Docker快速搭建Clickhouse集羣(3分片3副本)
背景
前文簡單介紹了下Clickhouse的安裝和客戶端使用，在實際生產環境中，Clickhouse經常是以集羣模式部署的，因爲不少系統不知足sse4.2指令，這裏使用docker來搭建一個Clickhouse的集羣。html
1. 環境說明
1.1 機器列表
機器名node |
IP算法 |
配置sql |
操做系統docker |
部署的服務數據庫 |
備註centos |
server01服務器 |
192.168.21.21運維 |
8c8gssh |
centos7.3 |
clickhouserver（cs01-01）和 clickhouserver（cs01-02） |
clickhouse01-01: 實例1, 端口: tcp 9000, http 8123, 同步端口9009, 類型: 分片1, 副本1 clickhouse01-02: 實例2, 端口: tcp 9001, http 8124, 同步端口9010, 類型: 分片2, 副本2 (clickhouse2的副本) |
server02 |
192.168.21.69 |
8c8g |
centos7.3 |
clickhouserver（cs02-01）和 clickhouserver（cs02-02） |
clickhouse02-01: 實例1, 端口: tcp 9000, http 8123, 同步端口9009, 類型: 分片2, 副本1 clickhouse02-02: 實例2, 端口: tcp 9001, http 8124, 同步端口9010, 類型: 分片3, 副本2 (clickhouse3的副本) |
server03 |
192.168.21.6 |
8c8g |
centos7.3 |
clickhouserver（cs03-01）和 clickhouserver（cs03-02） |
clickhouse03-01: 實例1, 端口: tcp 9000, http 8123, 同步端口9009, 類型: 分片3, 副本1 clickhouse03-02: 實例2, 端口: tcp 9001, http 8124, 同步端口9010, 類型: 分片1, 副本2 (clickhouse1的副本) |
1.2 機器初始化
1.2.1 配置host
執行：vi /etc/hosts， 加入下面三行：
192.168.21.21 server01
192.168.21.69 server02
192.168.21.6 server03
1.2.2 安裝docker
每臺機器上均安裝相同版本的docker
參照：docker環境搭建
1.2.3 安裝Zookeeper
1.3 目錄初始化
1.3.1 建立對應本地路徑
分別在三臺服務器，建立數據存儲目錄：
mkdir /data/clickhouse
1.3.2 獲取clickhouse-server的配置
1）在server01服務器作以下操做
先按照官方教程的docker命令啓動Clickhouse-Server
docker run -d --name clickhouse-server --ulimit nofile=262144:262144 --volume=/data/clickhouse/:/var/lib/clickhouse yandex/clickhouse-server
2）啓動完成後，複製容器內的配置文件到本機目錄下
#拷貝容器內容的配置到/etc目錄下 docker cp clickhouse-server:/etc/clickhouse-server/ /etc/ #將server01上的目錄重命名 cp -rf /etc/clickhouse-server/ /etc/clickhouse-server01/ cp -rf /etc/clickhouse-server/ /etc/clickhouse-server02/
3）而後將/etc/clickhouse-server/ 分別拷貝到每一個機器上
#拷貝配置到server02上 scp /etc/clickhouse-server/ server02:/etc/clickhouse-server01/ scp /etc/clickhouse-server/ server02:/etc/clickhouse-server02/ #拷貝配置到server03上 scp /etc/clickhouse-server/ server03:/etc/clickhouse-server01/ scp /etc/clickhouse-server/ server03:/etc/clickhouse-server02/
2. 集羣環境搭建
2.1 集羣環境拓撲圖
集羣環境說明：
clickhouse01-01: 實例1, 端口: tcp 9000, http 8123, 同步端口9009, 類型: 分片1, 副本1
clickhouse01-02: 實例2, 端口: tcp 9001, http 8124, 同步端口9010, 類型: 分片2, 副本2 (clickhouse2的副本)
clickhouse02-01: 實例1, 端口: tcp 9000, http 8123, 同步端口9009, 類型: 分片2, 副本1
clickhouse02-02: 實例2, 端口: tcp 9001, http 8124, 同步端口9010, 類型: 分片3, 副本2 (clickhouse3的副本)
clickhouse03-01: 實例1, 端口: tcp 9000, http 8123, 同步端口9009, 類型: 分片3, 副本1
clickhouse03-02: 實例2, 端口: tcp 9001, http 8124, 同步端口9010, 類型: 分片1, 副本2 (clickhouse1的副本)
2.2 配置集羣
2.2.1 待修改的配置文件
須要修改的配置有兩個(若是有須要也能夠配置user.xml)：
- /etc/clickhouse-server/config.xml
- /etc/clickhouse-server/metrika.xml（新增文件）
2.2.2 server1上配置clickhouse-server的實例
2.2.2.1 clickhouse-01-01的配置：
1）/etc/clickhouse-server01/config.xml(其餘實例此配置內容和這個同樣就行）
修改include from節點爲實際的引用到的文件
<!-- If element has 'incl' attribute, then for it's value will be used corresponding substitution from another file. By default, path to file with substitutions is /etc/metrika.xml. It could be changed in config in 'include_from' element. Values for substitutions are specified in /yandex/name_of_substitution elements in that file. --> <include_from>/etc/clickhouse-server/metrika.xml</include_from> <listen_host>0.0.0.0</listen_host> <listen_host>127.0.0.1</listen_host>
