Help Center/ GaussDB(DWS)/ Performance White Paper/ Test Methods/ Constructing Data for TPC-H and TPC-CDS Tests/ TPC-H Data Construction
Updated on 2022-07-26 GMT+08:00
View PDF
Share

TPC-H Data Construction

  1. Obtain TPC-H tools from the Official website.
  2. Log in to the ECS and run the following command to create a directory for storing the TPC-H tools: 

    1
2
    		 
    mkdir -p /data1/script/tpch-kit/tpch1000X
mkdir -p /data2/script/tpch-kit/tpch1000X

  3. Use SFTP to upload the obtained TPC-H tools to the /data1/script/tpch-kit directory of the ECS and run the following command to decompress the tools: 

    1
    		 
    cd /data1/script/tpch-kit && unzip tpch_v3.0.0.zip

  4. Run the following command to compile and generate the data construction tool dbgen:

    Before compilation, modify the makefile.suite and tpcd.h files in the dbgen directory.

    1. Modifying makefile.suite 
       1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
      		 
      #Change the parameters of makefile.suite as follows (line 103 to line 111):

CC      = gcc
# Current values for DATABASE are: INFORMIX, DB2, TDAT (Teradata)
#                                  SQLSERVER, SYBASE, ORACLE, VECTORWISE
# Current values for MACHINE are:  ATT, DOS, HP, IBM, ICL, MVS,
#                                  SGI, SUN, U2200, VMS, LINUX, WIN32
# Current values for WORKLOAD are:  TPCH

DATABASE = POSTGRESQL # The specified parameter of the program does not contain postgresql. Modify tpcd.h to add the POSTGRESQL script.
MACHINE = LINUX
WORKLOAD = TPCH
    2. Modifying tpcd.h 
      1
2
3
4
5
6
7
8
9
      		 
      //Add the following statements to the tpcd.h file:
#ifdef POSTGRESQL
#define GEN_QUERY_PLAN  "EXPLAIN"
#define START_TRAN      "BEGIN TRANSACTION"
#define END_TRAN        "COMMIT;"
#define SET_OUTPUT      ""
#define SET_ROWCOUNT    "LIMIT %d\n"
#define SET_DBASE       ""
#endif /* POSTGRESQL */

    1
    		 
    cd TPC-H_Tools_v3.0.0/dbgen && make

  1. Log in to the ECS and run the following commands to generate data for the TPC-H 1000X test. In this example, TPC-H 1000X data is generated on two data disks synchronously.

    The total size of TPC-H 1000X data file is about 1100 GB. Make sure that the ECS disk space is sufficient.

    1. Go to the /data1/script/tpch-kit/TPC-H_Tools_v3.0.0/dbgen directory and run the following command: 
      1
      		 
      for c in {1..5};do (./dbgen -s 1000 -C 10 -S ${c} –f > /dev/null 2>&1 &);done
    2. Copy the dbgen script. 
      1
      		 
      cp -r  /data1/script/tpch-kit/TPC-H_Tools_v3.0.0/dbgen  /data2/script/tpch-kit/TPC-H_Tools_v3.0.0/dbgen

      Go to the /data2/script/tpch-kit/TPC-H_Tools_v3.0.0/dbgen directory and run the following command:

      1
      		 
      for c in {6..10};do (./dbgen -s 1000 -C 10-S ${c} –f > /dev/null 2>&1 &);done

      Where,

      • -s specifies the data scale. In this example, the value is 1000.
      • -C specifies the number of chunks. In this example, the value is 10.
      • -S specifies the sequence number of the current chunk. You do not need to change the value.

  2. Run the following commands to check the data file generation progress: You can also run the ps ux|grep dsdgen command to check whether the file generation process stops. 

    1
2
    		 
    du -sh /data1/script/tpch-kit/TPC-H_Tools_v3.0.0/dbgen/*.tbl*
du -sh /data2/script/tpch-kit/TPC-H_Tools_v3.0.0/dbgen/*.tbl*

  3. Copy data to a specified directory. 

    1
2
    		 
    mv /data1/script/tpch-kit/TPC-H_Tools_v3.0.0/dbgen/*.tbl* /data1/script/tpch-kit/tpch1000X
mv /data2/script/tpch-kit/TPC-H_Tools_v3.0.0/dbgen/*.tbl* /data2/script/tpch-kit/tpch1000X