【PostgreSQL查看SQL执行过程主机的资源使用情况】

在分析SQL的时候,可以临时设置client_min_messages=log结合如下几个参数分别查看SQL在explain, parser, execute 过程系统资源使用分析。

生产环境不建议开启,建议临时分析会话级别开启,用于定位问题。

复制代码
postgres=# select name from pg_settings where name like '%stats';
        name
---------------------
 log_executor_stats
 log_parser_stats
 log_planner_stats
 log_statement_stats
(4 rows)

一、log_executor_stats 查看执行过程资源使用

language 复制代码
postgres=# set client_min_messages =log;
SET
postgres=# set log_executor_stats=on;
SET
postgres=# select 1;
LOG:  EXECUTOR STATISTICS
DETAIL:  ! system usage stats:
!       0.000000 s user, 0.000011 s system, 0.000010 s elapsed
!       [0.000000 s user, 0.002140 s system total]
!       4216 kB max resident size
!       0/0 [0/0] filesystem blocks in/out
!       0/2 [0/915] page faults/reclaims, 0 [0] swaps
!       0 [0] signals rcvd, 0/0 [0/0] messages rcvd/sent
!       0/0 [4/0] voluntary/involuntary context switches
 ?column?
----------
        1
(1 row)

二、log_parser_stats 查看解析过程资源使用

language 复制代码
postgres=# set client_min_messages =log;
SET
postgres=# set log_parser_stats=on;
SET
postgres=# select 1;
LOG:  PARSER STATISTICS
DETAIL:  ! system usage stats:
!       0.000017 s user, 0.000017 s system, 0.000032 s elapsed
!       [0.000934 s user, 0.000934 s system total]
!       3692 kB max resident size
!       0/0 [0/0] filesystem blocks in/out
!       0/17 [0/819] page faults/reclaims, 0 [0] swaps
!       0 [0] signals rcvd, 0/0 [0/0] messages rcvd/sent
!       0/0 [3/0] voluntary/involuntary context switches
LOG:  PARSE ANALYSIS STATISTICS
DETAIL:  ! system usage stats:
!       0.000013 s user, 0.000013 s system, 0.000026 s elapsed
!       [0.000983 s user, 0.000983 s system total]
!       3948 kB max resident size
!       0/0 [0/0] filesystem blocks in/out
!       0/19 [0/854] page faults/reclaims, 0 [0] swaps
!       0 [0] signals rcvd, 0/0 [0/0] messages rcvd/sent
!       0/0 [3/0] voluntary/involuntary context switches
LOG:  REWRITER STATISTICS
DETAIL:  ! system usage stats:
!       0.000003 s user, 0.000003 s system, 0.000005 s elapsed
!       [0.000994 s user, 0.000994 s system total]
!       3948 kB max resident size
!       0/0 [0/0] filesystem blocks in/out
!       0/4 [0/858] page faults/reclaims, 0 [0] swaps
!       0 [0] signals rcvd, 0/0 [0/0] messages rcvd/sent
!       0/0 [3/0] voluntary/involuntary context switches
 ?column?
----------
        1
(1 row)

三、log_planner_stats查看PLANNER过程资源使用

language 复制代码
postgres=# set client_min_messages =log;
SET
postgres=# set log_planner_stats=on;
SET
postgres=# select 1;
LOG:  PLANNER STATISTICS
DETAIL:  ! system usage stats:
!       0.000000 s user, 0.000098 s system, 0.000097 s elapsed
!       [0.000000 s user, 0.002105 s system total]
!       3952 kB max resident size
!       0/0 [0/0] filesystem blocks in/out
!       0/53 [0/895] page faults/reclaims, 0 [0] swaps
!       0 [0] signals rcvd, 0/0 [0/0] messages rcvd/sent
!       0/0 [3/0] voluntary/involuntary context switches
 ?column?
----------
        1
(1 row)

四、log_statement_stats

language 复制代码
postgres=# set client_min_messages =log;
SET
postgres=# set log_statement_stats=on;
SET
postgres=# select 1;
LOG:  QUERY STATISTICS
DETAIL:  ! system usage stats:
!       0.000000 s user, 0.000249 s system, 0.000250 s elapsed
!       [0.000000 s user, 0.002201 s system total]
!       4216 kB max resident size
!       0/0 [0/0] filesystem blocks in/out
!       0/114 [0/916] page faults/reclaims, 0 [0] swaps
!       0 [0] signals rcvd, 0/0 [0/0] messages rcvd/sent
!       0/0 [3/0] voluntary/involuntary context switches
 ?column?
----------
        1
(1 row)
相关推荐
程序边界19 小时前
解锁时序数据新玩法:金仓数据库实战体验分享
数据库
汉堡go19 小时前
python_chapter6
前端·数据库·python
范纹杉想快点毕业19 小时前
嵌入式工程师一年制深度进阶学习计划(纯技术深耕版)
linux·运维·服务器·c语言·数据库·算法
storyseek20 小时前
关于Milvus向量数据库的基础
数据库·milvus
步步为营DotNet20 小时前
深入探究.NET中Stream:灵活高效的数据流处理核心
服务器·数据库·.net
❀͜͡傀儡师20 小时前
基于提供的镜像构建PostGIS、pgvector 的 PostgreSQL 18镜像的Dockerfile
数据库·postgresql·postgis
Mr.徐大人ゞ20 小时前
5.PG基础之索引
postgresql
消失的旧时光-194320 小时前
第五课:数据库不是存数据那么简单 —— MySQL 与索引的后端视角
数据库·mysql
nice_lcj52020 小时前
MySQL中GROUP_CONCAT函数详解 | 按日期分组拼接销售产品经典案例
数据库·mysql
key1s21 小时前
在 clickhouse时间降序排序解决方案
数据库