postgresql pgbench

pgbench is a benchmarking tool for PostgreSQL ,pgbench是postgresql自带的一个性能基准测试工具。

可以通过pgbench --help看一下参数的相关信息

Usage:
  pgbench [OPTIONS]... [DBNAME]

Initialization options:
  -i           invokes initialization mode
  -F NUM       fill factor
  -s NUM       scaling factor

Benchmarking options:
  -c NUM       number of concurrent database clients (default: 1)
  -C           establish new connection for each transaction
  -D VARNAME=VALUE
               define variable for use by custom script
  -f FILENAME  read transaction script from FILENAME
  -j NUM       number of threads (default: 1)
  -l           write transaction times to log file
  -M {simple|extended|prepared}
               protocol for submitting queries to server (default: simple)
  -n           do not run VACUUM before tests
  -N           do not update tables "pgbench_tellers" and "pgbench_branches"
  -r           report average latency per command
  -s NUM       report this scale factor in output
  -S           perform SELECT-only transactions
  -t NUM       number of transactions each client runs (default: 10)
  -T NUM       duration of benchmark test in seconds
  -v           vacuum all four standard tables before tests

Common options:
  -d           print debugging output
  -h HOSTNAME  database server host or socket directory
  -p PORT      database server port number
  -U USERNAME  connect as specified database user
  --help       show this help, then exit
  --version    output version information, then exit

初始化：scale为10，pgbench_accounts记录有100W，

pgbench -i -s 10 pgbench -h 127.0.0.1 -p 1931 -U testuser

这里初始化会创建4张表，如下

 Schema |       Name       | Type  |  Owner   |  Size  | Description 
--------+------------------+-------+----------+--------+-------------
 public | pgbench_accounts | table | testuser | 130 MB | 
 public | pgbench_branches | table | testuser | 168 kB | 
 public | pgbench_history  | table | testuser | 960 kB | 
 public | pgbench_tellers  | table | testuser | 456 kB |

测试：100个client并发，每个有100个事务

pgbench  -c 100 -t 100 pgbench
starting vacuum...end.
transaction type: TPC-B (sort of)
scaling factor: 10
query mode: simple
number of clients: 100
number of threads: 1
number of transactions per client: 100
number of transactions actually processed: 10000/10000
tps = 114.339158 (including connections establishing)  --包含网络开销的事务数
tps = 114.643324 (excluding connections establishing)  --不包含网络开销的事务数

pgbench  -c 100 -t 100 -j 100 pgbench
starting vacuum...end.
transaction type: TPC-B (sort of)
scaling factor: 10
query mode: simple
number of clients: 100
number of threads: 100
number of transactions per client: 100
number of transactions actually processed: 10000/10000
tps = 134.473603 (including connections establishing)
tps = 134.616838 (excluding connections establishing)

这里注意-c必须是-j的倍数，也就是client是threads的倍数。

pgbench  -c 100  -j 100 -T 10 pgbench
starting vacuum...end.
transaction type: TPC-B (sort of)
scaling factor: 10
query mode: simple
number of clients: 100
number of threads: 100
duration: 10 s
number of transactions actually processed: 2144
tps = 209.617862 (including connections establishing)
tps = 211.282453 (excluding connections establishing)

注意：-t和-T不能同时使用，

脚本的内容：

\set nbranches :scale
\set ntellers 10 * :scale
\set naccounts 100000 * :scale
\setrandom aid 1 :naccounts
\setrandom bid 1 :nbranches
\setrandom tid 1 :ntellers
\setrandom delta -5000 5000

BEGIN;
UPDATE pgbench_accounts SET abalance = abalance + :delta WHERE aid = :aid;
SELECT abalance FROM pgbench_accounts WHERE aid = :aid;
UPDATE pgbench_tellers SET tbalance = tbalance + :delta WHERE tid = :tid;
UPDATE pgbench_branches SET bbalance = bbalance + :delta WHERE bid = :bid;
INSERT INTO pgbench_history (tid, bid, aid, delta, mtime) VALUES (:tid, :bid, :aid, :delta, CURRENT_TIMESTAMP);
END;

举例测试一下机器insert的速度有多快

insertsize.sql内容：create table data(filler text);

insert-size.sql内容：insert into data (filler) values (repeat('X',:scale));

$psql -d pgbench -f insertsize.sql 

$pgbench -s 100 -c 10 -t 10000 pgbench -f insert-size.sql 
starting vacuum...end.
transaction type: Custom query
scaling factor: 100
query mode: simple
number of clients: 10
number of threads: 1
number of transactions per client: 10000
number of transactions actually processed: 100000/100000
tps = 608.144907 (including connections establishing)
tps = 608.234303 (excluding connections establishing)