TPC-H quickstart demonstrates federated performance of Teiid server against 2 running Postgresql servers.
Bash (Linux host), Apache Maven, DBGEN (http://www.tpc.org/tpch/spec/tpch_2_17_0.zip), Postgresclient (psql command)
The DBGEN tool must be compiled. First rename makefile.suite to makefile and change the following settings in it
CC = gcc DATABASE= SQLSERVER MACHINE = LINUX WORKLOAD = TPCH
Then you can run
make
Generated data set Open configuration file tpch/setenv.sh and change value of DBGEN variable run
./generate-data.sh <scale>
where scale is a flot number. It is a number of gigabytes of data to generate. Two postgresql servers must be running. For each of one, run the following command to load the data
load-data-into-db.sh <hostname> <port> <dbname> <username>
The Teiid server must be running with 2 datasources configured. One data datasource per Postgresql instance. VDB config-files/tpch-vdb.xml must be reconfigured to use the two datasource and then be deployed
Open configuration file tpch/setenv.sh and change value of DRIVER_PATH, TEIID_USER, TEIID_PASSWD variables. Also make sure DBGEN variable is valid
run
./generate-queries.sh <scale>
Where scale is an integer that determines volume of query sets to generate. The final number of queries generated will be scale*22
run
./run-test.sh