AI 助理
备案控制台
开发者社区 数据库 文章 正文

We can ignore the performance influence when use sync replication in PostgreSQL 9.1

2016-03-30 1201
版权
版权声明：
本文内容由阿里云实名注册用户自发贡献，版权归原作者所有，阿里云开发者社区不拥有其著作权，亦不承担相应法律责任。具体规则请查看《 阿里云开发者社区用户服务协议》和 《阿里云开发者社区知识产权保护指引》。如果您发现本社区中有涉嫌抄袭的内容，填写 侵权投诉表单进行举报，一经查实，本社区将立刻删除涉嫌侵权内容。
本文涉及的产品
云原生数据库 PolarDB PostgreSQL 版，标准版 2核4GB 50GB
云原生数据库 PolarDB MySQL 版，通用型 2核4GB 50GB
简介：
PostgreSQL 9.1 提供了同步复制的能力，但是我一直担心同步复制必将带来较大的性能影响。
下面从单节点，多个异步复制节点，多个复制节点（含同步复制节点）；以及同步事务，本地事务，异步事务这几个方面来测试一下同步复制带来的性能影响。

下面是测试过程和测试结果。
首先是测试环境如下 : 
服务器配置 : 
主库:   8核, 24G, 1块本地SAS10K转硬盘, 1块1GB网卡, x86服务器
standby库1:  8核, 8G,  1块本地SAS10K转硬盘, 1块1GB网卡, x86服务器
standby库2:  8核, 14G, 1块本地SAS10K转硬盘, 1块1GB网卡, x86服务器

环境搭建可以使用pg_basebackup, 这个备份工具也是9.1提供的, 利用流复制协议远程备份数据库的命令. 当然也可以使用传统的pg_start_backup 然后SCP等命令。
对pg_basebackup不熟悉的朋友可以参考我以前写的BLOG
或者参考man pg_basebackup. 

环境搭建好后就开始建立测试表和测试数据了 : 
测试表 : 
create table user_info
(userid int,
engname text,
cnname text,
occupation text,
birthday date,
signname text,
email text,
qq numeric,
crt_time timestamp without time zone,
mod_time timestamp without time zone
);

create table user_login_rec
(userid int,
login_time timestamp without time zone,
ip inet
);

create table user_logout_rec
(userid int,
logout_time timestamp without time zone,
ip inet
);


测试数据 : 
我这里弄了5000W测试数据进去，当然你像测试多点数据的话可以再多塞点数据进去。 
insert into user_info (userid,engname,cnname,occupation,birthday,signname,email,qq,crt_time,mod_time)
select generate_series(1,50000000),
'digoal.zhou',
'德哥',
'DBA',
'1970-01-01'
,E'公益是一辈子的事, I\'m Digoal.Zhou, Just do it!',
'digoal@126.com',
276732431,
clock_timestamp(),
NULL
;

set work_mem='2048MB';
set maintenance_work_mem='2048MB';
alter table user_info add constraint pk_user_info primary key (userid) using index tablespace digoal_idx;

建立PK。

接下来新建数据库操作函数，我这里测试的是模拟用户登录，退出。涉及的SQL是查询用户信息，写登录日志表和退出日志表。
登录函数 : 
create or replace function f_user_login 
(i_userid int,
OUT o_userid int,
OUT o_engname text,
OUT o_cnname text,
OUT o_occupation text,
OUT o_birthday date,
OUT o_signname text,
OUT o_email text,
OUT o_qq numeric
)
as $BODY$
declare
begin
select userid,engname,cnname,occupation,birthday,signname,email,qq
into o_userid,o_engname,o_cnname,o_occupation,o_birthday,o_signname,o_email,o_qq
from user_info where userid=i_userid;
insert into user_login_rec (userid,login_time,ip) values (i_userid,now(),inet_client_addr());
return;
end;
$BODY$
language plpgsql;


退出函数 :
 
create or replace function f_user_logout
(i_userid int,
OUT o_result int
)
as $BODY$
declare
begin
insert into user_logout_rec (userid,logout_time,ip) values (i_userid,now(),inet_client_addr());
o_result := 0;
return;
exception 
when others then
o_result := 1;
return;
end;
$BODY$
language plpgsql;


