complicated model. The method constructs a boundary tree using selected training data and the tree is able to approximate the complicated model with high fidelity. We show that traversing data points in the tree gives users significantly better understanding of the model and paves the way for trustworthy model sharing.

cs.LG↗

arXiv2017

HPDedup: A Hybrid Prioritized Data Deduplication Mechanism for Primary Storage in the Cloud

Eliminating duplicate data in primary storage of clouds increases the cost-efficiency of cloud service providers as well as reduces the cost of users for using cloud services. Existing primary deduplication techniques either use inline caching to exploit locality in primary workloads or use post-processing deduplication running in system idle time to avoid the negative impact on I/O performance. However, neither of them works well in the cloud servers running multiple services or applications for the following two reasons: Firstly, the temporal locality of duplicate data writes may not exist in some primary storage workloads thus inline caching often fails to achieve good deduplication ratio. Secondly, the post-processing deduplication allows duplicate data to be written into disks, therefore does not provide the benefit of I/O deduplication and requires high peak storage capacity. This paper presents HPDedup, a Hybrid Prioritized data Deduplication mechanism to deal with the storage system shared by applications running in co-located virtual machines or containers by fusing an inline and a post-processing process for exact deduplication. In the inline deduplication phase, HPDedup gives a fingerprint caching mechanism that estimates the temporal locality of duplicates in data streams from different VMs or applications and prioritizes the cache allocation for these streams based on the estimation. HPDedup also allows different deduplication threshold for streams based on their spatial locality to reduce the disk fragmentation. The post-processing phase removes duplicates whose fingerprints are not able to be cached due to the weak temporal locality from disks. Our experimental results show that HPDedup clearly outperforms the state-of-the-art primary storage deduplication techniques in terms of inline cache efficiency and primary deduplication efficiency.

cs.DC↗