2）/etc/clickhouse-server01/metrika.xml（全部實例的配置內容都和這個同樣就行）
<!--全部實例均使用這個集羣配置，不用個性化 --> <yandex> <!-- 集羣配置 --> <!-- clickhouse_remote_servers全部實例配置都同樣 --> <!-- 集羣配置 --> <clickhouse_remote_servers> <cluster_3s_1r> <!-- 數據分片1 --> <shard> <internal_replication>true</internal_replication> <replica> <host>server01</host> <port>9000</port> <user>default</user> <password></password> </replica> <replica> <host>server03</host> <port>9001</port> <user>default</user> <password></password> </replica> </shard> <!-- 數據分片2 --> <shard> <internal_replication>true</internal_replication> <replica> <host>server02</host> <port>9000</port> <user>default</user> <password></password> </replica> <replica> <host>server01</host> <port>9001</port> <user>default</user> <password></password> </replica> </shard> <!-- 數據分片3 --> <shard> <internal_replication>true</internal_replication> <replica> <host>server03</host> <port>9000</port> <user>default</user> <password></password> </replica> <replica> <host>server02</host> <port>9001</port> <user>default</user> <password></password> </replica> </shard> </cluster_3s_1r> </clickhouse_remote_servers> <!-- ZK --> <!-- zookeeper_servers全部實例配置都同樣 --> <zookeeper-servers> <node index="1"> <host>192.168.21.66</host> <port>2181</port> </node> <node index="2"> <host>192.168.21.57</host> <port>2181</port> </node> <node index="3"> <host>192.168.21.17</host> <port>2181</port> </node> </zookeeper-servers> <!-- marcos每一個實例配置不同 分片1, 副本1 --> <macros> <layer>01</layer> <shard>01</shard> <replica>cluster01-01-1</replica> </macros> <networks> <ip>::/0</ip> </networks> <!-- 數據壓縮算法 --> <clickhouse_compression> <case> <min_part_size>10000000000</min_part_size> <min_part_size_ratio>0.01</min_part_size_ratio> <method>lz4</method> </case> </clickhouse_compression> </yandex>
2.2.2.2 clickhouse-01-02的配置：
1）/etc/clickhouse-server02/metrika.xml（全部實例的配置內容都和這個同樣就行）
<!--全部實例均使用這個集羣配置，不用個性化 --> <yandex> <!-- 集羣配置 --> <!-- clickhouse_remote_servers全部實例配置都同樣 --> <clickhouse_remote_servers> <cluster_3s_1r> <!-- 數據分片1 --> <shard> <internal_replication>true</internal_replication> <replica> <host>server01</host> <port>9000</port> <user>default</user> <password></password> </replica> <replica> <host>server03</host> <port>9001</port> <user>default</user> <password></password> </replica> </shard> <!-- 數據分片2 --> <shard> <internal_replication>true</internal_replication> <replica> <host>server02</host> <port>9000</port> <user>default</user> <password></password> </replica> <replica> <host>server01</host> <port>9001</port> <user>default</user> <password></password> </replica> </shard> <!-- 數據分片3 --> <shard> <internal_replication>true</internal_replication> <replica> <host>server03</host> <port>9000</port> <user>default</user> <password></password> </replica> <replica> <host>server02</host> <port>9001</port> <user>default</user> <password></password> </replica> </shard> </cluster_3s_1r> </clickhouse_remote_servers> <!-- ZK --> <!