PostgreSQL操作随记:修订间差异

来自三线的随记
无编辑摘要
无编辑摘要
 
(未显示同一用户的7个中间版本)
第12行: 第12行:


  pg_dumpall
  pg_dumpall
需要注意导出的文件格式(using <code>pg_dump -Fc</code> generates the most flexible export format and is already compressed)
需要注意导出的文件格式,pg_dump不加format参数的话默认是plain text文件,即SQL语句,这种格式是不支持在pg_restore中使用的


(using <code>pg_dump -Fc</code> generates the most flexible export format and is already compressed)
example:
db=(db1 db2 db3)
for d in ${db[*]};do echo $d;set -x;pg_dump -U postgres -h pg_host -p pg_port -Fc ${d} -C -f ${d}.sql ;set +x;done;
pg_restore -h pg_host2 -U stolon -d test -C -O < db.sql
上述的例子pg_restore使用了-O参数以实现skip restoration of object ownership,以避免在还原数据库的时候出现<code>ERROR:role "xxxx" does not exist</code>错误,实际按需调整(-x, --no-privileges参数用于skip restoration of access privileges (grant/revoke))
另外还有额外参数-c,可以在pg_restore的时候配合使用,用以在恢复时,覆盖原有数据
-c, --clean                  clean (drop) database objects before recreating
ps:  
ps:  


第34行: 第45行:


=====查看连接数=====
=====查看连接数=====
select count(1) from pg_stat_activity;<br />
select count(1) from pg_stat_activity;


=====查看最大连接数值=====
=====查看最大连接数值=====
第44行: 第55行:
=====查看连接的客户端信息=====
=====查看连接的客户端信息=====
SELECT * FROM pg_stat_activity
SELECT * FROM pg_stat_activity
<br />
select datname,pid,application_name,state from pg_stat_activity;


SELECT datname,count(datname)  FROM pg_stat_activity group by datname;
keywords: PSQLException: FATAL: sorry, too many clients already
keywords: PSQLException: FATAL: sorry, too many clients already


===== stolon: 修改连接数配置 =====
===== stolon: 扩大连接数配置 (重启keeper生效) =====
  <nowiki>stolonctl --cluster-name=kube-stolon --store-backend=kubernetes --kube-resource-kind=configmap update --patch '{"pgParameters":{"max_connections":"300"}}'</nowiki>
  <nowiki>stolonctl --cluster-name=kube-stolon --store-backend=kubernetes --kube-resource-kind=configmap update --patch '{"pgParameters":{"max_connections":"300"}}'</nowiki>
也可以在stolon proxy中执行
<nowiki>stolonctl --cluster-name=${STPROXY_CLUSTER_NAME} --store-backend=${STPROXY_STORE_BACKEND} --kube-resource-kind=${STPROXY_KUBE_RESOURCE_KIND} update --patch '{"pgParameters":{"max_connections":"300"}}'</nowiki>
stolon: 缩小连接数配置(强烈不建议)
流程大体是这样: stolon下发减少连接数操作 --> 重启Master keeper --> Standby keeper 报错,stolon 下发复原连接数操作 --> 重启Standby keeper --> stolon再次下发减少连接数操作 --> 重启Standby keeper --> 重启Master keeper


=== common commands ===
=== common commands ===

2024年8月28日 (三) 23:17的最新版本

PostgreSQL默认用户: postgres

Connect

psql -h host -U user -W

backup tools

pg_dump
PGPASSWORD="12345" pg_dump
pg_dumpall

需要注意导出的文件格式,pg_dump不加format参数的话默认是plain text文件,即SQL语句,这种格式是不支持在pg_restore中使用的

(using pg_dump -Fc generates the most flexible export format and is already compressed)

example:

db=(db1 db2 db3)
for d in ${db[*]};do echo $d;set -x;pg_dump -U postgres -h pg_host -p pg_port -Fc ${d} -C -f ${d}.sql ;set +x;done;
pg_restore -h pg_host2 -U stolon -d test -C -O < db.sql

上述的例子pg_restore使用了-O参数以实现skip restoration of object ownership,以避免在还原数据库的时候出现ERROR:role "xxxx" does not exist错误,实际按需调整(-x, --no-privileges参数用于skip restoration of access privileges (grant/revoke))

另外还有额外参数-c,可以在pg_restore的时候配合使用,用以在恢复时,覆盖原有数据

-c, --clean                  clean (drop) database objects before recreating

ps:

连接数相关

配置文件

sed  "s/max_connections = .*/max_connections = 400/" /var/lib/postgresql/data/postgresql.conf|grep max_connection
sed -i "s/max_connections = .*/max_connections = 400/" /var/lib/postgresql/data/postgresql.conf|grep max_connection

commands

临时修改连接数(待验证)

alter system set max_connections=1024

查看连接数
select count(1) from pg_stat_activity;
查看最大连接数值

show max_connections;

查看保留连接数

show superuser_reserved_connections;

查看连接的客户端信息

SELECT * FROM pg_stat_activity

select datname,pid,application_name,state from pg_stat_activity;
SELECT datname,count(datname)  FROM pg_stat_activity group by datname;

keywords: PSQLException: FATAL: sorry, too many clients already

stolon: 扩大连接数配置 (重启keeper生效)
stolonctl --cluster-name=kube-stolon --store-backend=kubernetes --kube-resource-kind=configmap update --patch '{"pgParameters":{"max_connections":"300"}}'

也可以在stolon proxy中执行

stolonctl --cluster-name=${STPROXY_CLUSTER_NAME} --store-backend=${STPROXY_STORE_BACKEND} --kube-resource-kind=${STPROXY_KUBE_RESOURCE_KIND} update --patch '{"pgParameters":{"max_connections":"300"}}'

stolon: 缩小连接数配置(强烈不建议)

流程大体是这样: stolon下发减少连接数操作 --> 重启Master keeper --> Standby keeper 报错,stolon 下发复原连接数操作 --> 重启Standby keeper --> stolon再次下发减少连接数操作 --> 重启Standby keeper --> 重启Master keeper

common commands

show databases in psql

SELECT datname FROM pg_database;

desc table in psql

postgres=# \c keel
You are now connected to database "keel" as user "postgres".
keel=# \dt
                List of relations
 Schema |         Name         | Type  |  Owner
--------+----------------------+-------+----------
 public | app_permissions      | table | postgres
 public | app_versions         | table | postgres
 public | applications         | table | postgres
(30 rows)

keel=# \dt+
                              List of relations
 Schema |         Name         | Type  |  Owner   |    Size    | Description
--------+----------------------+-------+----------+------------+-------------
 public | app_permissions      | table | postgres | 8192 bytes |
 public | app_versions         | table | postgres | 72 kB      |
 public | applications         | table | postgres | 8192 bytes |
(30 rows)

enabling expanded display (like mysql \G)

# \x
Expanded display is on.
# \x
Expanded display is off.

run this for a single command by using the \x\g\x suffix to toggle expanded display on, run the query, then toggle it off again.

select * from foo \x\g\x

Note the lack of ; in the above line, this is intentional.