Gabatarwa zuwa tsarin madadin wal-g PostgreSQL

WAL-G kayan aiki ne mai sauƙi kuma mai inganci don tallafawa PostgreSQL zuwa gajimare. Dangane da babban aikin sa, shine magada ga mashahurin kayan aiki WAL-E, amma an sake rubutawa a cikin Go. Amma akwai sabon fasali guda ɗaya mai mahimmanci a cikin WAL-G - kwafin delta. kwafi delta WAL-G adana shafukan fayilolin da suka canza tun farkon sigar madadin. WAL-G yana aiwatar da fasahohi da yawa don daidaita ma'amala. WAL-G yayi sauri fiye da WAL-E.

Ana iya samun cikakkun bayanai na yadda wal-g ke aiki a cikin labarin: Mun overclock da madadin. Yandex lecture

Ka'idar ajiya ta S3 ta zama sananne don adana bayanai. Ɗaya daga cikin fa'idodin S3 shine ikon samun dama ta hanyar API, wanda ke ba ku damar tsara ma'amala mai sassauƙa tare da ma'ajiyar, gami da damar karanta jama'a, yayin sabunta bayanai a cikin ma'ajiyar yana faruwa ne kawai ta mutane masu izini.

Akwai aikace-aikacen ajiya na jama'a da masu zaman kansu da yawa waɗanda ke amfani da ka'idar S3. A yau za mu kalli sanannen bayani don tsara ƙananan ajiya - Minio.

Sabar PostgreSQL guda ɗaya tana da kyau don gwada wal-g, kuma ana amfani da Minio azaman maye gurbin S3.

Minio Server

Minio shigarwa

yum -y install yum-plugin-copr
yum copr enable -y lkiesow/minio
yum install -y minio

Shirya AccessKey da SecretKey a /etc/minio/minio.conf

vi /etc/minio/minio.conf

Idan ba za ku yi amfani da nginx kafin Minio ba, to kuna buƙatar canzawa

--address 127.0.0.1:9000

--address 0.0.0.0:9000

Ana ƙaddamar da Minio

systemctl start minio

Jeka shafin yanar gizon Minio http://ip-адрес-сервера-minio:9000 kuma ƙirƙirar guga (misali, pg-backups).

DB uwar garken

WAL-G a cikin rpm ni ne (Anton Patsev). Github, Fedora COPR.

Wanda ba shi da tsarin tushen RPM, yi amfani da hukuma umarni ta hanyar shigarwa.

Tare da binary wal-g, rpm ya ƙunshi rubutun da ke shigo da masu canji daga fayil /etc/wal-gd/server-s3.conf.

backup-fetch.sh
backup-list.sh
backup-push.sh
wal-fetch.sh
wal-g-run.sh
wal-push.sh

Shigar walg.

yum -y install yum-plugin-copr
yum copr enable -y antonpatsev/wal-g
yum install -y wal-g

Duba sigar wal-g.

wal-g --version
wal-g version v0.2.14

Shirya /etc/wal-gd/server-s3.conf zuwa bukatun ku.

Fayilolin daidaitawa da fayilolin bayanan da gungun bayanai ke amfani da su ana adana su a al'adance tare a cikin kundin bayanan tari, wanda aka fi sani da su. PGDATA

#!/bin/bash

export PG_VER="9.6"

export WALE_S3_PREFIX="s3://pg-backups" # бакет, который мы создали в S3
export AWS_ACCESS_KEY_ID="xxxx" # AccessKey из /etc/minio/minio.conf 
export AWS_ENDPOINT="http://ip-адрес-сервера-minio:9000"
export AWS_S3_FORCE_PATH_STYLE="true"
export AWS_SECRET_ACCESS_KEY="yyyy" # SecretKey из /etc/minio/minio.conf

export PGDATA=/var/lib/pgsql/$PG_VER/data/
export PGHOST=/var/run/postgresql/.s.PGSQL.5432 # Сокет для подключения к PostgreSQL

export WALG_UPLOAD_CONCURRENCY=2 # Кол-во потоков для закачки 
export WALG_DOWNLOAD_CONCURRENCY=2 # Кол-во потоков для скачивания
export WALG_UPLOAD_DISK_CONCURRENCY=2 # Кол-во потоков на диске для закачки
export WALG_DELTA_MAX_STEPS=7
export WALG_COMPRESSION_METHOD=brotli # Какой метод сжатия использовать.

