SnappyData VS Tabula

This repository contains an experimental comparison between SnappyData and Tabula.

The experiment studies the performance of Synopsis Data Engine of SnappyData 1.0.2.1 Enterprise version. OLAP column-stored is used here.

• Example SQL scripts are provided in /experiment/ folder.
• All other files except the /experiment/ folder are the original enterprise version binary files from SnappyData offical website. Please respect SnappyData's copyright.

Connect to SnappyData Client

Locate the following information of your SnappyData cluster:

client-bind-address: en4119507l.cidse.dhcp.asu.edu

default port: 1527

Create NYCtaxi

./bin/snappy run -file=./experiment/create-nyc.sql -param:input=../../experiment/testdata-nyc.csv --client-bind-address=en4119507l.cidse.dhcp.asu.edu

Change the input path to a proper path

Create sample table

./bin/snappy run -file=./experiment/createsample-nyc.sql -param:frac=0.1 --client-bind-address=en4119507l.cidse.dhcp.asu.edu

Change the frac to a proper number

Query

./bin/snappy run -file=./experiment/query-nyc.sql --client-bind-address=en4119507l.cidse.dhcp.asu.edu

Drop sample

./bin/snappy run -file=./experiment/dropsample-nyc.sql --client-bind-address=en4119507l.cidse.dhcp.asu.edu

Drop all

./bin/snappy run -file=./experiment/dropall-nyc.sql --client-bind-address=en4119507l.cidse.dhcp.asu.edu