接下来新建pgbouncer连接池，用于测试单场景（开启16个连接到数据库, 允许1W个客户端连接），另一个用来测试复合场景（分别开启12个连接到数据库, 允许2W个客户端连接）。

然后编辑pgbench需要的SQL文件 : 
pgbench测试脚本1 f_user_login.sql : 
\setrandom userid 1 50000000
select * from f_user_login(:userid);


pgbench测试脚本2 f_user_logout.sql : 
 
\setrandom userid 1 50000000
select * from f_user_logout(:userid);


测试结果 : 
一、未开启任何standby的测试(异步事务) : 
单项测试结果 : 
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_login.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 1999 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 4451365
tps = 24729.367434 (including connections establishing)
tps = 24729.796402 (excluding connections establishing)
statement latencies in milliseconds:
        0.002033        \setrandom userid 1 50000000
        1.289057        select * from f_user_login(:userid);
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_logout.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 1999 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 5022561
tps = 27902.772377 (including connections establishing)
tps = 27903.700179 (excluding connections establishing)
statement latencies in milliseconds:
        0.002034        \setrandom userid 1 50000000
        1.141897        select * from f_user_logout(:userid);


复合测试结果 : 
 
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_login.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 1999 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 2374574
tps = 13191.822223 (including connections establishing)
tps = 13191.979321 (excluding connections establishing)
statement latencies in milliseconds:
        0.002205        \setrandom userid 1 50000000
        2.420410        select * from f_user_login(:userid);
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_logout.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 2000 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 2864446
tps = 15913.315357 (including connections establishing)
tps = 15913.774923 (excluding connections establishing)
statement latencies in milliseconds:
        0.002180        \setrandom userid 1 50000000
        2.005673        select * from f_user_logout(:userid);




二、开启2台异步standby测试(异步事务) : 
postgres=# select * from pg_stat_replication;
 procpid | usesysid | usename | application_name | client_addr | client_hostname | client_port |         backend_start         |   s
tate   | sent_location | write_location | flush_location | replay_location | sync_priority | sync_state 
---------+----------+---------+------------------+-------------+-----------------+-------------+-------------------------------+----
-------+---------------+----------------+----------------+-----------------+---------------+------------
    2780 |    16464 | replica | standby1         | 172.xxx.xxx.1 |                 |       26038 | 2011-10-22 00:09:48.067483+08 | str
eaming | 4/90163068    | 4/90163068     | 4/90163068     | 4/90163068      |             0 | async
    2781 |    16464 | replica | standby2         | 172.xxx.xxx.2 |                 |       59262 | 2011-10-22 00:09:53.633187+08 | str
eaming | 4/90163068    | 4/90163068     | 4/90163068     | 4/90163068      |             0 | async


单项测试结果 : 
 
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_login.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 1999 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 4216608
tps = 23425.184203 (including connections establishing)
tps = 23425.862496 (excluding connections establishing)
statement latencies in milliseconds:
        0.002019        \setrandom userid 1 50000000
        1.361130        select * from f_user_login(:userid);
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_logout.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 1999 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 4769505
tps = 26496.706965 (including connections establishing)
tps = 26497.351435 (excluding connections establishing)
statement latencies in milliseconds:
        0.002039        \setrandom userid 1 50000000
        1.202742        select * from f_user_logout(:userid);


复合测试结果 : 
 
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_login.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 1999 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 2325629
tps = 12919.857136 (including connections establishing)
tps = 12920.135410 (excluding connections establishing)
statement latencies in milliseconds:
        0.002146        \setrandom userid 1 50000000
        2.471710        select * from f_user_login(:userid);
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_logout.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 2000 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 2817671
tps = 15634.878108 (including connections establishing)
tps = 15635.203940 (excluding connections establishing)
statement latencies in milliseconds:
        0.002186        \setrandom userid 1 50000000
        2.041436        select * from f_user_logout(:userid);



