Filters








166 Hits in 5.7 sec

An Efficient Abortable-locking Protocol for Multi-level NUMA Systems

Milind Chabbi, Abdelhalim Amer, Shasha Wen, Xu Liu
2017 Proceedings of the 22nd ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming - PPoPP '17  
Few locks admit a waiting thread to abandon its protocol on a timeout. State-of-the-art abortable locks are not fully locality aware, introduce high overheads, and unsuitable for frequent aborts.  ...  The popularity of Non-Uniform Memory Access (NUMA) architectures has led to numerous locality-preserving hierarchical lock designs, such as HCLH, HMCS, and cohort locks.  ...  Correctness. HMCS-T is a very involved protocol; the system is stateful.  ... 
doi:10.1145/3018743.3018768 fatcat:hldqj7uw3zcjxnelky7n5nmqui

An Efficient Abortable-locking Protocol for Multi-level NUMA Systems

Milind Chabbi, Abdelhalim Amer, Shasha Wen, Xu Liu
2017 SIGPLAN notices  
Few locks admit a waiting thread to abandon its protocol on a timeout. State-of-the-art abortable locks are not fully locality aware, introduce high overheads, and unsuitable for frequent aborts.  ...  The popularity of Non-Uniform Memory Access (NUMA) architectures has led to numerous locality-preserving hierarchical lock designs, such as HCLH, HMCS, and cohort locks.  ...  Correctness. HMCS-T is a very involved protocol; the system is stateful.  ... 
doi:10.1145/3155284.3018768 fatcat:mow7mm4ztfaxnfstxm5suonwna

Architecture of a Database System

Joseph M. Hellerstein, Michael Stonebraker, James Hamilton
2007 Foundations and Trends in Databases  
Historically, DBMSs were among the earliest multi-user server systems to be developed, and thus pioneered many systems design techniques for scalability and reliability now in use in many other contexts  ...  This paper presents an architectural discussion of DBMS design principles, including process models, parallel architecture, storage system design, transaction system implementation, query processor and  ...  Muralidhar, Pat Selinger, Mehul Shah, and Matt Welsh for background information and comments on early drafts of this paper.  ... 
doi:10.1561/1900000002 fatcat:ve6wkgbgrngppk3cjcbfsevuye

Speedy transactions in multicore in-memory databases

Stephen Tu, Wenting Zheng, Eddie Kohler, Barbara Liskov, Samuel Madden
2013 Proceedings of the Twenty-Fourth ACM Symposium on Operating Systems Principles - SOSP '13  
Silo is a new in-memory database that achieves excellent performance and scalability on modern multicore machines. Silo was designed from the ground up to use system memory and caches efficiently.  ...  Though this might seem to complicate the enforcement of a serial order, correct logging and recovery is provided by linking periodically-updated epochs with the commit protocol.  ...  Acknowledgements We thank our shepherd Timothy Roscoe and the anonymous reviewers for their feedback. Eugene Wu provided helpful comments on an earlier draft of the paper.  ... 
doi:10.1145/2517349.2522713 dblp:conf/sosp/TuZKLM13 fatcat:zhnvt3pinvfxpcirj3wbdf6lkq

Design and Analysis of a Distributed Multi-leg Stock Trading System

Jia Zou, Gong Su, Arun Iyengar, Yu Yuan, Yi Ge
2011 2011 31st International Conference on Distributed Computing Systems  
We present the design, optimization and analysis of a highly flexible and efficient multi-leg stock trading system.  ...  While the expressive power of multi-leg trading brings significant value to investors, it also poses major challenges to stock exchange architecture design, due to additional complexities introduced in  ...  Architectural Design Considerations Now we discuss various architectural design decisions and optimizations for implementing an efficient multi-threaded EV system supporting highly flexible and efficient  ... 
doi:10.1109/icdcs.2011.70 dblp:conf/icdcs/ZouSIYG11 fatcat:eig7yy5ihbhorm5k4hsgelf7vi

HaTS: Hardware-Assisted Transaction Scheduler

Zhanhao Chen, Ahmed Hassan, Masoomeh Javidi Kishi, Jacob Nelson, Roberto Palmieri, Michael Wagner
2020 International Conference on Principles of Distributed Systems  
At higher contention levels, without efficient scheduling, transactions abort each other more frequently, possibly with a domino effect that can easily lead to performance similar to, if not worse than  ...  for transactional read and write operations 1 ; the granularity of conflict detection is the cache line; and if a cache line is evicted or invalidated, the transaction is aborted (reproducing the idea  ...  The machine provides 96 physical cores and a total of 768 GB of memory divided into 4 NUMA zones.  ... 
doi:10.4230/lipics.opodis.2019.10 dblp:conf/opodis/ChenHKNP19 fatcat:3w6t72ilcbh4rlmmbmcvrc2drq

Fast Serializable Multi-Version Concurrency Control for Main-Memory Database Systems

Thomas Neumann, Tobias Mühlbauer, Alfons Kemper
2015 Proceedings of the 2015 ACM SIGMOD International Conference on Management of Data - SIGMOD '15  
Multi-Version Concurrency Control (MVCC) is a widely employed concurrency control mechanism, as it allows for execution modes where readers never block writers.  ...  The novel idea is based on an adaptation of precision locking and verifies that the (extensional) writes of recently committed transactions do not intersect with the (intensional) read predicate space  ...  If both are writes, T would have aborted due to to our protocol of immediately aborting upon detecting wwconflicts.  ... 
doi:10.1145/2723372.2749436 dblp:conf/sigmod/0001MK15 fatcat:soyptpo245gfvp4ymqxvjniss4

