Scalable and quantitative contention generation for performance evaluation on OLTP databases

doi:10.1007/s11704-022-1056-2

Frontiers of Computer Science

2023, Vol. 17

Issue (2): 172202 https://doi.org/10.1007/s11704-022-1056-2

本期目录

Scalable and quantitative contention generation for performance evaluation on OLTP databases

Chunxi ZHANG, Yuming LI, Rong ZHANG(

), Weining QIAN, Aoying ZHOU

School of Data Science and Engineering, East China Normal University, Shanghai 200062, China

全文: PDF(20135 KB) HTML

Abstract：

Massive scale of transactions with critical requirements become popular for emerging businesses, especially in E-commerce. One of the most representative applications is the promotional event running on Alibaba’s platform on some special dates, widely expected by global customers. Although we have achieved significant progress in improving the scalability of transactional database systems (OLTP), the presence of contention operations in workloads is still one of the fundamental obstacles to performance improving. The reason is that the overhead of managing conflict transactions with concurrency control mechanisms is proportional to the amount of contentions. As a consequence, generating contented workloads is urgent to evaluate performance of modern OLTP database systems. Though we have kinds of standard benchmarks which provide some ways in simulating contentions, e.g., skew distribution control of transactions, they can not control the generation of contention quantitatively; even worse, the simulation effectiveness of these methods is affected by the scale of data. So in this paper we design a scalable quantitative contention generation method with fine contention granularity control. We conduct a comprehensive set of experiments on popular opensourced DBMSs compared with the latest contention simulation method to demonstrate the effectiveness of our generation work.

Key words： high contention OLTP database performance evaluation database benchmarking

收稿日期: 2021-02-02 出版日期: 2022-08-01

Corresponding Author(s): Rong ZHANG

引用本文:

. [J]. Frontiers of Computer Science, 2023, 17(2): 172202.
Chunxi ZHANG, Yuming LI, Rong ZHANG, Weining QIAN, Aoying ZHOU. Scalable and quantitative contention generation for performance evaluation on OLTP databases. Front. Comput. Sci., 2023, 17(2): 172202.

链接本文:

https://academic.hep.com.cn/fcs/CN/10.1007/s11704-022-1056-2
https://academic.hep.com.cn/fcs/CN/Y2023/V17/I2/172202

Fig.1

Tab.1

N	The number of nodes.
$T$	The collection of threads, $\| T \|$ =M.
$T B$	The collection of tables, $\| T B \|$ =b.
$T X$	The collection of transactions, $\| T X \|$ =m.
$T X R$	Transaction ratio, $\| T X R \|$ =m. $?$ $t x r i ∈ T X R$ , $0 ≤ t x r i ≤ 1$ & $∑ 1 m t x r i = 1$ .
$C m, b$	Access status matrix between transactions $T X$ and tables $T B$ . If $t x i$ accesses $t b j$ , $c i, j = 1$ , or else $c i, j$ =0, $t x i ∈ T X$ & $t b j ∈ T B$ .
$L X m, b$	Transaction latency matrix of $T X$ on $T B$ . $L 1 i, j$ ( $L 0 i, j$ ) represents the processing latency of contention (non-contention) transaction $t x i$ on table $t b j$ , $t x i ∈ T X$ & $t b j ∈ T B$ .
$C R$	3-dimension contention ratio matrix between transactions $T X$ on $T B$ with $c r i, j, k$ as contention ratio between $t x i$ and $t x j$ on table $t b k$ , $\| C R \|$ = m $2 ?$ b & $0 ≤ c r i, j, k ≤ 1$ .
$C I$	3-dimension contention intensity matrix between transactions $T X$ on $T B$ , with $c i i, j, k$ as contention intensity between $t x i$ and $t x j$ on table $t b k$ , $\| C I \|$ =m $2 ?$ b & $0 ≤ c i i, j, k ≤ M$ .