Lokacin saita WAL-G, kuna saka WALG_DELTA_MAX_STEPS - adadin matakan da majinin delta ya fi girma daga maajiyar tushe, sa'annan ku saka tsarin kwafin delta. Ko dai kayi kwafi daga delta na ƙarshe, ko kuma ka yi delta daga ainihin cikakken madadin. Wannan yana da mahimmanci idan yanayin ɓangaren ma'auni ɗaya yana canzawa koyaushe a cikin bayananku, bayanai iri ɗaya suna canzawa koyaushe.

Shigar da bayanai.

yum install -y https://download.postgresql.org/pub/repos/yum/reporpms/EL-7-x86_64/pgdg-redhat-repo-latest.
noarch.rpm
yum install -y postgresql96 postgresql96-server mc

Mun fara da database.

/usr/pgsql-9.6/bin/postgresql96-setup initdb
Initializing database ... OK

Idan kuna gwadawa akan uwar garken 1, to kuna buƙatar sake saita ma'aunin wal_level don adanawa don PostgreSQL ƙasa da sigar 10, da kwafi don nau'in PostgreSQL 10 da sama.

wal_level = archive

Bari mu adana bayanan WAL kowane daƙiƙa 60 ta amfani da PostgreSQL kanta. A kan samfur, za ku sami wani darajar archive_timeout daban.

archive_mode = on
archive_command = '/usr/local/bin/wal-push.sh %p'
archive_timeout = 60 # Каждые 60 секунд будет выполнятся команда archive_command.

Fara PostgreSQL

systemctl start postgresql-9.6

A cikin na'ura mai ba da hanya tsakanin hanyoyin sadarwa, muna duba rajistan ayyukan PostgreSQL don kurakurai: (canza postgresql-Wed.log zuwa na yanzu).

tail -fn100 /var/lib/pgsql/9.6/data/pg_log/postgresql-Wed.log

Mu je psql.

su - postgres
psql

Ƙirƙiri bayanan bayanai a cikin psql

Ƙirƙiri tebur a cikin gwajin bayanai1.

create database test1;

Canja zuwa gwajin bayanai.

postgres=# c test1;

Muna ƙirƙirar tebur indexing_table.

test1=# CREATE TABLE indexing_table(created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW());

Ƙara bayanai.

Mun fara saka bayanai. Muna jiran minti 10-20.

#!/bin/bash
# postgres
while true; do
psql -U postgres -d test1 -c "INSERT INTO indexing_table(created_at) VALUES (CURRENT_TIMESTAMP);"
sleep 60;
done

Tabbatar yin cikakken madadin.

su - postgres
/usr/local/bin/backup-push.sh

Muna duba bayanan da ke cikin tebur a cikin gwajin bayanai1

select * from indexing_table;
2020-01-29 09:41:25.226198+
2020-01-29 09:42:25.336989+
2020-01-29 09:43:25.356069+
2020-01-29 09:44:25.37381+
2020-01-29 09:45:25.392944+
2020-01-29 09:46:25.412327+
2020-01-29 09:47:25.432564+
2020-01-29 09:48:25.451985+
2020-01-29 09:49:25.472653+
2020-01-29 09:50:25.491974+
2020-01-29 09:51:25.510178+

Kirtani shine lokacin yanzu.

Dubi jerin cikakkun bayanai

/usr/local/bin/backup-list.sh

Gwajin farfadowa

Cikakken farfadowa tare da mirgina duk WAL akwai.

Dakatar da Postgresql.

Share komai daga babban fayil /var/lib/pgsql/9.6/data.

Gudanar da rubutun /usr/local/bin/backup-fetch.sh azaman mai amfani da postgres.

su - postgres
/usr/local/bin/backup-fetch.sh

Cirar Ajiyayyen cikakke.