In-Memory Big Data Management and Processing: A Survey

Hao Zhang, Gang Chen, Beng Chin Ooi, Kian-Lee Tan, Meihui Zhang
2015 IEEE Transactions on Knowledge and Data Engineering  
We are witnessing a revolution in the design of database systems that exploits main memory as its data storage layer.  ...  In this survey, we aim to provide a thorough review of a wide range of in-memory data management and processing proposals and systems, including both data storage systems and data processing frameworks  ...  We would like to thank the anonymous reviewers, and also Bingsheng He, Eric Lo and Bogdan Marius Tudor, for their insightful comments and suggestions.  ... 
doi:10.1109/tkde.2015.2427795 fatcat:u7r3rtvhxbainfeazfduxcdwrm

Comprehensive Framework of RDMA-enabled Concurrency Control Protocols [article]

Chao Wang, Kezhao Huang, Xuehai Qian
2021 arXiv   pre-print
We focus on the correct and efficient implementation using key techniques, such as co-routines, outstanding requests, and doorbell batching, with two-sided and one-sided communication primitives.  ...  Most importantly, we obtain the execution stage latency breakdowns with one-sided and two-sided primitive for each protocol, which are analyzed to develop more efficient hybrid implementations.  ...  While the protocols themselves are known, we rethink their correct and efficient implementations in the context of RDMA.  ... 
arXiv:2002.12664v4 fatcat:2ubvm4qz4jeilbstn3dfplhk6m

Cicada

Hyeontaek Lim, Michael Kaminsky, David G. Andersen
2017 Proceedings of the 2017 ACM International Conference on Management of Data - SIGMOD '17  
To provide high performance under diverse workloads, Cicada reduces overhead and contention at several levels of the system by leveraging optimistic and multi-version concurrency control schemes and multiple  ...  On the TPC-C and YCSB benchmarks, Cicada outperforms Silo, TicToc, FOEDUS, MOCC, twophase locking, Hekaton, and ERMIA in most scenarios, achieving up to 3X higher throughput than the next fastest design  ...  Optimistic multi-version execution, an optimistic timestamp ordering scheme [34] designed for multi-core in-memory OLTP, reduces both memory access-level and transaction-level interference at records  ... 
doi:10.1145/3035918.3064015 dblp:conf/sigmod/LimKA17 fatcat:ikod5evn2bcvbmlllwih2354me

Taurus: Lightweight Parallel Logging for In-Memory Database Management Systems (Extended Version) [article]

Yu Xia, Xiangyao Yu, Andrew Pavlo, Srinivas Devadas
2020 arXiv   pre-print
Existing single-stream logging schemes are unsuitable for in-memory database management systems (DBMSs) as the single log is often a performance bottleneck.  ...  To overcome this problem, we present Taurus, an efficient parallel logging scheme that uses multiple log streams, and is compatible with both data and command logging.  ...  In the pseudo-code, the code for RLV update is protected with an atomic section for correctness. We use a lock-free design to avoid this critical section in our implementation.  ... 
arXiv:2010.06760v1 fatcat:c6hjdjjhgnckrgv72lslbxr2wq

Lock cohorting

David Dice, Virendra J. Marathe, Nir Shavit
2012 SIGPLAN notices  
This paper presents lock cohorting, a general new technique for designing NUMA-aware locks that is as simple as it is powerful.  ...  Moreover, it allows us to derive a CLH-based cohort abortable lock, the first NUMA-aware queue lock to support abortability.  ...  As we show, one can convert the BO lock (which is abortable by design) and the abortable CLH lock [14] into abortable (NUMA-aware) cohort locks, which to our knowledge, are the first set of NUMAaware  ... 
doi:10.1145/2370036.2145848 fatcat:4u65zq73f5f75m6ysva72i66f4

Lock cohorting

David Dice, Virendra J. Marathe, Nir Shavit
2012 Proceedings of the 17th ACM SIGPLAN symposium on Principles and Practice of Parallel Programming - PPoPP '12  
This paper presents lock cohorting, a general new technique for designing NUMA-aware locks that is as simple as it is powerful.  ...  Moreover, it allows us to derive a CLH-based cohort abortable lock, the first NUMA-aware queue lock to support abortability.  ...  As we show, one can convert the BO lock (which is abortable by design) and the abortable CLH lock [14] into abortable (NUMA-aware) cohort locks, which to our knowledge, are the first set of NUMAaware  ... 
doi:10.1145/2145816.2145848 dblp:conf/ppopp/DiceMS12 fatcat:o62jiu67kngrni4smmnxamjvzq

Lock Cohorting

David Dice, Virendra J. Marathe, Nir Shavit
2015 ACM Transactions on Parallel Computing  
This paper presents lock cohorting, a general new technique for designing NUMA-aware locks that is as simple as it is powerful.  ...  Moreover, it allows us to derive a CLH-based cohort abortable lock, the first NUMA-aware queue lock to support abortability.  ...  As we show, one can convert the BO lock (which is abortable by design) and the abortable CLH lock [14] into abortable (NUMA-aware) cohort locks, which to our knowledge, are the first set of NUMAaware  ... 
doi:10.1145/2686884 fatcat:o4fbwxw6ufbndlmtv66cam2izq

The Impact of Timestamp Granularity in Optimistic Concurrency Control [article]

Yihe Huang, Hao Bai, Eddie Kohler, Barbara Liskov, Liuba Shrira
2018 arXiv   pre-print
But at high contention levels, OCC is susceptible to frequent aborts, leading to wasted work and degraded performance.  ...  and transactional systems.  ...  lock design, and TicToc.  ... 
arXiv:1811.04967v1 fatcat:r3gxwv2l5bczfi7jkqc45ukmem
« Previous Showing results 1 — 15 out of 166 results