-
Notifications
You must be signed in to change notification settings - Fork 6
/
HOW-TO-RUN.txt
executable file
·173 lines (130 loc) · 6.23 KB
/
HOW-TO-RUN.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
Instructions for running BenchmarkSQL on PostgreSQL
---------------------------------------------------
0. Requirements
Use of JDK7 is required.
1. Create the benchmarksql user and a database
As Unix user postgres use the psql shell to connect to the postgres
database and issue the CREATE USER and CREATE DATABASE commands.
[postgres#localhost ~] $ psql postgres
psql (9.5.2)
Type "help" for help.
postgres=# CREATE USER benchmarksql WITH ENCRYPTED PASSWORD 'changeme';
postgres=# CREATE DATABASE benchmarksql OWNER benchmarksql;
postgres=# \q
[postgres#localhost ~] $
2. Compile the BenchmarkSQL source code
As your own UNIX user change into the toplevel directory of the
benchmarksql git repository checkout or the directory that was
created by unpacking the release tarball/zipfile. Use the ant
command to compile the code.
[wieck@localhost ~] $ cd benchmarksql
[wieck@localhost benchmarksql] $ ant
Buildfile: /nas1/home/wieck/benchmarksql.git/build.xml
init:
[mkdir] Created dir: /home/wieck/benchmarksql/build
compile:
[javac] Compiling 11 source files to /home/wieck/benchmarksql/build
dist:
[mkdir] Created dir: /home/wieck/benchmarksql/dist
[jar] Building jar: /home/wieck/benchmarksql/dist/BenchmarkSQL-5.0.jar
BUILD SUCCESSFUL
Total time: 1 second
[wieck@localhost benchmarksql] $
3. Create the benchmark configuration file
Change the the run directory, copy the props.pg file and edit
the copy to match your system setup and desired scaling.
[wieck@localhost benchmarksql] $ cd run
[wieck@localhost run] $ cp props.pg my_postgres.properties
[wieck@localhost run] $ vi my_postgres.properties
[wieck@localhost run] $
Note that the provided example configuration is meant to test
the functionality of your setupr. That benchmarksql can connect
to the database and execute transactions. That configuration
is NOT a benchmark run. To make it into one you need to have a
configuration that matches your database server size and
workload. Leave the sizing for now and perform a first functional
test.
The BenchmarkSQL database has an initial size of approximately
100-100MB per configured warehouse. A typical setup would be
a database of 2-5 times the physical RAM of the server.
Likewise the number of concurrent database connections (config
parameter terminals) should be something about 2-6 times the
number of CPU threads.
Last but not least benchmark runs are normally done for hours,
if not days. This is because on the database sizes above it
will take that long to reach a steady state and make sure that
all performance relevant functionality of the database, like
checkpointing and vacuuming, is included in the measurement.
So you can see that with a modern server, that has 32-256 CPU
threads and 64-512GBi, of RAM we are talking about thousands of
warehouses and hundreds of concurrent database connections.
4. Build the schema and initial database load
Execute the runDatabaseBuild.sh script with your configuration file.
[wieck@localhost run]$ ./runDatabaseBuild.sh my_postgres.properties
# ------------------------------------------------------------
# Loading SQL file ./sql.common/tableCreates.sql
# ------------------------------------------------------------
create table bmsql_config (
cfg_name varchar(30) primary key,
cfg_value varchar(50)
);
create table bmsql_warehouse (
w_id integer not null,
w_ytd decimal(12,2),
[...]
Starting BenchmarkSQL LoadData
driver=org.postgresql.Driver
conn=jdbc:postgresql://localhost:5432/benchmarksql
user=benchmarksql
password=***********
warehouses=30
loadWorkers=10
fileLocation (not defined)
csvNullValue (not defined - using default 'NULL')
Worker 000: Loading ITEM
Worker 001: Loading Warehouse 1
Worker 002: Loading Warehouse 2
Worker 003: Loading Warehouse 3
[...]
Worker 000: Loading Warehouse 30 done
Worker 008: Loading Warehouse 29 done
# ------------------------------------------------------------
# Loading SQL file ./sql.common/indexCreates.sql
# ------------------------------------------------------------
alter table bmsql_warehouse add constraint bmsql_warehouse_pkey
primary key (w_id);
alter table bmsql_district add constraint bmsql_district_pkey
primary key (d_w_id, d_id);
[...]
vacuum analyze;
[wieck@localhost run]$
5. Run the configured benchmark
[wieck@localhost run]$ ./runBenchmark.sh my_postgres.properties
The benchmark should run for the number of configured concurrent
connections (terminals) and the duration or number of transactions.
The end result of the benchmark will be reported like this:
01:58:09,081 [Thread-1] INFO jTPCC : Term-00,
01:58:09,082 [Thread-1] INFO jTPCC : Term-00, Measured tpmC (NewOrders) = 179.55
01:58:09,082 [Thread-1] INFO jTPCC : Term-00, Measured tpmTOTAL = 329.17
01:58:09,082 [Thread-1] INFO jTPCC : Term-00, Session Start = 2016-05-25 01:58:07
01:58:09,082 [Thread-1] INFO jTPCC : Term-00, Session End = 2016-05-25 01:58:09
01:58:09,082 [Thread-1] INFO jTPCC : Term-00, Transaction Count = 10
At this point you have a working setup.
6. Scale the benchmark configuration.
Change the my_postgres.properties file to the correct scaling
(number of warehouses and concurrent connections/terminals). Switch
from using a transaction count to time based:
runTxnsPerTerminal=0
runMins=180
Rebuild the database (if needed) by running
[wieck@localhost run]$ ./runDatabaseDestroy.sh my_postgres.properties
[wieck@localhost run]$ ./runDatabaseBuild.sh my_postgres.properties
Then run the benchmark again.
Rinse and repeat.
7. Result report
BenchmarkSQL collects detailed performance statistics and (if
configured) OS performance data. The example configuration file
defaults to a directory starting with my_result_.
Use the generateReport.sh DIRECTORY script to create an HTML file
with graphs. This requires R to be installed, which is beyond the
scope of this HOW-TO.