三、开启1台同步standby, 1台异步standby测试 : 
这里测了三种情况，一种是同步事务synchronous_commit=on，一种是本地事务synchronous_commit=local，一种是异步事务synchronous_commit=off。
3.1 异步事务测试结果 : 
postgres=# select * from pg_stat_replication;
 procpid | usesysid | usename | application_name | client_addr | client_hostname | client_port |         backend_start         |   s
tate   | sent_location | write_location | flush_location | replay_location | sync_priority | sync_state 
---------+----------+---------+------------------+-------------+-----------------+-------------+-------------------------------+----
-------+---------------+----------------+----------------+-----------------+---------------+------------
    2780 |    16464 | replica | standby1         | 172.xxx.xxx.1 |                 |       26038 | 2011-10-22 00:09:48.067483+08 | str
eaming | 4/901630F8    | 4/901630F8     | 4/901630F8     | 4/901630F8      |             1 | sync
    2781 |    16464 | replica | standby2         | 172.xxx.xxx.2 |                 |       59262 | 2011-10-22 00:09:53.633187+08 | str
eaming | 4/901630F8    | 4/901630F8     | 4/901630F8     | 4/901630F8      |             2 | potential


3.1.1 单项测试结果 : 
 
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_login.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 1999 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 4159570
tps = 23108.282523 (including connections establishing)
tps = 23108.654858 (excluding connections establishing)
statement latencies in milliseconds:
        0.002064        \setrandom userid 1 50000000
        1.379831        select * from f_user_login(:userid);
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_logout.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 1999 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 4961886
tps = 27565.527965 (including connections establishing)
tps = 27566.353175 (excluding connections establishing)
statement latencies in milliseconds:
        0.002029        \setrandom userid 1 50000000
        1.155964        select * from f_user_logout(:userid);


3.1.2 复合测试结果 :
 
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_login.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 1999 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 2247088
tps = 12483.605632 (including connections establishing)
tps = 12483.992644 (excluding connections establishing)
statement latencies in milliseconds:
        0.002179        \setrandom userid 1 50000000
        2.558105        select * from f_user_login(:userid);
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_logout.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 2000 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 2705016
tps = 15027.514772 (including connections establishing)
tps = 15027.784844 (excluding connections establishing)
statement latencies in milliseconds:
        0.002145        \setrandom userid 1 50000000
        2.124289        select * from f_user_logout(:userid);


3.2 本地事务测试结果
3.2.1 单项测试结果 : 
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_login.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 1999 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 1880394
tps = 10446.398754 (including connections establishing)
tps = 10446.716731 (excluding connections establishing)
statement latencies in milliseconds:
        0.002059        \setrandom userid 1 50000000
        3.058239        select * from f_user_login(:userid);
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_logout.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 1999 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 2283510
tps = 12685.954741 (including connections establishing)
tps = 12686.191268 (excluding connections establishing)
statement latencies in milliseconds:
        0.002034        \setrandom userid 1 50000000
        2.517571        select * from f_user_logout(:userid);


3.2.2 复合测试结果 : 
 
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_login.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 1999 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 1010209
tps = 5611.969675 (including connections establishing)
tps = 5612.109360 (excluding connections establishing)
statement latencies in milliseconds:
        0.002196        \setrandom userid 1 50000000
        5.696539        select * from f_user_login(:userid);
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_logout.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 2000 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 1043739
tps = 5798.298837 (including connections establishing)
tps = 5798.383639 (excluding connections establishing)
statement latencies in milliseconds:
        0.002193        \setrandom userid 1 50000000
        5.513390        select * from f_user_logout(:userid);


3.3 同步事务测试结果 : 
3.3.1 单项测试结果 : 
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_login.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 1999 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 1778910
tps = 9882.600708 (including connections establishing)
tps = 9882.886514 (excluding connections establishing)
statement latencies in milliseconds:
        0.002006        \setrandom userid 1 50000000
        3.233093        select * from f_user_login(:userid);
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_logout.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 1999 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 2202408
tps = 12235.373102 (including connections establishing)
tps = 12235.684730 (excluding connections establishing)
statement latencies in milliseconds:
        0.002018        \setrandom userid 1 50000000
        2.610460        select * from f_user_logout(:userid);


3.3.2 复合测试结果 : 
 
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_login.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 1999 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 945110
tps = 5201.559767 (including connections establishing)
tps = 5201.652220 (excluding connections establishing)
statement latencies in milliseconds:
        0.002197        \setrandom userid 1 50000000
        6.122413        select * from f_user_login(:userid);
postgres@db5-> pgbench -c 32 -M extended -f ./f_user_logout.sql -j 32 -n -r -T 180 -h 127.0.0.1 -p 2000 -U digoal digoal
transaction type: Custom query
scaling factor: 1
query mode: extended
number of clients: 32
number of threads: 32
duration: 180 s
number of transactions actually processed: 1007326
tps = 5546.141117 (including connections establishing)
tps = 5546.243610 (excluding connections establishing)
statement latencies in milliseconds:
        0.002194        \setrandom userid 1 50000000
        5.735635        select * from f_user_logout(:userid);


