ACM Transactions on Computer Systems最新文献

Cross-domain applications are rapidly adopting blockchain techniques for immutability, availability, integrity, and interoperability. However, for most applications, global consensus is unnecessary and may not even provide sufficient guarantees.

We propose a new distributed data structure: Attested Data Structures (ADS), which generalize not only blockchains, but also many other structures used by distributed applications. As in blockchains, data in ADSs is immutable and self-authenticating. ADSs go further by supporting application-defined proofs (attestations). Attestations enable applications to plug in their own mechanisms to ensure availability and integrity.

We present Charlotte, a framework for composable ADSs. Charlotte deconstructs conventional blockchains into more primitive mechanisms. Charlotte can be used to construct blockchains, but does not impose the usual global-ordering overhead. Charlotte offers a flexible foundation for interacting applications that define their own policies for availability and integrity. Unlike traditional distributed systems, Charlotte supports heterogeneous trust: different observers have their own beliefs about who might fail, and how. Nevertheless, each observer has a consistent, available view of data.

Charlotte’s data structures are interoperable and composable: applications and data structures can operate fully independently, or can share data when desired. Charlotte defines a language-independent format for data blocks and a network API for servers.

To demonstrate Charlotte’s flexibility, we implement several integrity mechanisms, including consensus and proof of work. We explore the power of disentangling availability and integrity mechanisms in prototype applications. The results suggest that Charlotte can be used to build flexible, fast, composable applications with strong guarantees.

We present an extensive study focused on partial network partitioning. Partial network partitions disrupt the communication between some but not all nodes in a cluster. First, we conduct a comprehensive study of system failures caused by this fault in 13 popular systems. Our study reveals that the studied failures are catastrophic (e.g., lead to data loss), easily manifest, and are mainly due to design flaws. Our analysis identifies vulnerabilities in core systems mechanisms including scheduling, membership management, and ZooKeeper-based configuration management.

Second, we dissect the design of nine popular systems and identify four principled approaches for tolerating partial partitions. Unfortunately, our analysis shows that implemented fault tolerance techniques are inadequate for modern systems; they either patch a particular mechanism or lead to a complete cluster shutdown, even when alternative network paths exist.

Finally, our findings motivate us to build Nifty, a transparent communication layer that masks partial network partitions. Nifty builds an overlay between nodes to detour packets around partial partitions. Nifty provides an approach for applications to optimize their operation during a partial partition. We demonstrate the benefit of this approach through integrating Nifty with VoltDB, HDFS, and Kafka.

Issue time prediction processors use dataflow dependencies and predefined instruction latencies to predict issue times of repeated instructions. In this work, we make two key observations: (1) memory accesses often take additional time to arrive than the static, predefined access latency that is used to describe these systems. This is due to contention in the memory hierarchy and variability in DRAM access times, and (2) we find that these memory access delays often repeat across iterations of the same code. We propose a new processor microarchitecture that replaces a complex reservation-station-based scheduler with an efficient, scalable alternative. Our scheduling technique tracks real-time delays of loads to accurately predict instruction issue times and uses a reordering mechanism to prioritize instructions based on that prediction. To accomplish this in an energy-efficient manner we introduce (1) an instruction delay learning mechanism that monitors repeated load instructions and learns their latest delay, (2) an issue time predictor that uses learned delays and dataflow dependencies to predict instruction issue times, and (3) priority queues that reorder instructions based on their issue time prediction. Our processor achieves 86.2% of the performance of a traditional out-of-order processor, higher than previous efficient scheduler proposals, while consuming 30% less power.

Intellectual Property (IP) reuse is a well known practice in chip design processes. Nowadays, network-on-chips (NoCs) are increasingly used as IP and sold by various vendors to be integrated in a multiprocessor system-on-chip (MPSoC). However, IP reuse exposes the design to IP theft, and an attacker can launch IP stealing attacks against NoC IPs. With the growing adoption of MPSoC, such attacks can result in huge financial losses. In this article, we propose four NoC IP protection techniques using fingerprint embedding: ON-OFF router-based fingerprinting (ORF), ON-OFF link-based fingerprinting (OLF), Router delay-based fingerprinting (RTDF), and Row delay-based fingerprinting (RWDF). ORF and OLF techniques use patterns of ON-OFF routers and links, respectively, while RTDF and RWDF techniques use router delays to embed fingerprints. We show that all of our proposed techniques require much less hardware overhead compared to an existing NoC IP security solution (square spiral routing) and also provide better security from removal and masking attacks. In particular, our proposed techniques require between 40.75% and 48.43% less router area compared to the existing solution. We also show that our solutions do not affect the normal packet latency and hence do not degrade the NoC performance.