-- zookeeper_servers全部實例配置都同樣 --> <zookeeper-servers> <node index="1"> <host>192.168.21.66</host> <port>2181</port> </node> <node index="2"> <host>192.168.21.57</host> <port>2181</port> </node> <node index="3"> <host>192.168.21.17</host> <port>2181</port> </node> </zookeeper-servers> <!-- marcos每一個實例配置不同 分片2, 副本2--> <macros> <layer>01</layer> <shard>02</shard> <replica>cluster01-02-2</replica> </macros> <networks> <ip>::/0</ip> </networks> <!-- 數據壓縮算法 --> <clickhouse_compression> <case> <min_part_size>10000000000</min_part_size> <min_part_size_ratio>0.01</min_part_size_ratio> <method>lz4</method> </case> </clickhouse_compression> </yandex>
2.2.3 server2上配置clickhouse-server的實例
2.2.3.1 clickhouse-02-01的配置：
/etc/clickhouse-server01/metrika.xml
<!--全部實例均使用這個集羣配置，不用個性化 --> <yandex> <!-- 集羣配置 --> <!-- clickhouse_remote_servers全部實例配置都同樣 --> <clickhouse_remote_servers> <cluster_3s_1r> <!-- 數據分片1 --> <shard> <internal_replication>true</internal_replication> <replica> <host>server01</host> <port>9000</port> <user>default</user> <password></password> </replica> <replica> <host>server03</host> <port>9001</port> <user>default</user> <password></password> </replica> </shard> <!-- 數據分片2 --> <shard> <internal_replication>true</internal_replication> <replica> <host>server02</host> <port>9000</port> <user>default</user> <password></password> </replica> <replica> <host>server01</host> <port>9001</port> <user>default</user> <password></password> </replica> </shard> <!-- 數據分片3 --> <shard> <internal_replication>true</internal_replication> <replica> <host>server03</host> <port>9000</port> <user>default</user> <password></password> </replica> <replica> <host>server02</host> <port>9001</port> <user>default</user> <password></password> </replica> </shard> </cluster_3s_1r> </clickhouse_remote_servers> <!-- ZK --> <!-- zookeeper_servers全部實例配置都同樣 --> <zookeeper-servers> <node index="1"> <host>192.168.21.66</host> <port>2181</port> </node> <node index="2"> <host>192.168.21.57</host> <port>2181</port> </node> <node index="3"> <host>192.168.21.17</host> <port>2181</port> </node> </zookeeper-servers> <!-- marcos每一個實例配置不同 分片2, 副本1--> <macros> <layer>01</layer> <shard>02</shard> <replica>cluster01-02-1</replica> </macros> <networks> <ip>::/0</ip> </networks> <!-- 數據壓縮算法 --> <clickhouse_compression> <case> <min_part_size>10000000000</min_part_size> <min_part_size_ratio>0.01</min_part_size_ratio> <method>lz4</method> </case> </clickhouse_compression> </yandex>
2.2.3.2 clickhouse-02-02的配置：
/etc/clickhouse-server02/metrika.xml
<!--全部實例均使用這個集羣配置，不用個性化 --> <yandex> <!-- 集羣配置 --> <!-- clickhouse_remote_servers全部實例配置都同樣 --> <clickhouse_remote_servers> <cluster_3s_1r> <!-- 數據分片1 --> <shard> <internal_replication>true</internal_replication> <replica> <host>server01</host> <port>9000</port> <user>default</user> <password></password> </replica> <replica> <host>server03</host> <port>9001</port> <user>default</user> <password></password> </replica> </shard> <!-- 數據分片2 --> <shard> <internal_replication>true</internal_replication> <replica> <host>server02</host> <port>9000</port> <user>default</user> <password></password> </replica> <replica> <host>server01</host> <port>9001</port> <user>default</user> <password></password> </replica> </shard> <!