为了对比的公平，把本地事务的性能和同步事务进行比较，得出的结论是同步复制带来的影响微乎其微。
把异步事务的性能和其他模式下的异步事务的性能进行比较，得出的结论是同步复制带来的影响微乎其微。

增加一组数据，采用同步事务测试复合场景过程中，主节点的CPU空闲率约48%，sync节点的空闲率约91%。async节点的空闲率约91%。
由于是同步事务，所以等待基本集中在xlog的写等待上。

【小结】
从测试结果来看，在sync节点的性能以及与Primary节点通讯网络带宽足够的情况下，同步复制带来的影响确实非常小。
PostgreSQL 9.1的性能令人满意，按照这个速度（异步事务），一台非常普通的X86服务器（另外两台备机如果提供查询的话可以满足更多的查询需求），一天可以提供20多亿次插入请求和10多亿次的查询请求。
结合PG分布式应用，再多点用户数只要加服务器就行了，非常容易扩展。

【测试sync节点DOWN机后自动将potential节点切换为sync节点】
1. 关闭1号standby节点, 然后去主节点观察. 2号standby节点已经变成sync状态了。 
postgres=# select * from pg_stat_replication;
 procpid | usesysid | usename | application_name | client_addr | client_hostname | client_port |         backend_start         |   s
tate   | sent_location | write_location | flush_location | replay_location | sync_priority | sync_state 
---------+----------+---------+------------------+-------------+-----------------+-------------+-------------------------------+----
-------+---------------+----------------+----------------+-----------------+---------------+------------
    2781 |    16464 | replica | standby2         | 172.xxx.xxx.2 |                 |       59262 | 2011-10-22 00:09:53.633187+08 | str
eaming | 5/17B32770    | 5/17B32770     | 5/17B32770     | 5/17B32770      |             2 | sync
(1 row)

2. 再次开启1号standby节点，1号节点恢复为sync状态，2号又变回potential.
 
postgres=# select * from pg_stat_replication;
 procpid | usesysid | usename | application_name | client_addr | client_hostname | client_port |         backend_start         |   s
tate   | sent_location | write_location | flush_location | replay_location | sync_priority | sync_state 
---------+----------+---------+------------------+-------------+-----------------+-------------+-------------------------------+----
-------+---------------+----------------+----------------+-----------------+---------------+------------
    2920 |    16464 | replica | standby1         | 172.xxx.xxx.1 |                 |       13088 | 2011-10-22 00:28:30.324383+08 | str
eaming | 5/17B32770    | 5/17B32770     | 5/17B32770     | 5/17B32770      |             1 | sync
    2781 |    16464 | replica | standby2         | 172.xxx.xxx.2 |                 |       59262 | 2011-10-22 00:09:53.633187+08 | str
eaming | 5/17B32770    | 5/17B32770     | 5/17B32770     | 5/17B32770      |             2 | potential
(2 rows)

3. 把1号和2号standby节点都关掉。主库的sync事务将进入等待状态,非sync事务依然可以进行读写。
发现所有standby节点都挂了之后应该考虑立即取消sync模式, 保持事务进行。

【配置文件】
standby节点的recovery.conf文件 : 
standby_mode = on
primary_conninfo = 'host=172.xxx.xxx.3 port=5432 user=replica application_name=standby1'         # e.g. 'host=localhost port=5432'
trigger_file = '/pgdata/digoal/5432/data02/pg_root/postgresql.trigger.5432'


