ClickHouse基本操作（一）

常用SQL

创建表

1 2 3 4 5 6 7

CREATE TABLE b6logs( eventDate Date, impid UInt64, uid String, idfa String, imei String ) ENGINE=MergeTree(eventDate, (impid, eventDate), 8192)

一般情况下, 都建议使用 MergeTree 引擎. 这个引擎必须要有一个 Date 的列来作为索引, 即上面的 eventDate.

导入CSV数据

1	cat xxx.csv | clickhouse-client --query="INSERT INTO b6logs FORMAT CSV";

指定分隔符

1	cat xxx.csv | clickhouse-client --format_csv_delimiter="|" --query="INSERT INTO b6logs FORMAT CSV";

导入数据时忽略错误

1	clickhouse-client --input_format_allow_errors_num=100000 --input_format_allow_errors_ratio=0.2

--input_format_allow_errors_num : 是允许的错误数

--input_format_allow_errors_ratio : 是允许的错误率, 范围是 [0-1]

导出 CSV 数据

1

clickhouse-client --query="select uid, idfa, imei from (select impid, uid from b2logs where impid >= 15289903030261609347 and impid <= 15289904230261609347) any inner join (select impid, idfa, imei from b6logs where impid >= 15289903030261609347 and impid <= 15289904230261609347) using(impid) format CSV" > 9c9dc608-269b-4f02-b122-ef5dffb2669d.log

即语法为 select xxxx format CSV

重命名表

1	rename table tbl1 to btl2;

删除表

1	drop table tbl;

添加列

1	alter table dsp_statis add column cost UInt32 default 0;

查看表结构

1

desc tbl;

更多语法, 参考官方文档. https://clickhouse.yandex/docs/en/query_language/queries/

MergeTree 引擎中删除分区

注意, 默认情况下 mergeTree 引擎是按月分区的, 删除分区的格式为 201808

如果想修改为按日分区, 则在建表时加上:

1 2 3

ENGINE = MergeTree PARTITION BY eventDate ORDER BY imp_id SETTINGS index_granularity = 8192; 然后就可以: alter table xxx drop partition '2018-08-08';

默认情况下, Clickhouse 不允许删除分区或表的大小大于 50GB 的分区或表. 可以通过修改server的配置文件来永久配置. 也可以临时设置一下来删除而不用重启服务.

永久配置

1 2 3 4 5 6 7 8

sudo vim /etc/clickhouse-server/config.xml   然后注释掉下面两行   <!-- <max_table_size_to_drop>0</max_table_size_to_drop> --> <!-- <max_partition_size_to_drop>0</max_partition_size_to_drop> -->   0表示不限制. 或者你可以设置为你想限制的最大的大小.

临时设置

创建个标志文件:

1	sudo touch '/home/username/clickhouse/flags/force_drop_table' && sudo chmod 666 '/home/username/clickhouse/flags/force_drop_table'

创建好之后, 就可以执行上面的删除分区或表的命令了.

查看表大小

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

SELECT database, table, formatReadableSize(size) AS size, formatReadableSize(bytes_on_disk) AS bytes_on_disk, formatReadableSize(data_uncompressed_bytes) AS data_uncompressed_bytes, formatReadableSize(data_compressed_bytes) AS data_compressed_bytes, compress_rate, rows, days, formatReadableSize(avgDaySize) AS avgDaySize FROM ( SELECT database, table, sum(bytes) AS size, sum(rows) AS rows, min(min_date) AS min_date, max(max_date) AS max_date, sum(bytes_on_disk) AS bytes_on_disk, sum(data_uncompressed_bytes) AS data_uncompressed_bytes, sum(data_compressed_bytes) AS data_compressed_bytes, (data_compressed_bytes / data_uncompressed_bytes) * 100 AS compress_rate, max_date - min_date AS days, size / (max_date - min_date) AS avgDaySize FROM system.parts WHERE active GROUP BY database, table ORDER BY database ASC, size DESC )

执行 SQL 文件

1	clickhouse-client -d 数据库 --multiquery < /tmp/your.sql.file

查看分区信息

1	select partition, name, active from system.parts WHERE table = 'visits'

性能相关收集

join 表性能

切记, 要用大表 join 小表. (不知道具体为什么, 从经验上看, 用大表作为驱动表, 性能远远快于用小表作为驱动表). (MySQL 里的话, 则是小表驱动大表).

优化 distinct count

之前

1	select yob, count(), count(distinct uid, idfa, imei) from nginx_bid_log where eventDate='2018-9-1' group by yob;

之后

1	select yob, count(), count(distinct(sipHash64(concat(uid, idfa, imei)))) from nginx_bid_log where eventDate='2018-9-1' group by yob;

查看数据分布

1	select histogram(100)(upstream_resp_time) from (select upstream_resp_time from nginx_bid_log where eventDate = '2018-12-13') format CSV;

histogram(100) 表示组距100 (即分成100等份的的分布) , 后面的 upstream_resp_time 是你的列名, 即按这个列的数据来进行统计.

bar

1	select upstream_resp_time, bar(列名, 最小值, 最大, step) from tableXX;

显示简单的图形.

hex 十六进制 转换为 十进制

1	SELECT reinterpretAsInt64(reverse(unhex('123')));

md5 分区

1 2 3 4 5

# 一 SELECT reinterpretAsInt64(reverse(unhex(substring(md5_field, 1, 1))));   # 二, md5 => hex => 十进制 => 取模 SELECT modulo(reinterpretAsInt64(reverse(unhex(substring(md5_field, 1, 1)))), 5);

参考资料

• http://www.clickhouse.com.cn/topic/5ad9521b9d28dfde2ddc5f86

• https://clickhouse.yandex/docs/en/query_language/queries/

• https://github.com/yandex/ClickHouse/blob/master/dbms/tests/queries/0_stateless/00418_input_format_allow_errors.sh

• https://groups.google.com/forum/m/#!topic/clickhouse/U_Uc4omvoqc