TY - GEN
T1 - Accelerating hybrid transactional/analytical processing using consistent dual-snapshot
AU - Li, Liang
AU - Wu, Gang
AU - Wang, Guoren
AU - Yuan, Ye
N1 - Publisher Copyright:
© Springer Nature Switzerland AG 2019.
PY - 2019
Y1 - 2019
N2 - To efficiently deal with OLTP and OLAP workload simultaneously, the conventional method is to deploy two separate processing systems, i.e., the transaction-friendly designed OLTP system and the analytic-friendly optimized OLAP system. To maintain the freshness of the data, extra ETL tools are required to propagate data from the OLTP to the OLAP system. However, low-speed ETL processing is the bottleneck in those business decision support systems. As a result, there has been tremendous interest in developing hybrid transactional/analytical processing (HTAP) systems. This paper proposes a wait-free HTAP (WHTAP) architecture, that can perform both OLTP and OLAP requests efficiently in a wait-free form. We develop and evaluate a prototype WHTAP system. Our experiments show that the system can obtain a similar OLTP performance as the TicToc system and a four to six times acceleration in analytical processing at the same time.
AB - To efficiently deal with OLTP and OLAP workload simultaneously, the conventional method is to deploy two separate processing systems, i.e., the transaction-friendly designed OLTP system and the analytic-friendly optimized OLAP system. To maintain the freshness of the data, extra ETL tools are required to propagate data from the OLTP to the OLAP system. However, low-speed ETL processing is the bottleneck in those business decision support systems. As a result, there has been tremendous interest in developing hybrid transactional/analytical processing (HTAP) systems. This paper proposes a wait-free HTAP (WHTAP) architecture, that can perform both OLTP and OLAP requests efficiently in a wait-free form. We develop and evaluate a prototype WHTAP system. Our experiments show that the system can obtain a similar OLTP performance as the TicToc system and a four to six times acceleration in analytical processing at the same time.
KW - HTAP
KW - In-memory database systems
KW - Transaction concurrency control
UR - https://www.scopus.com/pages/publications/85065552921
U2 - 10.1007/978-3-030-18576-3_4
DO - 10.1007/978-3-030-18576-3_4
M3 - Conference contribution
AN - SCOPUS:85065552921
SN - 9783030185756
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 52
EP - 69
BT - Database Systems for Advanced Applications - 24th International Conference, DASFAA 2019, Proceedings
A2 - Li, Guoliang
A2 - Yang, Jun
A2 - Natwichai, Juggapong
A2 - Tong, Yongxin
A2 - Gama, Joao
PB - Springer Verlag
T2 - 24th International Conference on Database Systems for Advanced Applications, DASFAA 2019
Y2 - 22 April 2019 through 25 April 2019
ER -