47 lines
1.7 KiB
Plaintext
47 lines
1.7 KiB
Plaintext
HISTORY
|
|
|
|
July 2004. M. Spiekermann customized the TPC data generator
|
|
|
|
|
|
NOTES:
|
|
|
|
I have only tested dbgen with the -s option, so theres no guarantee
|
|
that other options will produce proper results. Anyway, many of other
|
|
options make no sense for use with Secondo.
|
|
|
|
dbgen -s 0.1 will create approximately 100MB data
|
|
dbgen -s 1 will create 1 GB, etc. etc.
|
|
...
|
|
|
|
There is also a litte import script restoreTPC which can be executed
|
|
in SecondoTTYBDB (Precondition: A database is open) which creates
|
|
the eight relations forming the TPC Benchmark database schema.
|
|
|
|
The queries subdirectory contains the benchmark queries in standard SQL syntax.
|
|
There are some keywords included such that the program qgen can make some
|
|
transformations (to support dialects of different database vendors). However
|
|
for SECONDO we must transform them manually.
|
|
|
|
|
|
|
|
March 2012. Jiamin Lu adds the explanation of using the automatic TPC-H generator
|
|
|
|
The automatic TPC-H generator is a bash script file,
|
|
used to create a TPC-H data-set by setting a scale factor,
|
|
and load the data into a Secondo database.
|
|
|
|
The location of this script is $SECONDO_BUILD_DIR/Tools/Generators/TPC-H/secondo/tpcgen.sh
|
|
You can find the explanation about its parameters by commands:
|
|
|
|
./tpcgen.sh -h
|
|
|
|
An example of using this script is:
|
|
|
|
./tpcgen.sh -i -s 0.2 -d TPCHDB
|
|
|
|
This command first creates the TPC-H data set with the set scale factor 0.2,
|
|
then moves created data files having the suffix of "_tbl", into a temporal folder.
|
|
Afterwards, it starts a SecondoTTYNT process,
|
|
creates a Secondo database with the name that we set, TPCHDB.
|
|
At last, data are loaded into this database through a script called restore_objs,
|
|
which is kept in the same folder as this generator. |