-- 數據分片3 --> <shard> <internal_replication>true</internal_replication> <replica> <host>server03</host> <port>9000</port> <user>default</user> <password></password> </replica> <replica> <host>server02</host> <port>9001</port> <user>default</user> <password></password> </replica> </shard> </cluster_3s_1r> </clickhouse_remote_servers> <!-- ZK --> <!-- zookeeper_servers全部實例配置都同樣 --> <zookeeper-servers> <node index="1"> <host>192.168.21.66</host> <port>2181</port> </node> <node index="2"> <host>192.168.21.57</host> <port>2181</port> </node> <node index="3"> <host>192.168.21.17</host> <port>2181</port> </node> </zookeeper-servers> <!-- marcos每一個實例配置不同 分片3, 副本2--> <macros> <layer>01</layer> <shard>03</shard> <replica>cluster01-03-2</replica> </macros> <networks> <ip>::/0</ip> </networks> <!-- 數據壓縮算法 --> <clickhouse_compression> <case> <min_part_size>10000000000</min_part_size> <min_part_size_ratio>0.01</min_part_size_ratio> <method>lz4</method> </case> </clickhouse_compression> </yandex>
2.2.4 server3上配置clickhouse-server的實例
2.2.4.1 clickhouse-03-01的配置：
1）/etc/clickhouse-server01/metrika.xml
<!--全部實例均使用這個集羣配置，不用個性化 --> <yandex> <!-- 集羣配置 --> <!-- clickhouse_remote_servers全部實例配置都同樣 --> <clickhouse_remote_servers> <cluster_3s_1r> <!-- 數據分片1 --> <shard> <internal_replication>true</internal_replication> <replica> <host>server01</host> <port>9000</port> <user>default</user> <password></password> </replica> <replica> <host>server03</host> <port>9001</port> <user>default</user> <password></password> </replica> </shard> <!-- 數據分片2 --> <shard> <internal_replication>true</internal_replication> <replica> <host>server02</host> <port>9000</port> <user>default</user> <password></password> </replica> <replica> <host>server01</host> <port>9001</port> <user>default</user> <password></password> </replica> </shard> <!-- 數據分片3 --> <shard> <internal_replication>true</internal_replication> <replica> <host>server03</host> <port>9000</port> <user>default</user> <password></password> </replica> <replica> <host>server02</host> <port>9001</port> <user>default</user> <password></password> </replica> </shard> </cluster_3s_1r> </clickhouse_remote_servers> <!-- ZK --> <!-- zookeeper_servers全部實例配置都同樣 --> <zookeeper-servers> <node index="1"> <host>192.168.21.66</host> <port>2181</port> </node> <node index="2"> <host>192.168.21.57</host> <port>2181</port> </node> <node index="3"> <host>192.168.21.17</host> <port>2181</port> </node> </zookeeper-servers> <!-- marcos每一個實例配置不同 分片3, 副本1--> <macros> <layer>01</layer> <shard>03</shard> <replica>cluster01-03-1</replica> </macros> <networks> <ip>::/0</ip> </networks> <!-- 數據壓縮算法 --> <clickhouse_compression> <case> <min_part_size>10000000000</min_part_size> <min_part_size_ratio>0.01</min_part_size_ratio> <method>lz4</method> </case> </clickhouse_compression> </yandex>
2.2.4.2 clickhouse-03-02的配置：
1. /etc/clickhouse-server02/metrika.xml
