Download Latest Version BenchmarkSQL2-V2.0.tar.gz (8.3 MB)
Email in envelope

Get an email when there's a new version of benchmarksql2

Home
Name Modified Size InfoDownloads / Week
BenchmarkSQL2-V2.0.tar.gz 2016-05-25 8.3 MB
0
readme.txt 2016-05-25 2.3 kB
0
BenchmarkSQL2-V2.0-SRC.tar.gz 2016-05-25 63.4 kB
0
BenchmarkSQL2-v1.2-src.zip 2015-11-19 28.2 kB
1 1 weekly downloads
BenchmarkSQL2-v1.2.tar.gz 2015-11-19 10.1 MB
0
benchmarksql2-v1.1.zip 2015-11-05 6.3 MB
0
benchmarksql2-src-1.0.tar.gz 2015-10-23 21.7 kB
0
benchmarksql2-v1.0-release.tar.gz 2015-10-23 2.8 MB
0
Totals: 8 Items   27.7 MB 1 
Note:My english is poor, but i think you can understand it. :)

todo:
1. uniform code style.
2. implement TPC-W/DS/E/A/B/D benchmarks in the future.

2.0 fix
1. implement the TPC-H benchmark(without update stream)
2. improve performance of this program
3. use xml file as the configuration file
4. support auto-reconnect and continue testing when system under test recover from failures.
5. adjust columns position in tables, comfirming to TPC-C.
6. use log4j to log info
(for TPC-C, set log4j.appender.logfile.Threshold=trace to show error info when running performance test, default is info.
 for TPC-H, set log4j.appender.logfile.Threshold=trace to show QUERY info when running performance test, default is info.)
7. fix the bug in new-order transaction that can lead to unique constraint violation
8. fix the bug in tpmC calculation.

range for parameters in xml configuration file:
benchmark: TPC-C or TPC-H
scale:1-1000
terminals:1-1000
duration:1-100000000
xxxx_weight:0-1000

Note:for mysql-tpch, when running some queries, it may report the following error:
The total number of locks exceeds the lock table size.
you need to adjust parameters for mysql.

1.2 fix
1. support sqlserver-v2014(only sqlserver-v2014 was tested)
2. support db2-v10.5(only db2-v10.5 was tested)

1.1 fix
1 .only load 1 warehouse if not provided the number of warehouses(original value is 10).
2 .fix the bug in calculate recent tpm.
3 .support postgresql-v8.x&9.x(only postgresql-v9.4 was tested)
4 .support mysql-v5.7(only mysql-v5.7 was tested) 
5 .reduce the output of performance results.(show result every 1 second)

1.0 fix
1 .commit every 1000 records in LoadData(It's 10000 in benchmarksql)
2 .remove h_hist_id from c_history table(in tpc-c, there is no h_hist_id column in history table)
3 .change all column names into upper case, each table name begins with C_.
4 .remove benchmarksql schema from all sqls.
SELECT no_o_id FROM benchmarksql.new_order ... -> SELECT NO_O_ID FROM C_NEW_ORDER
5 .remove the retry mechanism in the new order transaction, just rollback when exception ocurrs.
6 .remove sequence hist_id_seq.
7 .provide a directory for each database, currently ,only oracle is tested(10g/11g/12c), and the scripts have been adjusted.
Source: readme.txt, updated 2016-05-25