ALBERT

Description: Automatic recycling of intermediate results to improve both query response time and throughput is a grand challenge for state-of-the-art databases. Tuples are loaded and streamed through a tuple-at-a-time processing pipeline, avoiding materialization of intermediates as much as possible. This limits the opportunities for reuse of overlapping computations to DBA-defined materialized views and function/result cache tuning. In contrast, the operator-at-a-time execution paradigm produces fully materialized results in each step of the query plan. To avoid resource contention, these intermediates are evicted as soon as possible. In this article we study an architecture that harvests the byproducts of the operator-at-a-time paradigm in a column-store system using a lightweight mechanism, the recycler. The key challenge then becomes the selection of the policies to admit intermediates to the resource pool, to determine their retention period, and devise the eviction strategy when facing resource limitations. The proposed recycling architecture has been implemented in an open-source system. An experimental analysis against the TPC-H ad-hoc decision support benchmark and a complex, real-world application (SkyServer) demonstrates its effectiveness in terms of self-organizing behavior and its significant performance gains. The results indicate the potentials of recycling intermediates and charts a route for further development of database kernels.

Description: As network infrastructures with 10 Gb/s bandwidth and beyond have become pervasive and as cost advantages of large commodity-machine clusters continue to increase, research and industry strive to exploit the available processing performance for large-scale database processing tasks. In this work we look at the use of high-speed networks for distributed join processing. We propose Data Roundabout as alight weight transport layer that uses Remote Direct Memory Access (RDMA) to gain access to the throughput opportunities in modern networks. The essence of Data Roundabout is a ring shaped network in which each host stores one portion of a large database instance. We leverage the available bandwidth to (continuously) pump data through the high-speed network. Based on Data Roundabout, we demonstrate cyclo-join, which exploits the cycling flow of data to execute distributed joins. The study uses different join algorithms (hash join and sort-merge join) to expose the pitfalls and the advantages of each algorithm in the data cycling arena. The experiments show the potential of a large distributed main-memory cache glued together with RDMA into a novel distributed database architecture.

Description: A grand challenge of distributed query processing is to devise a self-organizing architecture which exploits all hardware resources optimally to manage the database hot set, minimize query response time, and maximize throughput without single point global coordination. The Data Cyclotron architecture [Goncalves and Kersten 2010] addresses this challenge using turbulent data movement through a storage ring built from distributed main memory and capitalizing on the functionality offered by modern remote-DMA network facilities. Queries assigned to individual nodes interact with the storage ring by picking up data fragments, which are continuously flowing around, that is, the hot set. The storage ring is steered by the Level Of Interest (LOI) attached to each data fragment, which represents the cumulative query interest as it passes around the ring multiple times. A fragment with LOI below a given threshold, inversely proportional to the ring load, is pulled out to free up resources. This threshold is dynamically adjusted in a fully distributed manner based on ring characteristics and locally observed query behavior. It optimizes resource utilization by keeping the average data access latency low. The approach is illustrated using an extensive and validated simulation study. The results underpin the fragment hot set management robustness in turbulent workload scenarios. A fully functional prototype of the proposed architecture has been implemented using modest extensions to MonetDB and runs within a multirack cluster equipped with Infiniband. Extensive experimentation using both microbenchmarks and high-volume workloads based on TPC-H demonstrates its feasibility. The Data Cyclotron architecture and experiments open a new vista for modern distributed database architectures with a plethora of new research challenges.

Description: In any probabilistic seismic-hazard model, the earthquake activity that cannot be associated with well-characterized fault structures is taken into account as seismicity distributed over a geographical region. Ground-motion prediction equations (GMPEs) are generally based on predictor variables describing the spatial extension of a rupture. The approach taken to model rupture finiteness can therefore bias the estimation of seismic hazard. We study the effect of rupture finiteness in modeling distributed seismicity using the OpenQuake (OQ) engine, the open-source software for seismic hazard and risk assessment promoted by the Global Earthquake Model initiative. For a simple test case we show how the inclusion of rupture finiteness, with respect to the point-rupture approximation, leads to a significant increase in the probabilities of exceedance for a given level of motion. We then compare the OQ engine with the calculation software developed by the U.S. Geological Survey-National Seismic Hazard Mapping Project. By considering a gridded seismicity model for California, we show how different approaches for modeling finite ruptures affect seismic-hazard estimates. We show how sensitivity to rupture finiteness depends not only on the spatial distribution of activity rates but also on the GMPE model. Considering two sites in Los Angeles and San Francisco, we show that for a return period of 475 years, the percent difference in the associated ground-motion levels when using point and finite ruptures ranges from 19% to 46%; for a return period of 2475 years the difference ranges from 29% to 58%.

Description: The Aegean is the most seismically active and tectonically complex region in Europe. Damaging earthquakes have occurred here throughout recorded history, often resulting in considerable loss of life. The Monte Carlo method of probabilistic seismic hazard analysis (PSHA) is used to determine the level of ground motion likely to be exceeded in a given time period. Multiple random simulations of seismicity are generated to calculate, directly, the ground motion for a given site. Within the seismic hazard analysis we explore the impact of different seismic source models, incorporating both uniform zones and distributed seismicity. A new, simplified, seismic source model, derived from seismotectonic interpretation, is presented for the Aegean region. This is combined into the epistemic uncertainty analysis alongside existing source models for the region, and models derived by a K-means cluster analysis approach. Seismic source models derived using the K-means approach offer a degree of objectivity and reproducibility into the otherwise subjective approach of delineating seismic sources using expert judgment. Similar review and analysis is undertaken for the selection of peak ground acceleration (PGA) attenuation models, incorporating into the epistemic analysis Greek-specific models, European models and a Next Generation Attenuation model. Hazard maps for PGA on a “rock” site with a 10% probability of being exceeded in 50 years are produced and different source and attenuation models are compared. These indicate that Greek-specific attenuation models, with their smaller aleatory variability terms, produce lower PGA hazard, whilst recent European models and Next Generation Attenuation (NGA) model produce similar results. The Monte Carlo method is extended further to assimilate epistemic uncertainty into the hazard calculation, thus integrating across several appropriate source and PGA attenuation models. Site condition and fault-type are also integrated into the hazard mapping calculations. These hazard maps are in general agreement with previous maps for the Aegean, recognising the highest hazard in the Ionian Islands, Gulf of Corinth and Hellenic Arc. Peak Ground Accelerations for some sites in these regions reach as high as 500–600 cm s−2 using European/NGA attenuation models, and 400–500 cm s−2 using Greek attenuation models.