Ƙara recovery.conf zuwa /var/lib/pgsql/9.6/data babban fayil tare da abun ciki mai zuwa.

restore_command = '/usr/local/bin/wal-fetch.sh "%f" "%p"'

Mun fara PostgreSQL. PostgreSQL zai fara aikin dawowa daga WALs da aka adana, sannan kawai za a buɗe ma'ajin bayanai.

systemctl start postgresql-9.6
tail -fn100 /var/lib/pgsql/9.6/data/pg_log/postgresql-Wed.log

Farfadowa na wani lokaci.

Idan muna son dawo da bayanan har zuwa wani ɗan lokaci, to muna ƙara ma'aunin recovery_target_time zuwa recovery.conf - mun nuna a wane lokaci don dawo da bayanan.

restore_command = '/usr/local/bin/wal-fetch.sh "%f" "%p"'
recovery_target_time = '2020-01-29 09:46:25'

Bayan murmurewa, duba tebur indexing_table

 2020-01-29 09:41:25.226198+00
 2020-01-29 09:42:25.336989+00
 2020-01-29 09:43:25.356069+00
 2020-01-29 09:44:25.37381+00
 2020-01-29 09:45:25.392944+00

Mun fara PostgreSQL. PostgreSQL zai fara aikin dawowa daga WALs da aka adana, sannan kawai za a buɗe ma'ajin bayanai.

systemctl start postgresql-9.6
tail -fn100 /var/lib/pgsql/9.6/data/pg_log/postgresql-Wed.log

Gwaji

Samar da bayanan 1GB kamar yadda aka bayyana anan https://gist.github.com/ololobus/5b25c432f208d7eb31051a5f238dffff

Neman girman guga bayan samar da 1GB na bayanai.

postgres=# SELECT pg_size_pretty(pg_database_size('test1'));
pg_size_pretty
----------------
1003 MB

s4cmd kayan aiki ne na layin umarni kyauta don aiki tare da bayanan da ke zaune a cikin ajiyar Amazon S3. An rubuta mai amfani a cikin yaren shirye-shiryen Python, kuma saboda wannan ana iya amfani da shi a duka Windows da Linux Tsarukan aiki.

Shigar da s4cmd

pip install s4cmd

LZ4

s4cmd --endpoint-url=http://ip-адрес-сервера-minio:9000 --access-key=xxxx --secret-key=yyyy du -r s3://pg-backups
840540822       s3://pg-backups/wal_005/
840 МБ в формате lz4 только WAL логов

Полный бекап с lz4 - 1GB данных
time backup_push.sh
real 0m18.582s

Размер S3 бакета после полного бекапа

581480085       s3://pg-backups/basebackups_005/
842374424   s3://pg-backups/wal_005
581 МБ занимает полный бекап

LZMA

После генерации 1ГБ данных
338413694       s3://pg-backups/wal_005/
338 мб логов в формате lzma

Время генерации полного бекапа
time backup_push.sh
real    5m25.054s

Размер бакета в S3
270310495       s3://pg-backups/basebackups_005/
433485092   s3://pg-backups/wal_005/

270 мб занимает полный бекап в формате lzma

Brotli

После генерации 1ГБ данных
459229886       s3://pg-backups/wal_005/
459 мб логов в формате brotli

Время генерации полного бекапа
real    0m23.408s

Размер бакета в S3
312960942       s3://pg-backups/basebackups_005/
459309262   s3://pg-backups/wal_005/

312 мб занимает полный бекап в формате brotli

Kwatanta sakamako akan ginshiƙi.

Gabatarwa zuwa tsarin madadin wal-g PostgreSQL

Kamar yadda kake gani, Brotli yana kama da girmansa zuwa LZMA, amma ana yin ajiyar ajiyar a cikin lokacin LZ4.

Taɗi na al'ummar PostgreSQL masu magana da Rashanci: https://t.me/pgsql

Da fatan za a ba Github tauraro idan kuna amfani wal-g

source: www.habr.com

Add a comment