TPC-H benchmark tools Microsoft SQL Server Linux

This repository facilitates the use of the TPC-H benchmark.

It's posible to run Query in :

• single mode : 1 Query run sequentially or simulate multiple users who execute the same query in parallel
• power mode : 22 queries run sequentially, in single stream.
• throughput mode : X amount of streams, each running the 22 queries in parallel

Get the results and calculating the QppH

Requirements

• The Bourne Again Shell - bash
• various typical Unix-shell command-line tools: unzip, sed, echo and so on
• Microsoft SQL Server instance up
• Database TPC-H create database and loaded data

Getting started

Edit script RunQuery.sh and initialize the following variables :

LOGIN=[ login user database ]
PASS=[ Password ] 
BASE=[ Database Name ]
HOSTNAME=[ Hostname ]

Run Workload : Query in single mode , power mode or throughput mode

./RunQuery.sh [ Run Name ] [ Number iteration ] [ Query Number ] 

[ Run Name ] : String name of run
[ Number iteration ] : Number de run in parallel
[ Query Number ] : 1 to 22 - 0 for all Query in synchronous mode

Show Results

./ShowResult.sh [ Query Number ] [ Mode ]

[ Query Number ] : 1 to 22 - 0 for all Query in synchronous mode
[ Mode ] : 0 : Power - 1 : Throughput - 2 : Single

This script generate 1 csv file per mode :

Power      : Results_power.txt
Throughput : Results_throughput.txt
Single     : Results_single.txt

for calculating the QppH use the file TPC-H-Calculator.xls