【测试数据样本】 
 
digoal=> select * from user_info limit 5;
 userid |   engname   | cnname | occupation |  birthday  |                    signname                    |     email      |    qq  
   |          crt_time          | mod_time 
--------+-------------+--------+------------+------------+------------------------------------------------+----------------+--------
---+----------------------------+----------
      1 | digoal.zhou | 德哥   | DBA        | 1970-01-01 | 公益是一辈子的事, I'm Digoal.Zhou, Just do it! | digoal@126.com | 2767324
31 | 2011-10-21 22:01:20.029311 | 
      2 | digoal.zhou | 德哥   | DBA        | 1970-01-01 | 公益是一辈子的事, I'm Digoal.Zhou, Just do it! | digoal@126.com | 2767324
31 | 2011-10-21 22:01:20.029424 | 
      3 | digoal.zhou | 德哥   | DBA        | 1970-01-01 | 公益是一辈子的事, I'm Digoal.Zhou, Just do it! | digoal@126.com | 2767324
31 | 2011-10-21 22:01:20.029433 | 
      4 | digoal.zhou | 德哥   | DBA        | 1970-01-01 | 公益是一辈子的事, I'm Digoal.Zhou, Just do it! | digoal@126.com | 2767324
31 | 2011-10-21 22:01:20.029438 | 
      5 | digoal.zhou | 德哥   | DBA        | 1970-01-01 | 公益是一辈子的事, I'm Digoal.Zhou, Just do it! | digoal@126.com | 2767324
31 | 2011-10-21 22:01:20.029442 | 
(5 rows)

Time: 0.612 ms
digoal=> select * from user_login_rec limit 5;
  userid  |         login_time         |      ip      
----------+----------------------------+--------------
 23375965 | 2011-10-21 22:56:04.722263 | 172.xxx.xxx.xxx
 22441124 | 2011-10-21 22:56:04.727708 | 172.xxx.xxx.xxx
 46581992 | 2011-10-21 22:56:04.721366 | 172.xxx.xxx.xxx
 17439690 | 2011-10-21 22:56:04.72087  | 172.xxx.xxx.xxx
 36373256 | 2011-10-21 22:56:04.722019 | 172.xxx.xxx.xxx
(5 rows)

Time: 0.516 ms
digoal=> select * from user_logout_rec limit 5;
  userid  |        logout_time         |      ip      
----------+----------------------------+--------------
  8899868 | 2011-10-21 22:17:03.468211 | 172.xxx.xxx.xxx
 47715166 | 2011-10-21 22:17:03.467715 | 172.xxx.xxx.xxx
  4865187 | 2011-10-21 22:17:03.468816 | 172.xxx.xxx.xxx
 25229626 | 2011-10-21 22:17:03.470679 | 172.xxx.xxx.xxx
  7699796 | 2011-10-21 22:17:03.470818 | 172.xxx.xxx.xxx
(5 rows)

Time: 0.541 ms

【其他】
同步复制大多数的瓶颈来自备机的IO能力以及主机和备机之间的网络延时. 可以通过以下工具进行分析.
1. IO分析工具
iostat
2. 网络分析工具
sar -n DEV 1 10000
mtr
qperf
文章标签：
云原生数据库 PolarDB
关系型数据库
流计算
数据库管理
PostgreSQL
数据库
关键词：
can PostgreSQL
sync PostgreSQL
相关实践学习
使用PolarDB和ECS搭建门户网站
本场景主要介绍基于PolarDB和ECS实现搭建门户网站。
阿里云数据库产品家族及特性
阿里云智能数据库产品团队一直致力于不断健全产品体系，提升产品性能，打磨产品功能，从而帮助客户实现更加极致的弹性能力、具备更强的扩展能力、并利用云设施进一步降低企业成本。以云原生+分布式为核心技术抓手，打造以自研的在线事务型(OLTP)数据库Polar DB和在线分析型(OLAP)数据库Analytic DB为代表的新一代企业级云原生数据库产品体系， 结合NoSQL数据库、数据库生态工具、云原生智能化数据库管控平台，为阿里巴巴经济体以及各个行业的企业客户和开发者提供从公共云到混合云再到私有云的完整解决方案，提供基于云基础设施进行数据从处理、到存储、再到计算与分析的一体化解决方案。本节课带你了解阿里云数据库产品家族及特性。
德哥
目录
相关文章
玄惭
|
Unix 关系型数据库 PostgreSQL
PostgreSQL stream repication can implement between FreeBSD and CentOS
玄惭
1935 0
德哥
|
关系型数据库 PostgreSQL
How to control who can see PostgreSQL function's source code
德哥
1672 0
德哥
|
安全 关系型数据库
PostgreSQL 9.5 new feature - can define row security policy for table
德哥
1400 0
德哥
|
关系型数据库 PostgreSQL
[example]Sync data from PostgreSQL database to another PostgreSQL database
德哥
1260 0
德哥
|
Oracle 关系型数据库 PostgreSQL
a foolish sync method about sync data from PostgreSQL to Oracle
德哥
1575 0
开发工程师
|
关系型数据库 分布式数据库 PolarDB
《阿里云产品手册2022-2023 版》——PolarDB for PostgreSQL
《阿里云产品手册2022-2023 版》——PolarDB for PostgreSQL
开发工程师
368 0
-开发达人-
|
存储 缓存 关系型数据库
PolarDB for PostgreSQL 开源必读手册-开源PolarDB for PostgreSQL架构介绍（上）
PolarDB for PostgreSQL 开源必读手册
-开发达人-
502 0
-开发达人-
|
存储 SQL 并行计算
PolarDB for PostgreSQL 开源必读手册-开源PolarDB for PostgreSQL架构介绍（中）
PolarDB for PostgreSQL 开源必读手册-开源PolarDB for PostgreSQL架构介绍
-开发达人-
429 0
-开发达人-
|
存储 算法 安全
PolarDB for PostgreSQL 开源必读手册-开源PolarDB for PostgreSQL架构介绍（下）
PolarDB for PostgreSQL 开源必读手册-开源PolarDB for PostgreSQL架构介绍
-开发达人-
387 0
-开发达人-
|
关系型数据库 分布式数据库 开发工具
PolarDB for PostgreSQL 开源必读手册-PolarDB安装与配置（上）
PolarDB for PostgreSQL 开源必读手册
-开发达人-
857 0

