Recently, very fast deterministic and randomized algorithms have been obtained for connectivity and minimum spanning tree in the unicast congested clique. In contrast, no solution faster than a simple parallel impleme...
详细信息
The proceedings contain 31 papers. The special focus in this conference is on NASA Formal Methods. The topics include: An automata-theoretic approach to modeling systems and specifications over infinite data;parametri...
ISBN:
(纸本)9783319572871
The proceedings contain 31 papers. The special focus in this conference is on NASA Formal Methods. The topics include: An automata-theoretic approach to modeling systems and specifications over infinite data;parametric model checking timed automata under non zenoness assumption;multi-timed bisimulation for distributed timed automata;on learning sparse Boolean formulae for explaining AI decisions;event-based runtime verification of temporal properties using time basic Petri nets;model-counting approaches for nonlinear numerical constraints;input space partitioning to enable massively parallel proof;compositional model checking of interlocking systems for lines with multiple stations;modular model-checking of a byzantine fault-tolerant protocol;improved learning for stochastic timed models by state-merging algorithms;verifying safety and persistence properties of hybrid systems using flowpipes and continuous invariants;floating-point format inference in mixed-precision;a verification technique for deterministic parallel programs;systematic predicate abstraction using variable roles;from hazard analysis to hazard mitigation planning;some lessons learnt from an application to a robot anti-collision function;reasoning about safety-critical information flow between pilot and computer;compositional falsification of cyber-physical systems with machine learning components;verifying a class of certifying distributed programs;qualification of a model checker for avionics software verification;formalized past LTL specification and analysis of requirements.
The coalescence protocol plays an important role in the population protocol model. The conceptual structure of the protocol is for two agents holding two non-zero values a, b respectively to take a transition (a, b) ...
详细信息
The proceedings contain 20 papers. The topics discussed include: HyperDrive: exploring hyperparameters with pop scheduling;sieve: actionable insights from monitored metrics in distributedsystems;Rafiki: a middleware ...
ISBN:
(纸本)9781450347204
The proceedings contain 20 papers. The topics discussed include: HyperDrive: exploring hyperparameters with pop scheduling;sieve: actionable insights from monitored metrics in distributedsystems;Rafiki: a middleware for parameter tuning of NoSQL datastores for dynamic metagenomics workloads;Rivulet: a fault-tolerant platform for smart-home applications;binary compatible graphics support in Android for running iOS apps;Sense-Aid: a framework for enabling network as a service for participatory sensing;ORCA: an ORChestration automata for configuring VNFs;improving spark application throughput via memory aware task co-location: a mixture of experts approach;Swayam: distributed autoscaling to meet SLAs of machine learning inference services with resource efficiency;data-driven serverless functions for object storage;a programming model for application-defined multipath TCP scheduling;POLM2: automatic profiling for object lifetime-aware memory management for hotspot big data applications;SPECTRE: supporting consumption policies in window-based parallel complex event processing;efficient covering for top-k filtering in content-based publish/subscribe systems;StreamApprox: approximate computing for stream analytics;X-Search: revisiting private web search using Intel SGX;rectify: black-box intrusion recovery in PaaS Clouds;scheduler activations for interference-resilient SMP virtual machine scheduling;DoubleDecker: a cooperative disk caching framework for derivative clouds;and Ginja: one-dollar cloud-based disaster recovery for databases.
In recent years, the generated and collected data is increasing at an almost exponential rate. At the same time, the data’s value has been identified in terms of insights that can be provided. However, retrieving the...
详细信息
ISBN:
(纸本)9783885796596
In recent years, the generated and collected data is increasing at an almost exponential rate. At the same time, the data’s value has been identified in terms of insights that can be provided. However, retrieving the value requires powerful analysis tools, since valuable insights are buried deep in large amounts of noise. Unfortunately, analytic capacities did not scale well with the growing data. Many existing tools run only on a single computer and are limited in terms of data size by its memory. A very promising solution to deal with large-scale data is scaling systems and exploiting parallelism. In this paper, we propose Gilbert, a distributed sparse linear algebra system, to decrease the imminent lack of analytic capacities. Gilbert offers a MATLAB®-like programming language for linear algebra programs, which are automatically executed in parallel. Transparent parallelization is achieved by compiling the linear algebra operations first into an intermediate representation. This language-independent form enables high-level algebraic optimizations. Different optimization strategies are evaluated and the best one is chosen by a cost-based optimizer. The optimized result is then transformed into a suitable format for parallel execution. Gilbert generates execution plans for Apache Spark® and Apache Flink®, two massively parallel dataflow systems. distributed matrices are represented by square blocks to guarantee a well-balanced trade-off between data parallelism and data granularity. An exhaustive evaluation indicates that Gilbert is able to process varying amounts of data exceeding the memory of a single computer on clusters of different sizes. Two well known machine learning (ML) algorithms, namely PageRank and Gaussian non-negative matrix factorization (GNMF), are implemented with Gilbert. The performance of these algorithms is compared to optimized implementations based on Spark and Flink. Even though Gilbert is not as fast as the optimized algorithms, it simpli
The proceedings contain 26 papers. The special focus in this conference is on Data Privacy Management. The topics include: A proof calculus for attack trees in isabelle;confidentiality of interactions in concurrent ob...
ISBN:
(纸本)9783319678153
The proceedings contain 26 papers. The special focus in this conference is on Data Privacy Management. The topics include: A proof calculus for attack trees in isabelle;confidentiality of interactions in concurrent object-oriented systems;using oblivious RAM in genomic studies;searchable encrypted relational databases;private verification of access on medical data: an initial study;threshold single password authentication;towards a toolkit for utility and privacy-preserving transformation of semi-structured data using data pseudonymization;order-preserving encryption using approximate integer common divisors;privacy-preserving deterministic automata evaluation with encrypted data blocks;securing proof-of-stake blockchain protocols;atomically trading with roger: gambling on the success of a hardfork;a user-centric system for verified identities on the bitcoin blockchain;towards a concurrent and distributed route selection for payment channel networks;secure event tickets on a blockchain.
Elastic distributed storage systems have been increasingly studied in recent years because power consumption has become a major problem in data centers. Much progress has been made in improving the agility of resizing...
详细信息
Elastic distributed storage systems have been increasingly studied in recent years because power consumption has become a major problem in data centers. Much progress has been made in improving the agility of resizing small- and large-scale distributed storage systems. However, most of these studies focus on metadata based distributed storage systems. On the other hand, emerging consistent hashing based distributed storage systems are considered to allow better scalability and are highly attractive. We identify challenges in achieving elasticity in consistent hashing based distributed storage. These challenges cannot be easily solved by techniques used in current studies. In this paper, we propose an elastic consistent hashing based distributed storage to solve two problems. First, in order to allow a distributed storage to resize quickly, we modify the data placement algorithm using a primary server design and achieve an equal-work data layout. Second, we propose a selective data re-integration technique to reduce the performance impact when resizing a cluster. Our experimental and trace analysis results confirm that our proposed elastic consistent hashing works effectively and allows significantly better elasticity.
暂无评论