pg_flame postgresql EXPLAIN ANALYZE 火焰圖工具

pg_flame 是golang 編寫的一個將pg的EXPLAIN ANALYZE 轉換爲火焰圖，使用簡單
如下是一個簡單的demo

環境準備

• docker-compose 文件

 

version: "3"

services:

  postgres:

    image: postgres:9.6.11

    ports:

    - "5432:5432"

    environment:

    - "POSTGRES_PASSWORD:dalong"

 

 

• 源碼編譯
  clone 代碼

 

git clone <a href="https://github.com/mgartner/pg_flame.git">https://github.com/mgartner/pg_flame.git</a>

goreleaser release --snapshot --skip-publish --rm-dist

簡單修改.goreleaser.yml支持windows

# This is an example goreleaser.yaml file with some sane defaults.

# Make sure to check the documentation at http://goreleaser.com

builds:

- env:

  - CGO_ENABLED=0

  goos:

    - linux

    - darwin

    - windows

archives:

- replacements:

    darwin: Darwin

    linux: Linux

    windows: windows

    386: i386

    amd64: x86_64

checksum:

  name_template: 'checksums.txt'

snapshot:

  name_template: "{{ .Tag }}-next"

changelog:

  sort: asc

• sql 腳本

CREATE TABLE users (

    id SERIAL PRIMARY KEY,

    userename text,

    userage integer

);

​

-- Indices -------------------------------------------------------

​

CREATE UNIQUE INDEX users_pkey ON users(id int4_ops);

​

​

​

CREATE TABLE userapps (

    id SERIAL PRIMARY KEY,

    appname text,

    appversion text,

    userid integer

);

​

-- Indices -------------------------------------------------------

​

CREATE UNIQUE INDEX userapps_pkey ON userapps(id int4_ops);

​

​

INSERT INTO "public"."userapps"("id","appname","appversion","userid")

VALUES

(1,E'login',E'v1',1),

(2,E'view',E'v2.0',2);

​

​

INSERT INTO "public"."users"("id","userename","userage")

VALUES

(1,E'dalong',11),

(2,E'app',22);

 

 

使用

• 啓動pg

docker-compose up -d

• 生成json 格式的 EXPLAIN ANALYZE

psql -U postgres -h 127.0.0.1 postgres -qAtc 'EXPLAIN (ANALYZE, BUFFERS, FORMAT JSON) select * from userapps a join users b on a.id=b.id' > plan.json

• 生成報告結果

cat plan.json | ./dist/pg_flame_darwin_amd64/pg_flame > result.html

• 效果

 

 

 

說明

pg_flame 多平臺構建使用了goreleaser，若是須要構建跨平臺的語言包，須要安裝，並且上邊數據比較少，看的不是很明顯

參考資料

https://github.com/mgartner/pg_flame