热门文章

最新文章

  • 1
    Hadoop迁移MaxCompute神器之DataX-On-Hadoop使用指南
  • 2
    Nginx的启动、停止与重启
  • 3
    Hive之数据倾斜的原因和解决方法
  • 4
    开发人员各级岗位胜任力模型
  • 5
    Node.js 中流操作实践
  • 6
    传奇龙版地图查看器
  • 7
    调整DC的网络对网络是否会有影响?
  • 8
    How to maintain Oracle10g Recyclebin?
  • 9
    你知道自己感染了恶意软件吗？
  • 10
    Groovy 快速入门
  • 1
    Python学习的自我理解和想法（9）
    89
  • 2
    Kimi 上线视觉思考模型，K1 系列强化学习模型正式开放，无需借助外部 OCR 处理图像与文本进行思考并回答
    101
  • 3
    CosyVoice 2.0：阿里开源升级版语音生成大模型，支持多语言和跨语言语音合成，提升发音和音色等的准确性
    79
  • 4
    Megrez-3B-Omni：无问芯穹开源最强端侧全模态模型，支持理解图像、音频和文本三种模态数据
    50
  • 5
    FreeScale：无需微调即可提升模型的图像生成能力，生成 8K 分辨率的高质量图像
    37
  • 6
    Freestyler：微软联合西工大和香港大学推出说唱音乐生成模型，支持控制生成的音色、风格和节奏等
    37
  • 7
    《C 语言与统计假设检验：洞察人工智能模型性能差异》
    82
  • 8
    《C 语言赋能：物联网环境下人工智能应用的能耗优化之道》
    22
  • 9
    《C 语言助力神经网络剪枝：优化模型的卓越之道》
    25
  • 10
    《C 语言复数运算：赋能人工智能信号处理新境界》
    20

    • 相关课程

    更多
  • PolarDB for PostgreSQL 开源人才初级认证培训课程
  • PolarDB for PostgreSQL 内核解读系列课程
  • 7天突破PolarDB for PostgreSQL 2022版
  • PolarDB for PostgreSQL 入门
  • PostgreSQL云上开发实践
  • PostgreSQL实战进阶

    • 相关电子书

    更多
  • 金融级 PostgreSQL监控及优化
  • PostgreSQL在哈啰的实践-周飞
  • PostgreSQL高并发数据库应用数据

    • 相关实验场景

    更多
  • PolarDB for PostgreSQL 闪回特性体验
  • 搭建PostgreSQL主从架构
  • 如何开通PostgreSQL的专属集群

    • 推荐镜像

    更多
  • postgresql
  • mongodb
  • mysql
    • 下一篇
    fs.oss.accessKeyId和fs.oss.accessKeySecret。