Aggregation is common in data analytics and crucial to distilling information from large datasets, but current data analytics frameworks do not fully exploit the potential for optimization in such phases. The lack of optimization is particularly notable in current “online” approaches that store data in main memory across nodes, shifting the bottleneck away from disk I/O toward network and compute resources, thus increasing the relative performance impact of distributed aggregation phases.

We present ROME, an aggregation system for use within data analytics frameworks or in isolation. ROME uses a set of novel heuristics based primarily on basic knowledge of aggregation functions combined with deployment constraints to efficiently aggregate results from computations performed on individual data subsets across nodes (e.g., merging sorted lists resulting from top-k). The user can either provide minimal information that allows our heuristics to be applied directly, or ROME can autodetect the relevant information at little cost. We integrated ROME as a subsystem into the Spark and Flink data analytics frameworks. We use real-world data to experimentally demonstrate speedups up to 3× over single-level aggregation overlays, up to 21% over other multi-level overlays, and 50% for iterative algorithms like gradient descent at 100 iterations.

In this work, we present libHetMP, an OpenMP runtime for automatically and transparently distributing parallel computation across heterogeneous nodes. libHetMP targets platforms comprising CPUs with different instruction set architectures (ISA) coupled by a high-speed memory interconnect, where cross-ISA binary incompatibility and non-coherent caches require application data be marshaled to be shared across CPUs. Because of this, work distribution decisions must take into account both relative compute performance of asymmetric CPUs and communication overheads. libHetMP drives workload distribution decisions without programmer intervention by measuring performance characteristics during cross-node execution. A novel HetProbe loop iteration scheduler decides if cross-node execution is beneficial and either distributes work according to the relative performance of CPUs when it is or places all work on the set of homogeneous CPUs providing the best performance when it is not. We evaluate libHetMP using compute kernels from several OpenMP benchmark suites and show a geometric mean 41% speedup in execution time across asymmetric CPUs. Because some workloads may showcase irregular behavior among iterations, we extend libHetMP with a second scheduler called HetProbe-I. The evaluation of HetProbe-I shows it can further improve speedup for irregular computation, in some cases up to a 24%, by triggering periodic distribution decisions.

To process real-world datasets, modern data-parallel systems often require extremely large amounts of memory, which are both costly and energy inefficient. Emerging non-volatile memory (NVM) technologies offer high capacity compared to DRAM and low energy compared to SSDs. Hence, NVMs have the potential to fundamentally change the dichotomy between DRAM and durable storage in Big Data processing. However, most Big Data applications are written in managed languages and executed on top of a managed runtime that already performs various dimensions of memory management. Supporting hybrid physical memories adds a new dimension, creating unique challenges in data replacement. This article proposes Panthera, a semantics-aware, fully automated memory management technique for Big Data processing over hybrid memories. Panthera analyzes user programs on a Big Data system to infer their coarse-grained access patterns, which are then passed to the Panthera runtime for efficient data placement and migration. For Big Data applications, the coarse-grained data division information is accurate enough to guide the GC for data layout, which hardly incurs overhead in data monitoring and moving. We implemented Panthera in OpenJDK and Apache Spark. Based on Big Data applications’ memory access pattern, we also implemented a new profiling-guided optimization strategy, which is transparent to applications. With this optimization, our extensive evaluation demonstrates that Panthera reduces energy by 32–53% at less than 1% time overhead on average. To show Panthera’s applicability, we extend it to QuickCached, a pure Java implementation of Memcached. Our evaluation results show that Panthera reduces energy by 28.7% at 5.2% time overhead on average.