Many systems are designed to run graph algorithms efficiently in memory but they achieve only cache efficiency or work efficiency. We tackle this fundamental trade-off in existing systems by designing CoroGraph, a system that attains both cache efficiency and work efficiency for in-memory graph processing. CoroGraph adopts a novel hybrid execution model , which generates update messages at vertex granularity to prioritize promising vertices for work efficiency, and commits updates at partition granularity to share data access for cache efficiency. To overlap the random memory access of graph algorithms with computation, CoroGraph extensively uses coroutine , i.e., a lightweight function in C++ that can yield and resume with low overhead, to prefetch the required data. A suite of designs are incorporated to reap the full benefits of coroutine, which include prefetch pipeline, cache-friendly graph format, and stop-free synchronization. We compare CoroGraph with five state-of-the-art graph algorithm systems via extensive experiments. The results show that CoroGraph yields shorter algorithm execution time than all baselines in 18 out of 20 cases, and its speedup over the best-performing baseline can be over 2x. Detailed profiling suggests that CoroGraph achieves both cache efficiency and work efficiency with a low memory stall and a small number of processed edges.

Large-scale software systems are generally deployed on distributed machines. Logs are usually collected from those machines for comprehensive and accurate system fault analysis. However, there are potential challenges during log transmission from distributed machines to third-party data analytics services. First, uploading massive raw logs causes tremendous bandwidth consumption. Moreover, user privacy contained in logs is easy to get leaked during transmission. To address these issues, we introduce federated learning for anomaly detection using distributed log data. However, gradient updates of model parameters transmitted between the server (third-party data analytics services) and participants (distributed machines) in federated learning have been proved of possible recovery by attackers, so encryption of gradient updates is necessary for enhanced privacy protection. Considering that encryption time is proportional to the number of parameters, we propose a lightweight federated learning method for anomaly detection, named FLOGCNN, using distributed log data. The sever in FLOGCNN aggregates gradient updates according to the sample size of participants to generate an integrated model. For local training, participants apply an anomaly detection model based on one-dimensional convolution with much fewer parameters. Extensive experiments are conducted for FLOGCNN using open log datasets. Results demonstrate that FLOGCNN outperforms baseline methods on anomaly detection and reduces 97.08% parameters in comparison with one baseline method. Furthermore, we perform exploratory experiments on lightweight models and results manifest that logs with simple semantic information are suitable for lightweight anomaly detection models.

In this demo, we present SEBDB, a novel blockchain database that integrates immutability and transparency properties of blockchain with modeling and query ability of relational database. In summary, SEBDB has the following advantages: First, it adopts the linked structure and full replication of data among multiple participants to guarantee immutability and transparency. Second, it introduces the relational model to blockchain without introducing extra overhead, based on which relational queries are supported. SEBDB supports SQL-like language as the general interface to support convenient application development, in which intrinsic operations are re-defined and re-implemented to suit for blockchain platform. Third, it supports rich verifiable queries based on the proposed authenticated index, thin clients can participate in the system regardless of limitations of storage, network, and computing resources. We demonstrate the usability and scalability of SEBDB using a donation system.

Currently, users in a blockchain system must maintain all the data on the blockchain and query the data locally to ensure the integrity of the query results. However, since data is updated in an append-only way, resulting in a huge amount of data, it will take considerable maintenance costs to users. In this paper, we present an approach to support verifiable aggregate queries on blockchain systems that alleviates both storage and computing costs for users, while ensuring the integrity of the query results. We design an accumulator-based authenticated data structure (ADS) that supports verifiable multidimensional aggregate queries (i.e., aggregate queries with multiple selection predicates). The structure is built for each block, based on which verifiable multidimensional aggregate queries within a single block or involving multiple blocks are supported. We further optimize the performance by merging ADSs on different blocks to reduce the verification time at the client side and reduce the verification object (VO) size. Extensive experiments demonstrate the effectiveness and efficiency of our proposed approach.

Blockchain has been adopted in many applications to construct trust among multiple participants, such as supply chain management, digital assets transfer, philanthropy, etc. Blockchain platforms are often used as decentralized databases. However, existing blockchain platforms are far less convenient to use than traditional databases. They are lack of the capability of modelling complex tasks conveniently and efficiently, especially when both on-chain and off-chain data are involved at the same time. In this paper, we propose and implement a novel blockchain database, called SEBDB, which leverages the existing databases' functionality which are optimized for decades. Comparing to existing works, SEBDB is the first platform which considers both useability and scalability. Specifically, first, weaddrelationaldata semantics into blockchain platform, where each transaction is a tuple with multiple attributes in a pre-defined table. Second, we use SQL-like language as the general interface, instead of code-level APIs, to support convenient application development, in which intrinsic operations are re-defined and re-implemented to suit for blockchain platform. Third, as RDBMS has achieved great success in the past decades, our system, though not relying on RDBMS, treats it as an important component. Finally, we define a mini-benchmark to evaluate the performance of the blockchain database. Extensive experiments demonstrate the effectiveness and efficiency of our proposed system.

Non-uniform memory access (NUMA) architecture is widely used in high-end servers and computing systems due to its scalability. In recent years, the number of processor cores in NUMA systems increases rapidly with the development of multi-core processors. Along with the growing of system scales, simulation of NUMA systems becomes a challenge to traditional general-purpose simulators by reason of their low simulation performance. This paper presents SimNUMA, an execution-driven full-system simulator dedicated for NUMA systems. In the design of SimNUMA, to improve simulation performance significantly, the same type of processor with the target machine is used in the host system, and a new method to capture remote-memory accesses efficiently is proposed, in addition, parallel simulation is used to achieve scalability and improve performance. The modeling and simulation of interconnection networks are also supported. The simulator is tested in accuracy, scalability and performance, results show that the simulation slowdown is rather satisfying. Finally, the paper gives simulation experiments for different scales of target NUMA systems.