Filters








116 Hits in 5.8 sec

CedrusDB: Persistent Key-Value Store with Memory-Mapped Lazy-Trie [article]

Maofan Yin, Hongbo Zhang, Robbert van Renesse, Emin Gün Sirer
2021 arXiv   pre-print
A "lazy-trie" is a variant of the hash-trie data structure that achieves near-optimal height, has practical storage overhead, and can be maintained on-disk with standard write-ahead logging.  ...  We implemented CedrusDB, persistent key-value store based on a lazy-trie. The lazy-trie is kept on disk while made available in memory using standard memory-mapping.  ...  LSM-trie [50] uses a static hash-trie merge structure that keeps reorganizing data for more efficient compaction. SuRF [51] uses an LSM design with a trie-based filter to optimize range queries.  ... 
arXiv:2005.13762v3 fatcat:wkbsap2jbjf67eyesbc6zis4du

Cache-Aware Lock-Free Concurrent Hash Tries [article]

Aleksandar Prokopec, Phil Bagwell, Martin Odersky
2017 arXiv   pre-print
Remove operations ensure that the unneeded memory is freed and that the trie is kept compact.  ...  A pseudocode for these operations is presented and a proof of correctness is given -- we show that the implementation is linearizable and lock-free.  ...  Universal methodologies for constructing lock-free data structures exist [9] , but they serve as a theoretical foundation and are in general too inefficient to be practical -developing efficient lock-free  ... 
arXiv:1709.06056v1 fatcat:qyudvjylwzds3akqfhmmowjs3y

Reducing the Storage Overhead of Main-Memory OLTP Databases with Hybrid Indexes

Huanchen Zhang, David G. Andersen, Andrew Pavlo, Michael Kaminsky, Lin Ma, Rui Shen
2016 Proceedings of the 2016 International Conference on Management of Data - SIGMOD '16  
The index periodically migrates entries from the first stage to the second, which uses a more compact, read-optimized data structure.  ...  To reduce this overhead, we propose using a two-stage index: The first stage ingests all incoming entries and is kept small for fast read and write operations.  ...  National Science Foundation under awards IIS-1409802, CNS-1345305, and CCF-1535821, as well as Intel via the Intel Science and Technology Center for Cloud Computing (ISTC-CC).  ... 
doi:10.1145/2882903.2915222 dblp:conf/sigmod/ZhangAPKMS16 fatcat:rh4qexnquvfolddctr3fqory44

KenLM: Faster and Smaller Language Model Queries

Kenneth Heafield
2011 Conference on Machine Translation  
We present KenLM, a library that implements two data structures for efficient language model queries, reducing both time and memory costs.  ...  The TRIE data structure is a trie with bit-level packing, sorted records, interpolation search, and optional quantization aimed at lower memory consumption.  ...  This material is based upon work supported by the National Science Foundation Graduate Research Fellowship under Grant No. 0750271 and by the DARPA GALE program.  ... 
dblp:conf/wmt/Heafield11 fatcat:gskdwespmbcpvdru6mgzdghhbq

Concurrent tries with efficient non-blocking snapshots

Aleksandar Prokopec, Nathan Grasso Bronson, Phil Bagwell, Martin Odersky
2012 Proceedings of the 17th ACM SIGPLAN symposium on Principles and Practice of Parallel Programming - PPoPP '12  
To ensure spaceefficiency, removal operations compress the trie when necessary. We show how to implement an efficient lock-free snapshot operation for concurrent hash tries.  ...  The hash trie supports standard mutable lock-free operations such as insertion, removal, lookup and their conditional variants.  ...  Doug Lea at the State University of New York at Oswego for his help. We would also like to thank Eva Darulova and Tihomir Gvero at the EPFL for the useful discussions we've lead.  ... 
doi:10.1145/2145816.2145836 dblp:conf/ppopp/ProkopecBBO12 fatcat:qgifwoagj5fgrndkdsb6e6ham4

Height Optimized Tries

Robert Binna, Eva Zangerle, Martin Pichl, Günther Specht, Viktor Leis
2022 ACM Transactions on Database Systems  
We present the Height Optimized Trie (HOT), a fast and space-efficient in-memory index structure.  ...  The core algorithmic idea of HOT is to dynamically vary the number of bits considered at each node, which enables a consistently high fanout and thereby good cache efficiency.  ...  Therefore, different approaches based on the concept of lock-free index structures or write-only minimal locks have been proposed [21, [26] [27] [28] Lock-free data structures often use a single compare-and-swap  ... 
doi:10.1145/3506692 fatcat:fxlo2iqanbbbvop7l2qpb5f6ra

Recipe

Se Kwon Lee, Jayashree Mohan, Sanidhya Kashyap, Taesoo Kim, Vijay Chidambaram
2019 Proceedings of the 27th ACM Symposium on Operating Systems Principles - SOSP '19  
For example, we built P-CLHT, our PM implementation of the CLHT hash table by modifying only 30 LOC.  ...  We present a set of conditions that enable the identification of this class of DRAM indexes, and the actions to be taken to convert each index to be persistent.  ...  Acknowledgments We would like to thank our shepherd, Peter Chen, the anonymous reviewers, and members of Systems and Storage Lab and LASR group for their feedback and guidance. We are also  ... 
doi:10.1145/3341301.3359635 dblp:conf/sosp/LeeMKKC19 fatcat:gxbpwvv7tnecfaorv7wyv7tx2i

An Efficient and Scalable Memory Allocator for Multithreaded Tabled Evaluation of Logic Programs

