1
# This file describes how to run MySQL benchmark suite with PostgreSQL
5
# Don't run the --fast test on a PostgreSQL 7.1.1 database on
6
# which you have any critical data; During one of our test runs
7
# PostgreSQL got a corrupted database and all data was destroyed!
8
# When we tried to restart postmaster, It died with a
9
# 'no such file or directory' error and never recovered from that!
11
# Another time vacuum() filled our system disk with had 6G free
12
# while vaccuming a table of 60 M.
16
# The test was run on a Intel Xeon 2x 550 Mzh machine with 1G memory,
17
# 9G hard disk. The OS is Suse 7.1, with Linux 2.4.2 compiled with SMP
19
# Both the perl client and the database server is run
20
# on the same machine. No other cpu intensive process was used during
23
# During the test we run PostgreSQL with -o -F, not async mode (not ACID safe)
24
# because when we started postmaster without -o -F, PostgreSQL log files
25
# filled up a 9G disk until postmaster crashed.
26
# We did however notice that with -o -F, PostgreSQL was a magnitude slower
27
# than when not using -o -F.
30
# First, install postgresql-7.1.2.tar.gz
32
# Adding the following lines to your ~/.bash_profile or
33
# corresponding file. If you are using csh, use �setenv�.
35
export POSTGRES_INCLUDE=/usr/local/pg/include
36
export POSTGRES_LIB=/usr/local/pg/lib
38
PATH=$PATH:/usr/local/pg/bin
39
MANPATH=$MANPATH:/usr/local/pg/man
42
# Add the following line to /etc/ld.so.conf:
51
# untar the postgres source distribution, cd to postgresql-*
52
# and run the following commands:
54
CFLAGS=-O3 ./configure
58
mkdir /usr/local/pg/data
59
chown postgres /usr/local/pg/data
61
/usr/local/pg/bin/initdb -D /usr/local/pg/data
62
/usr/local/pg/bin/postmaster -o -F -D /usr/local/pg/data &
63
/usr/local/pg/bin/createdb test
67
# Second, install packages DBD-Pg-1.00.tar.gz and DBI-1.18.tar.gz,
68
# available from http://www.perl.com/CPAN/
70
export POSTGRES_LIB=/usr/local/pg/lib/
71
export POSTGRES_INCLUDE=/usr/local/pg/include/postgresql
77
# Now we run the test that can be found in the sql-bench directory in the
78
# MySQL 3.23 source distribution.
80
# We did run two tests:
83
run-all-tests --comment="Intel Xeon, 2x550 Mhz, 512M, pg started with -o -F" --user=postgres --server=pg --cmp=mysql
85
# When running with --fast we run the following vacuum commands on
86
# the database between each major update of the tables:
93
# The time for vacuum() is accounted for in the book-keeping() column, not
94
# in the test that updates the database.
96
run-all-tests --comment="Intel Xeon, 2x550 Mhz, 512M, pg started with -o -F" --user=postgres --server=pg --cmp=mysql --fast
98
# If you want to store the results in a output/RUN-xxx file, you should
99
# repeate the benchmark with the extra option --log --use-old-result
100
# This will create a the RUN file based of the previous results
102
run-all-tests --comment="Intel Xeon, 2x550 Mhz, 512M, pg started with -o -F" --user=postgres --server=pg --cmp=mysql --log --use-old-result
103
run-all-tests --comment="Intel Xeon, 2x550 Mhz, 512MG, pg started with -o -F" --user=postgres --server=pg --cmp=mysql --fast --log --use-old-result
105
# Between running the different tests we dropped and recreated the PostgreSQL
106
# database to ensure that PostgreSQL should get a clean start,
107
# independent of the previous runs.