<!--全部實例均使用這個集羣配置，不用個性化 --> <yandex> <!-- 集羣配置 --> <!-- clickhouse_remote_servers全部實例配置都同樣 --> <clickhouse_remote_servers> <cluster_3s_1r> <!-- 數據分片1 --> <shard> <internal_replication>true</internal_replication> <replica> <host>server01</host> <port>9000</port> <user>default</user> <password></password> </replica> <replica> <host>server03</host> <port>9001</port> <user>default</user> <password></password> </replica> </shard> <!-- 數據分片2 --> <shard> <internal_replication>true</internal_replication> <replica> <host>server02</host> <port>9000</port> <user>default</user> <password></password> </replica> <replica> <host>server01</host> <port>9001</port> <user>default</user> <password></password> </replica> </shard> <!-- 數據分片3 --> <shard> <internal_replication>true</internal_replication> <replica> <host>server03</host> <port>9000</port> <user>default</user> <password></password> </replica> <replica> <host>server02</host> <port>9001</port> <user>default</user> <password></password> </replica> </shard> </cluster_3s_1r> </clickhouse_remote_servers> <!-- ZK --> <!-- zookeeper_servers全部實例配置都同樣 --> <zookeeper-servers> <node index="1"> <host>192.168.21.66</host> <port>2181</port> </node> <node index="2"> <host>192.168.21.57</host> <port>2181</port> </node> <node index="3"> <host>192.168.21.17</host> <port>2181</port> </node> </zookeeper-servers> <!-- marcos每一個實例配置不同 分片1, 副本2--> <macros> <layer>01</layer> <shard>01</shard> <replica>cluster01-01-2</replica> </macros> <networks> <ip>::/0</ip> </networks> <!-- 數據壓縮算法 --> <clickhouse_compression> <case> <min_part_size>10000000000</min_part_size> <min_part_size_ratio>0.01</min_part_size_ratio> <method>lz4</method> </case> </clickhouse_compression> </yandex>
2.3 運行clickhouse集羣
2.3.1 運行clickhouse01-01實例
登錄到server01：ssh server01
#先刪除cs01-01容器 docker rm -f cs01-01 #運行Clickhouse-server docker run -d \ --name cs01-01 \ --ulimit nofile=262144:262144 \ --volume=/data/clickhouse01/:/var/lib/clickhouse \ --volume=/etc/clickhouse-server01/:/etc/clickhouse-server/ \ --add-host server01:192.168.21.21 \ --add-host server02:192.168.21.69 \ --add-host server03:192.168.21.6 \ --add-host i-r9es2e0q:192.168.21.21 \ --add-host i-o91d619w:192.168.21.69 \ --add-host i-ldipmbwa:192.168.21.6 \ --hostname $(hostname) \ -p 9000:9000 \ -p 8123:8123 \ -p 9009:9009 \ yandex/clickhouse-server
說明：
--add-host參數：由於咱們在配置文件中使用了hostname來指代咱們的服務器，爲了讓容器可以識別，因此須要加此參數，對應的host配置會自動被添加到容器主機的/etc/hosts裏面
--hostname參數：clickhouse中的system.clusters表會顯示集羣信息，其中is_local的屬性若是不配置hostname的話clickhouse沒法識別是不是當前本機。is_local都爲0的話會影響集羣操做，is_local經過clickhouse-client登陸到任一clickhouse-server上查看：SELECT * FROM system.clusters;
--p參數：暴露容器中的端口到本機端口中。
2.3.2 運行clickhouse01-02實例
登錄到server01：ssh server01;
docker run -d \ --name cs01-02 \ --ulimit nofile=262144:262144 \ --volume=/data/clickhouse02/:/var/lib/clickhouse \ --volume=/etc/clickhouse-server02/:/etc/clickhouse-server/ \ --add-host server01:192.168.21.21 \ --add-host server02:192.168.21.69 \ --add-host server03:192.168.21.6 \ --add-host i-r9es2e0q:192.168.21.21 \ --add-host i-o91d619w:192.168.21.69 \ --add-host i-ldipmbwa:192.168.21.6 \ --hostname $(hostname) \ -p 9001:9000 \ -p 8124:8123 \ -p 9010:9009 \ yandex/clickhouse-server
