本节简单介绍了Greenplum开源的列式存储zedstore的安装和使用.
目前创新互联已为上千多家的企业提供了网站建设、域名、雅安服务器托管、网站托管、服务器租用、企业网站设计、上虞网站维护等服务，公司将坚持客户导向、应用为本的策略，正道将秉承"和谐、参与、激情"的文化，与客户和合作伙伴齐心协力一起成长，共同发展。
安装
从Github上下载源码,与普通PG一样,编译安装即可
[root@localhost postgres-zedstore]# ./configure --enable-debug --with-python --with-perl --with-tcl --with-gssapi --with-pam --with-openssl --with-pam --without-ldap --with-libxml --with-libxslt --enable-dtrace --enable-depend --enable-cassert --with-systemd CFLAGS="-O0 -DOPTIMIZER_DEBUG -g3 -gdwarf-2" --prefix=/appdb/zedstore
checking build system type... x86_64-pc-linux-gnu
...
[root@localhost postgres-zedstore]# make -j4
...
[root@localhost postgres-zedstore]# make install
...
PostgreSQL installation complete.
Heap vs ZedStore
创建用户,初始化数据库
[zedstore@localhost ~]$ initdb -E utf8 -D /data/zedstore/testdb
The files belonging to this database system will be owned by user "zedstore".
This user must also own the server process.
The database cluster will be initialized with locale "en_US.UTF-8".
The default text search configuration will be set to "english".
Data page checksums are disabled.
creating directory /data/zedstore/testdb ... ok
creating subdirectories ... ok
selecting dynamic shared memory implementation ... posix
selecting default max_connections ... 100
selecting default shared_buffers ... 128MB
selecting default timezone ... PRC
creating configuration files ... ok
running bootstrap script ... ok
performing post-bootstrap initialization ... ok
syncing data to disk ... ok
initdb: warning: enabling "trust" authentication for local connections
You can change this by editing pg_hba.conf or using the option -A, or
--auth-local and --auth-host, the next time you run initdb.
Success. You can now start the database server using:
pg_ctl -D /data/zedstore/testdb -l logfile start
下面来对比一下head am和zedstore的性能差异
PG
testdb=# create table t_olap(id int,c1 int,c2 varchar(20));
CREATE TABLE
testdb=# insert into t_olap select x,x,'c2'||x from generate_series(1,5000000) as x;
INSERT 0 5000000
testdb=#
执行查询
testdb=# explain analyze select avg(id),sum(c1),max(c2) from t_olap;
QUERY PLAN
-------------------------------------------------------------------------------------------------
Finalize Aggregate (cost=69209.40..69209.40 rows=1 width=72) (actual time=925.540..925.540 rows=1 loops=1)
-> Gather (cost=69209.17..69209.38 rows=2 width=72) (actual time=925.284..932.688 rows=3 loops=1)
Workers Planned: 2
Workers Launched: 2
-> Partial Aggregate (cost=68209.17..68209.18 rows=1 width=72) (actual time=911.539..911.539 rows=1 loops=3)
-> Parallel Seq Scan on t_olap (cost=0.00..52584.24 rows=2083324 width=17) (actual time=0.037..240.287 rows=
1666667 loops=3)
Planning Time: 22.703 ms
Execution Time: 933.020 ms
(8 rows)
执行时间为933ms
ZedStore
[zedstore@localhost testdb]$ psql -d testdb
psql (13devel)
Type "help" for help.
testdb=# \dA+
List of access methods
Name | Type | Handler | Description
----------+-------+--------------------------+----------------------------------------
brin | Index | brinhandler | block range index (BRIN) access method
btree | Index | bthandler | b-tree index access method
gin | Index | ginhandler | GIN index access method
gist | Index | gisthandler | GiST index access method
hash | Index | hashhandler | hash index access method
heap | Table | heap_tableam_handler | heap table access method
spgist | Index | spghandler | SP-GiST index access method
zedstore | Table | zedstore_tableam_handler | zedstore table access method
(8 rows)
testdb=# create table t_olap(id int,c1 int,c2 varchar(20)) using zedstore;
CREATE TABLE
testdb=# \d+ t_olap
Table "public.t_olap"
Column | Type | Collation | Nullable | Default | Storage | Stats target | Description
--------+-----------------------+-----------+----------+---------+----------+--------------+-------------
id | integer | | | | plain | |
c1 | integer | | | | plain | |
c2 | character varying(20) | | | | extended | |
Access method: zedstore
testdb=# insert into t_olap select x,x,'c2'||x from generate_series(1,5000000) as x;
INSERT 0 5000000
执行查询
testdb=# explain analyze select avg(id),sum(c1),max(c2) from t_olap;
QUERY PLAN
-------------------------------------------------------------------------------------------------
Finalize Aggregate (cost=31425.10..31425.11 rows=1 width=72) (actual time=1707.755..1707.756 rows=1 loops=1)
-> Gather (cost=31424.87..31425.08 rows=2 width=72) (actual time=1659.121..1710.512 rows=3 loops=1)
Workers Planned: 2
Workers Launched: 2
-> Partial Aggregate (cost=30424.87..30424.88 rows=1 width=72) (actual time=1647.216..1647.217 rows=1 loops=3)
-> Parallel Seq Scan on t_olap (cost=0.00..24130.07 rows=839307 width=17) (actual time=0.418..1124.465 rows=
1666667 loops=3)
Planning Time: 1.907 ms
Execution Time: 1753.191 ms
(8 rows)
执行时间为1753ms,列式存储似乎没有发挥作用?待续.
参考资料
PostgreSQL 基于access method api的列存zedstore