Tab.2

$C t x, t b$	$t b 1$	$t b 2$	$t b 3$
$t x 1$	1	1	0
$t x 2$	0	1	1

Tab.3

$L 0 t x, t b$	$t b 1$	$t b 2$	$t b 3$
$t x 1$	$l 0 1, 1$	$l 0 1, 2$	?
$t x 2$	?	$l 0 2, 1$	$l 0 2, 3$

Tab.4

$L 1 t x, t b$	$t b 1$	$t b 2$	$t b 3$
$t x 1$	$l 1 1, 1$	$l 1 1, 2$	?
$t x 2$	?	$l 1 2, 2$	$l 1 2, 3$

Tab.5

$C R t x, t x, t b : C I t x, t x, t b$	$t x 1 t b 1$	$t x 2 t b 1$	$t x 1 t b 2$	$t x 2 t b 2$	$t x 1 t b 3$	$t x 2 t b 3$
$t x 1$	$0.6 : 3$	?	$0.3 : 3$	$0.4 : 4$	?	?
$t x 2$	?	?	$0.4 : 4$	$0.2 : 2$	?	$0.5 : 4$

Tab.6

$C R t x, t b : C I t x, t b$	$t b 1$	$t b 2$	$t b 3$
$t x 1$	$0.6 : 3$	$0.7 : 4$	?
$t x 2$	?	$0.6 : 3$	$0.5 : 4$

Tab.7

	$t b 1$	$t b 2$	$t b 3$
$C R t b : C I t b$	$0.6 : 3$	$0.64 : 4$	$0.5 : 4$

Tab.8

Fig.2

Name	Characteristics
Schema	TB name, TB size, key, foreign key; Attr name, Attr type, domain
Transaction	Basic Info	transID, transRatio, compileMode
	Contention	contentionGeMode, contentionGranularity
	Operation*	contentionorNot, contentionParaPos, contentionTB, [ $t x i$ , $c r$ , $c i$ ]*

Tab.9

Id	txr	oID	Workload setting	Contention	Type
$t x 0$	1	$o p 0$	update usertable set $f i e l d 1$ =? where $y c s b k e y$ =?;	1	$Y C S B$
$t x 1$	1/0.5	$o p 1$	update usertable set $f i e l d 1$ =? where $y c s b k e y$ =?;	1
$t x 1$	1/0.5	$o p 2$	update usertable set $f i e l d 1$ =? where $y c s b k e y$ =?;	0
$t x 2$	1/0.5	$o p 3$	update usertable set $f i e l d 1$ =? where $y c s b k e y$ =?;	1
$t x 2$	1/0.5	$o p 4$	select $?$ from usertable where $y c s b k e y$ =?;	0
$t x 3$	1	$o p 5$	select $s l p r i c e$ from seckillplan where $s l s k p k e y$ =?;	0	$P e a k B e n c h$
		$o p 6$	update seckillplan set $s l s k p c o u n t$ = $s l s k p c o u n t$ + 1 where $s l s k p k e y$ =? and $s l s k p c o u n t < s l p l a n c o u n t$ ;	1
		$o p 7$	replace into orders ( $o o r d e r k e y$ , $o c u s t k e y$ , $o s k p k e y$ , $o p r i c e$ , $o o r d e r d a t e$ , $o s t a t e$ ) values (?,?,?,?,?, 0);	0
		$o p 8$	replace into orderitem ( $o i o r d e r k e y$ , $o i i t e m k e y$ , $o i c o u n t$ , $o i p r i c e$ ) values (?,?, 1,?);	0

Tab.10

$T P S$	$t x 0$	$t x 1$	$t x 2$	$t x 3$
$M y S Q L$	$38256$	$32582$	$23874$	$13434$
$P G$	$95238$	$56324$	$44765$	$24865$