2.3.3 運行clickhouse02-01實例
登錄到server02：ssh server02
docker run -d \ --name cs02-01 \ --ulimit nofile=262144:262144 \ --volume=/data/clickhouse01/:/var/lib/clickhouse \ --volume=/etc/clickhouse-server01/:/etc/clickhouse-server/ \ --add-host server01:192.168.21.21 \ --add-host server02:192.168.21.69 \ --add-host server03:192.168.21.6 \ --add-host i-r9es2e0q:192.168.21.21 \ --add-host i-o91d619w:192.168.21.69 \ --add-host i-ldipmbwa:192.168.21.6 \ --hostname $(hostname) \ -p 9000:9000 \ -p 8123:8123 \ -p 9009:9009 \ yandex/clickhouse-server
2.3.4 運行clickhouse02-02實例
登錄到server02：ssh server02
docker run -d \ --name cs02-02 \ --ulimit nofile=262144:262144 \ --volume=/data/clickhouse02/:/var/lib/clickhouse \ --volume=/etc/clickhouse-server02/:/etc/clickhouse-server/ \ --add-host server01:192.168.21.21 \ --add-host server02:192.168.21.69 \ --add-host server03:192.168.21.6 \ --add-host i-r9es2e0q:192.168.21.21 \ --add-host i-o91d619w:192.168.21.69 \ --add-host i-ldipmbwa:192.168.21.6 \ --hostname $(hostname) \ -p 9001:9000 \ -p 8124:8123 \ -p 9010:9009 \ yandex/clickhouse-server
2.3.5 運行clickhouse03-01實例
登錄到server03：ssh server03
docker run -d \ --name cs03-01 \ --ulimit nofile=262144:262144 \ --volume=/data/clickhouse01/:/var/lib/clickhouse \ --volume=/etc/clickhouse-server01/:/etc/clickhouse-server/ \ --add-host server01:192.168.21.21 \ --add-host server02:192.168.21.69 \ --add-host server03:192.168.21.6 \ --add-host i-r9es2e0q:192.168.21.21 \ --add-host i-o91d619w:192.168.21.69 \ --add-host i-ldipmbwa:192.168.21.6 \ --hostname $(hostname) \ -p 9000:9000 \ -p 8123:8123 \ -p 9009:9009 \ yandex/clickhouse-server
2.3.6 運行clickhouse03-02實例
登錄到server03：ssh server03
docker run -d \ --name cs03-02 \ --ulimit nofile=262144:262144 \ --volume=/data/clickhouse02/:/var/lib/clickhouse \ --volume=/etc/clickhouse-server02/:/etc/clickhouse-server/ \ --add-host server01:192.168.21.21 \ --add-host server02:192.168.21.69 \ --add-host server03:192.168.21.6 \ --add-host i-r9es2e0q:192.168.21.21 \ --add-host i-o91d619w:192.168.21.69 \ --add-host i-ldipmbwa:192.168.21.6 \ --hostname $(hostname) \ -p 9001:9000 \ -p 8124:8123 \ -p 9010:9009 \ yandex/clickhouse-server
2.4 clickhouse集羣的數據操做
2.4.1 運行客戶端鏈接clickhouse server
1）隨便在哪一臺實例的機器上執行以下（連不一樣的clickhouse實例只須要改下host和port的值便可）
docker run -it \ --rm \ --add-host server01:192.168.21.21 \ --add-host server02:192.168.21.69 \ --add-host server03:192.168.21.6 \ yandex/clickhouse-client \ --host server01 \ --port 9000
2）執行下以下命令查看下配置信息：
#須要看下，是否和metrika.xml配置的分片和副本信息一致，若是不一致，須要check下每一個clickhouse-server實例的配置
#應該is_local顯示爲0，且分片和副本信息都正確
SELECT * FROM system.clusters;
2.4.2 建立本地複製表和分佈式表
全部實例配置完上面這些以後，分別執行啓動命令啓動，而後全部實例都執行下面語句建立數據庫：
例如在實例01-02上執行：
#執行Clickhouse client進入Clickhouse docker run -it --rm --add-host server01:192.168.21.21 --add-host server02:192.168.21.69 --add-host server03:192.168.21.6 \ yandex/clickhouse-client --host server01 --port 9001 ClickHouse client version 19.17.5.18 (official build). Connecting to server01:9001 as user default. Connected to ClickHouse server version 19.17.5 revision 54428. i-r9es2e0q :) CREATE DATABASE test; CREATE DATABASE test Ok.
