75 Hits in 3.2 sec

Scheduling OLTP Transactions via Machine Learning [article]

Yangjun Sheng and Anthony Tomasic and Tieying Zhang and Andrew Pavlo
2019 arXiv   pre-print
We then empirically measure the performance impact of different scheduling algorithms on OLTP and social networking workloads.  ...  To incorporate reasoning about transaction conflicts, we develop a supervised machine learning model that estimates the probability of conflict.  ...  More precisely, any Scheduling OLTP Transactions via Machine Learning instance of attribute operator value in a WHERE clause of a SELECT, DELETE or UPDATE statement is a feature.  ... 
arXiv:1903.02990v2 fatcat:ge3vphyowrhsxbnlposoo67rg4

The case for determinism in database systems

Alexander Thomson, Daniel J. Abadi
2010 Proceedings of the VLDB Endowment  
This effectively removes all nondeterminism from typical OLTP workloads, allowing active replication with no synchronization overhead whatsoever.  ...  Further, our system eliminates the requirement for two-phase commit for any kind of distributed transaction, even across multiple nodes within the same replica.  ...  The overhead of adding the additional read transaction to learn the dependency is almost negligible.  ... 
doi:10.14778/1920841.1920855 fatcat:svy4tmqvl5arfmilhlxstjzh4i

The end of an architectural era: it's time for a complete rewrite [chapter]

Michael Stonebraker, Samuel Madden, Daniel J. Abadi, Stavros Harizopoulos, Nabil Hachem, Pat Helland
2018 Making Databases Work: the Pragmatic Wisdom of Michael Stonebraker  
The experimental evidence comes from comparing a new OLTP prototype, H-Store, which we have built at M.I.T. to a popular RDBMS on the standard transactional benchmark, TPC-C.  ...  In previous papers [SC05, SBC+07] , some of us predicted the end of "one size fits all" as a commercial relational DBMS paradigm.  ...  We predict the next fifteen years will have the same feel.  ... 
doi:10.1145/3226595.3226637 fatcat:gd2lcosm4nag7oywcv6pnpqey4

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  
aborts  ...  communication (via RDMA or library-based networking), synchronization (via transactional memory or atomic primitives) and fault-tolerance (via remote logging) [60] . Integration of Storage Service  ... 
doi:10.1109/tkde.2015.2427795 fatcat:u7r3rtvhxbainfeazfduxcdwrm

Greenplum: A Hybrid Database for Transactional and Analytical Workloads [article]

Zhenghua Lyu, Huan Hubert Zhang, Gang Xiong, Haozhou Wang, Gang Guo, Jinbao Chen, Asim Praveen, Yu Yang, Xiaoming Gao, Ashwin Agrawal, Alexandra Wang, Wen Lin (+7 others)
2021 arXiv   pre-print
Demand for enterprise data warehouse solutions to support real-time Online Transaction Processing (OLTP) queries as well as long-running Online Analytical Processing (OLAP) workloads is growing.  ...  When transactions that update data are guaranteed to reside on exactly one segment we introduce one-phase commit to speed up query processing.  ...  or aborted yet, transaction B has to wait.  ... 
arXiv:2103.11080v3 fatcat:lzb4kdxaefd3rkddejrgq4fvj4

Improving High Contention OLTP Performance via Transaction Scheduling [article]

Guna Prasaad, Alvin Cheung, Dan Suciu
2018 arXiv   pre-print
Strife executes transactions in batches, where each batch is partitioned into clusters of conflict-free transactions and a small set of residual transactions.  ...  Modern transactional systems perform poorly on workloads with transactions accessing a few highly contended data items.  ...  [20] predict and choose the optimizations (such as intelligent scheduling) that a distributed OLTP system can employ during runtime using a combination of offline machine-learning and markov models.  ... 
arXiv:1810.01997v1 fatcat:4eqvpanelfgczo22gymsaspjei


Djellel Eddine Difallah, Andrew Pavlo, Carlo Curino, Philippe Cudre-Mauroux
2013 Proceedings of the VLDB Endowment  
The key contributions of OLTP-Bench are its ease of use and extensibility, support for tight control of transaction mixtures, request rates, and access distributions over time, as well as the ability to  ...  To remedy these problems, we present OLTP-Bench, an extensible "batteries included" DBMS benchmarking testbed.  ...  For each new transaction, the Worker invokes the corresponding transaction's control code (i.e., program logic with parameterized queries) and either commits or aborts the transaction.  ... 
doi:10.14778/2732240.2732246 fatcat:dpduljjyivf6dkfmikezxwmzym

An Analysis of Concurrency Control Protocols for In-Memory Databases with CCBench (Extended Version) [article]