Tab.11

Fig.3

Fig.4

Fig.5

Fig.6

Fig.7

Fig.8

Fig.9

Fig.10

1	G, Huang X, Cheng J, Wang Y, Wang D, He T, Zhang F, Li S, Wang W, Cao Q Li. X-engine: an optimized storage engine for large-scale ecommerce transaction processing. In: Proceedings of 2019 International Conference on Management of Data. 2019, 651– 665
2	R, Appuswamy A C, Anadiotis D, Porobic M K, Iman A Ailamaki . Analyzing the impact of system architecture on the scalability of OLTP engines for high-contention workloads. Proceedings of the VLDB Endowment, 2017, 11( 2): 121– 134
3	R, Harding Aken D, Van A, Pavlo M Stonebraker . An evaluation of distributed concurrency control. Proceedings of the VLDB Endowment, 2017, 10( 5): 553– 564
4	H, Lan Z, Bao Y Peng. A survey on advancing the DBMS query optimizer: cardinality estimation, cost model, and plan enumeration. Data Science and Engineering, 2021, 6(1): 86– 101
5	Y, Ji Y, Chai X, Zhou L, Ren Y Qin. Smart intra-query fault tolerance for massive parallel processing databases. Data Science And Engineering, 2020, 5(1): 65– 79
6	X, Yu G, Bezerra A Pavlo. Staring into the abyss: an evaluation of concurrency control with one thousand cores. Proceedings of the VLDB Endowment, 2014, 8(3): 209– 220
7	F R Johnson. Scalable storage managers for the multicore era. Carnegie Mellon University, Dissertation, 2010
8	J M, Faleiro D J Abadi. FIT: a distributed database performance tradeoff. IEEE Data Engineering Bulletin, 2015, 38(1): 10– 17
9	A, Gurajada D, Gala F, Zhou A, Pathak Z F Ma . BTrim: hybrid in-memory database architecture for extreme transaction processing. Proceedings of the VLDB Endowment, 2018, 11( 12): 1889– 1901
10	K, Ren J M, Faleiro D J Abadi. Design principles for scaling multi-core OLTP under high contention. In: Proceedings of the 2016 International Conference on Management of Data. 2016, 1583– 1598
11	N, Zhou X, Zhou X, Zhang X, Du S Wang . Reordering transaction execution to boost high-frequency trading applications. Data Science and Engineering, 2017, 2( 4): 301– 315
12	S, Chen A, Ailamaki M, Athanassoulis P B, Gibbons R, Johnson I, Pandis R Stoica . TPC-E vs. TPC-C: characterizing the new TPC-E benchmark via an I/O comparison study. Acm Sigmod Record, 2011, 39( 3): 5– 10
13	B F, Cooper A, Silberstein E, Tam R, Ramakrishnan R Sears. Benchmarking cloud serving systems with YCSB. In: Proceedings of the 1st ACM Symposium on Cloud Computing. 2010. 143− 154
14	C, Zhang Y, Li R, Zhang W, Qian A Zhou. Benchmarking on intensive transaction processing. Frontiers of Computer Science, 2020, 14(5): 145204
15	P, Bodik A, Fox M J, Franklin M I, Jordan A David. Patterson: characterizing, modeling, and generating workload spikes for stateful services. In: Proceedings of the 1st ACM Symposium on Cloud Computing. 2010, 241− 252
16	A Wolski. TATP benchmark description.
17	M Stonebraker. A measure of transaction processing power. In: Stonebraker M, ed. Readings in Database Systems. 2nd ed. San Francisco: Morgan Kaufmann Publishers Inc. 1994, 442− 454
18	M J, Cahill U, Röhm A D Fekete . Fekete: serializable isolation for snapshot databases. ACM Transactions on Database Systems, 2009, 34( 4): 20
19	D E, Difallah A, Pavlo C, Curino P Cudre-Mauroux. OLTP-bench: an extensible testbed for benchmarking relational databases. Proceedings of the VLDB Endowment, 2013, 7(4): 277– 288
20	MySQL database. See Mysql.com website, 2021
21	PostgreSQL database. See Postgresql.org website, 2021

[1]

FCS-21056-OF-CZ_suppl_1

Download

Viewed

Full text

Abstract

Cited

Shared

Discussed