2.4.3 建立複製表
而後對於全部實例分別建立對應的複製表，這裏測試建立一個簡單的表：
#在clickhouse01-01 9000上執行:
docker run -it --rm --add-host server01:192.168.21.21 --add-host server02:192.168.21.69 --add-host server03:192.168.21.6 \
yandex/clickhouse-client --host server01 --port 9000
#而後執行
CREATE TABLE test.device_thing_data (
time UInt64,
user_id String,
device_id String,
source_id String,
thing_id String,
identifier String,
value_int32 Int32,
value_float Float32,
value_double Float64,
value_string String,
value_enum Enum8('0'=0,'1'=1,'2'=2,'3'=3,'4'=4,'5'=5,'6'=6,'7'=7,'8'=8),
value_string_ex String,
value_array_string Array(String),
value_array_int32 Array(Int32),
value_array_float Array(Float32),
value_array_double Array(Float64),
action_date Date,
action_time DateTime
) Engine= ReplicatedMergeTree('/clickhouse/tables/01-01/device_thing_data','cluster01-01-1') PARTITION BY toYYYYMM(action_date) ORDER BY (user_id,device_id,thing_id,identifier,time,intHash64(time)) SAMPLE BY intHash64(time) SETTINGS index_granularity=8192
#在clickhouse01-02 9001上執行:
docker run -it --rm --add-host server01:192.168.21.21 --add-host server02:192.168.21.69 --add-host server03:192.168.21.6 \
yandex/clickhouse-client --host server01 --port 9001
而後執行
CREATE TABLE test.device_thing_data (
time UInt64,
user_id String,
device_id String,
source_id String,
thing_id String,
identifier String,
value_int32 Int32,
value_float Float32,
value_double Float64,
value_string String,
value_enum Enum8('0'=0,'1'=1,'2'=2,'3'=3,'4'=4,'5'=5,'6'=6,'7'=7,'8'=8),
value_string_ex String,
value_array_string Array(String),
value_array_int32 Array(Int32),
value_array_float Array(Float32),
value_array_double Array(Float64),
action_date Date,
action_time DateTime
) Engine= ReplicatedMergeTree('/clickhouse/tables/01-02/device_thing_data','cluster01-02-2') PARTITION BY toYYYYMM(action_date) ORDER BY (user_id,device_id,thing_id,identifier,time,intHash64(time)) SAMPLE BY intHash64(time) SETTINGS index_granularity=8192
#在clickhouse02-01 9000上執行:
docker run -it --rm --add-host server01:192.168.21.21 --add-host server02:192.168.21.69 --add-host server03:192.168.21.6 \
yandex/clickhouse-client --host server02 --port 9000
#而後執行
CREATE TABLE test.device_thing_data (
time UInt64,
user_id String,
device_id String,
source_id String,
thing_id String,
identifier String,
value_int32 Int32,
value_float Float32,
value_double Float64,
value_string String,
value_enum Enum8('0'=0,'1'=1,'2'=2,'3'=3,'4'=4,'5'=5,'6'=6,'7'=7,'8'=8),
value_string_ex String,
value_array_string Array(String),
value_array_int32 Array(Int32),
value_array_float Array(Float32),
value_array_double Array(Float64),
action_date Date,
action_time DateTime
) Engine= ReplicatedMergeTree('/clickhouse/tables/01-02/device_thing_data','cluster01-02-1') PARTITION BY toYYYYMM(action_date) ORDER BY (user_id,device_id,thing_id,identifier,time,intHash64(time)) SAMPLE BY intHash64(time) SETTINGS index_granularity=8192
#在clickhouse02-02 9001上執行:
docker run -it --rm --add-host server01:192.168.21.21 --add-host server02:192.168.21.69 --add-host server03:192.168.21.6 \
yandex/clickhouse-client --host server02 --port 9001
#而後執行
CREATE TABLE test.device_thing_data (
time UInt64,
user_id String,
device_id String,
source_id String,
thing_id String,
identifier String,
value_int32 Int32,
value_float Float32,
value_double Float64,