Miguel Areias, Ricardo Rocha
2012 2012 IEEE 18th International Conference on Parallel and Distributed Systems  
For that, we propose a memory allocator based on local and global pages, to split memory among specific data structures and different threads, together with a strategy where data structures of the same  ...  number of data structures in the table space.  ...  ACKNOWLEDGMENTS This work is partially funded by the ERDF (European  ... 
doi:10.1109/icpads.2012.91 dblp:conf/icpads/AreiasR12 fatcat:ibiaunniyrguzj2goromdsfydi

The myrmics memory allocator

Spyros Lyberis, Polyvios Pratikakis, Dimitrios S. Nikolopoulos, Martin Schulz, Todd Gamblin, Bronis R. de Supinski
2012 Proceedings of the 2012 international symposium on Memory Management - ISMM '12  
Productivity is improved through the use of dynamic regions in a global address space, which provide a convenient shared memory abstraction for dynamic and irregular data structures.  ...  One approach to harness the massive parallelism is to move to task-based programming models that rely on runtime systems for dependency analysis and scheduling.  ...  Accordingly, the U.S. Government retains and the publisher, by accepting the article for publication, acknowledges that the U.S.  ... 
doi:10.1145/2258996.2259001 dblp:conf/iwmm/LyberisPNSGS12 fatcat:yw4hleeq65crrbe3ndnnoyvlou

The myrmics memory allocator

Spyros Lyberis, Polyvios Pratikakis, Dimitrios S. Nikolopoulos, Martin Schulz, Todd Gamblin, Bronis R. de Supinski
2013 SIGPLAN notices  
Productivity is improved through the use of dynamic regions in a global address space, which provide a convenient shared memory abstraction for dynamic and irregular data structures.  ...  One approach to harness the massive parallelism is to move to task-based programming models that rely on runtime systems for dependency analysis and scheduling.  ...  Accordingly, the U.S. Government retains and the publisher, by accepting the article for publication, acknowledges that the U.S.  ... 
doi:10.1145/2426642.2259001 fatcat:ph3iin5a75fn7dvrxhq75vp6de

Analysis and Optimization of Persistent Memory Index Structures' Write Amplification

Youngjoo Woo, Taesoo Kim, Sungin Jung, Euiseong Seo
2021 IEEE Access  
For these reasons, PM’s write a new bucket for the key to be added, writes the value and amplification is heavily affected by data structures and data key there  ...  on different cache lines for record 0, 8, 16, and 24. This In the workload Load, since P-HOT performs the trie results in additional read and write operations.  ... 
doi:10.1109/access.2021.3136459 fatcat:w5spdozzkfbztj5qtm4laegm4y

The adaptive radix tree: ARTful indexing for main-memory databases

V. Leis, Alfons Kemper, T. Neumann
2013 2013 IEEE 29th International Conference on Data Engineering (ICDE)  
At the same time, ART is very space efficient and solves the problem of excessive worst-case space consumption, which plagues most radix trees, by adaptively choosing compact and efficient data structures  ...  Traditional in-memory data structures like balanced binary search trees are not efficient on modern hardware, because they do not optimally utilize on-CPU caches.  ...  Internally, we use four data structures with different capacities. Given the next key byte, each data structure allows to efficiently find, add, and remove a child node.  ... 
doi:10.1109/icde.2013.6544812 dblp:conf/icde/LeisK013 fatcat:sma67ydio5c5vpuok4axiruanu

LOOP: Layer-based overlay and optimized polymerization for multiple virtual tables

Zhian Mi, Tong Yang, Jianyuan Lu, Hao Wu, Yi Wang, Tian Pan, Haoyu Song, Bin Liu
2013 2013 21st IEEE International Conference on Network Protocols (ICNP)  
Experiments on the generalpurpose CPU show that LOOP achieves efficient storage, fast lookup, and fast incremental update. It compacts 18 FIBs with about 7M prefixes in total to only 4.6MB.  ...  One of the key advantages of LOOP is that it supports inserting and deleting virtual routers incrementally so it is ideal for fast and dynamic configuration of virtual networks.  ...  Inspired by Lulea algorithm [6] , LOOP adopts bitmap technique for data structure compaction.  ... 
doi:10.1109/icnp.2013.6733611 dblp:conf/icnp/MiYLWWPSL13 fatcat:y53ng5e4efdd5hixxpkyc2x2wa

On fast address-lookup algorithms

H.H.-Y. Tzeng, T. Przygienda
1999 IEEE Journal on Selected Areas in Communications  
The growth of the Internet and its acceptance has sparkled keen interest in the research community in respect to many apparent scaling problems for a large infrastructure based on IP technology.  ...  A self-contained problem of considerable practical and theoretical interest is the longest-prefix lookup operation, perceived as one of the decisive bottlenecks.  ...  ACKNOWLEDGMENT The authors would like to acknowledge the valuable comments from anonymous reviewers that improved the depth and structure of this paper.  ... 
doi:10.1109/49.772436 fatcat:ydivubanarbmri4mlfsmcejaim

Evaluating Persistent Memory Range Indexes: Part Two [Extended Version] [article]

Yuliang He, Duo Lu, Kaisong Huang, Tianzheng Wang
2022 arXiv   pre-print
Scalable persistent memory (PM) has opened up new opportunities for building indexes that operate and persist data directly on the memory bus, potentially enabling instant recovery, low latency and high  ...  This paper addresses these issues by analyzing and experimentally evaluating state-of-the-art PM range indexes built for real PM.  ...  ACKNOWLEDGMENTS We thank the anonymous reviewers and associate editor for their constructive feedback.  ... 
arXiv:2201.13047v2 fatcat:a3n67sioifcx5cej7lz33tnip4
« Previous Showing results 1 — 15 out of 116 results