Takayuki Tanabe, Takashi Hoshino, Hideyuki Kawashima, Jun Nemoto, Masahiro Tanaka, Osamu Tatebe
2021 arXiv   pre-print
(I4) OCC reads the same record two or more times if a concurrent transaction interruption occurs, which can improve performance.  ...  (I6) Even a state-of-the-art garbage collection method cannot improve the performance of multi-version protocols if there is a single long transaction mixed into the workload.  ...  Scheduling OLTP Transactions via Learned Abort Prediction. In aiDM, 2019. [61] R. E. Stearns and D. J. Rosenkrantz. Distributed Database Concurrency Controls Using Before-Values.  ... 
arXiv:2009.11558v2 fatcat:276w3o4jdrdy5cs5guk4iyixaa

Architecture of a Database System

Joseph M. Hellerstein, Michael Stonebraker, James Hamilton
2007 Foundations and Trends in Databases  
It must also make sure that the thread's data and control outputs are connected via the communications manager to the client.  ...  Relational Systems: The Life of a Query 145 is established between the client and the database server directly, e.g., via the ODBC or JDBC connectivity protocol.  ...  (e.g., upon transaction commit or abort).  ... 
doi:10.1561/1900000002 fatcat:ve6wkgbgrngppk3cjcbfsevuye

Self-tuning Database Technology and Information Services [chapter]

Gerhard Weikum, Axel Moenkeberg, Christof Hasse, Peter Zabback
2002 VLDB '02: Proceedings of the 28th International Conference on Very Large Databases  
MPL limitation, admission control, and scheduling in general are solved for conventional OLTP workloads, but are still widely open for mixed workloads with both OLTP and OLAP parts and potential contention  ...  These victims were chosen based on their numbers of currently held locks and previous aborts (to prevent starvation), and only blocked transactions that blocked other transactions were considered as candidates  ... 
doi:10.1016/b978-155860869-6/50011-1 dblp:conf/vldb/WeikumMHZ02 fatcat:nxladtdxk5dlta4hy26wwjfape

Mostly-optimistic concurrency control for highly contended dynamic workloads on a thousand cores

Tianzheng Wang, Hideaki Kimura
2016 Proceedings of the VLDB Endowment  
Although pessimistic locking can protect reads, it floods cachecoherence backbones in deep memory hierarchies and can also cause numerous deadlock aborts.  ...  Larger thresholds take longer to learn new hot spots.  ...  Long Scan Workloads Combining OLAP and OLTP: Real workloads often contain both analytic accesses (e.g., a long cursor scan) and transactional updates.  ... 
doi:10.14778/3015274.3015276 fatcat:4vfbm646rrhshfr4zql6ivypye


Yi Lu, Xiangyao Yu, Samuel Madden
2019 Proceedings of the VLDB Endowment  
Our experiments on two popular benchmarks (YCSB and TPC-C) show that high availability via replication can coexist with fast serializable transaction execution in distributed in-memory databases, with  ...  In the partitioned phase, single-partition transactions are run on multiple machines in parallel to exploit more concurrency.  ...  HRDB [51] tolerates Byzantine faults among replicas by scheduling transactions with a commit barrier.  ... 
doi:10.14778/3342263.3342270 fatcat:midlq4f6vnaengjrhvm35tfsfu

Managing Query Compilation Memory Consumption to Improve DBMS Throughput [article]

Boris Baryshnikov, Cipri Clinciu, Conor Cunningham, Leo Giakoumakis, Slava Oks, Stefano Stefani
2006 arXiv   pre-print
If the compilation of a query remains blocked for an excessively long period of time, its transaction is aborted with a "timeout" error returned to the client.  ...  The individual thread scheduling choices are made by the system based on global goals that effectively prioritize earlier over later compiles when making scheduling (and thus allocation) decisions.  ... 
arXiv:cs/0612112v1 fatcat:xi4ut7sshffyplwvvtt6dbnro4

Empirical performance evaluation of concurrency and coherency control protocols for database sharing systems

Erhard Rahm
1993 ACM Transactions on Database Systems  
The nodes of a DB-sharing system are locally coupled via a high-speed interconnect and share a common database at the disk level. This is also known as a "shared disk" approach.  ...  Database Sharing (DB-sharing) refers to a general approach for building a distributed high performance transaction system.  ...  After completion of a transaction, the scheduler requests the next transaction from the reference manager.  ... 
doi:10.1145/151634.151639 fatcat:up6m3xv7cbcaba7hwppbqqbvma

Effective resource and workload management in data centers

Lei Lu, Evgenia Smirni
2014 2014 IEEE Network Operations and Management Symposium (NOMS)  
Vector Scheduling with Predictive Model Our vector scheduling algorithm consists of three steps: Given an e > 0 and a guess for the optimal maximum load vector Application Profiling: Service Demand  ...  To make the testing more I real, we select the execution mode to be "advanced transactional" in which each thread performs transactions.  ... 
doi:10.1109/noms.2014.6838287 dblp:conf/noms/LuS14 fatcat:nk2l6njvzbhkfmhg6f5ewimf6m
« Previous Showing results 1 — 15 out of 75 results