value_string String,
value_enum Enum8('0'=0,'1'=1,'2'=2,'3'=3,'4'=4,'5'=5,'6'=6,'7'=7,'8'=8),
value_string_ex String,
value_array_string Array(String),
value_array_int32 Array(Int32),
value_array_float Array(Float32),
value_array_double Array(Float64),
action_date Date,
action_time DateTime
) Engine= ReplicatedMergeTree('/clickhouse/tables/01-03/device_thing_data','cluster01-03-2') PARTITION BY toYYYYMM(action_date) ORDER BY (user_id,device_id,thing_id,identifier,time,intHash64(time)) SAMPLE BY intHash64(time) SETTINGS index_granularity=8192
#在clickhouse03-01 9000上執行:
docker run -it --rm --add-host server01:192.168.21.21 --add-host server02:192.168.21.69 --add-host server03:192.168.21.6 \
yandex/clickhouse-client --host server03 --port 9000
#而後執行
CREATE TABLE test.device_thing_data (
time UInt64,
user_id String,
device_id String,
source_id String,
thing_id String,
identifier String,
value_int32 Int32,
value_float Float32,
value_double Float64,
value_string String,
value_enum Enum8('0'=0,'1'=1,'2'=2,'3'=3,'4'=4,'5'=5,'6'=6,'7'=7,'8'=8),
value_string_ex String,
value_array_string Array(String),
value_array_int32 Array(Int32),
value_array_float Array(Float32),
value_array_double Array(Float64),
action_date Date,
action_time DateTime
) Engine= ReplicatedMergeTree('/clickhouse/tables/01-03/device_thing_data','cluster01-03-1') PARTITION BY toYYYYMM(action_date) ORDER BY (user_id,device_id,thing_id,identifier,time,intHash64(time)) SAMPLE BY intHash64(time) SETTINGS index_granularity=8192
#在clickhouse03-02 9001上執行:
docker run -it --rm --add-host server01:192.168.21.21 --add-host server02:192.168.21.69 --add-host server03:192.168.21.6 \
yandex/clickhouse-client --host server03 --port 9001
而後執行
CREATE TABLE test.device_thing_data (
time UInt64,
user_id String,
device_id String,
source_id String,
thing_id String,
identifier String,
value_int32 Int32,
value_float Float32,
value_double Float64,
value_string String,
value_enum Enum8('0'=0,'1'=1,'2'=2,'3'=3,'4'=4,'5'=5,'6'=6,'7'=7,'8'=8),
value_string_ex String,
value_array_string Array(String),
value_array_int32 Array(Int32),
value_array_float Array(Float32),
value_array_double Array(Float64),
action_date Date,
action_time DateTime
) Engine= ReplicatedMergeTree('/clickhouse/tables/01-01/device_thing_data','cluster01-01-2') PARTITION BY toYYYYMM(action_date) ORDER BY (user_id,device_id,thing_id,identifier,time,intHash64(time)) SAMPLE BY intHash64(time) SETTINGS index_granularity=8192
2.4.3 建立分佈式表（用於查詢）
而後建立完上面複製表以後，能夠建立分佈式表，分佈式表只是做爲一個查詢引擎，自己不存儲任何數據，查詢時將sql發送到全部集羣分片，而後進行進行處理和聚合後將結果返回給客戶端，所以clickhouse限制聚合結果大小不能大於分佈式表節點的內存，固然這個通常條件下都不會超過；分佈式表能夠全部實例都建立，也能夠只在一部分實例建立，這個和業務代碼中查詢的示例一致，建議設置多個，當某個節點掛掉時能夠查詢其餘節點上的表，分佈式表的建表語句以下：
#在clickhouse-server集羣上一次性建立全部的分佈式表，操做卡主了。緣由不明 CREATE TABLE device_thing_data_all ON CLUSTER cluster_3s_1r AS test.device_thing_data ENGINE = Distributed(cluster_3s_1r, default, device_thing_data, rand()) #以下這個須要每一個機器上都操做一遍 CREATE TABLE device_thing_data_all AS test.device_thing_data ENGINE = Distributed(cluster_3s_1r, test, device_thing_data, rand())
2.4.4 測試可用性
#客戶端鏈接到某個clickhouse-server實例（例如cs01-01）
#查詢分佈式表，此時沒有查詢到數據
select * from image_label_all;
#在cs01-01上執行以下查看是否有數據
#查詢本地複製表，此時沒有查詢到數據
select * from test.device_thing_data;
#往復製表中表裏插入一條數據
INSERT INTO test.device_thing_data; (user_id) VALUES ('1')
#因爲剛纔在cs01-01上插入一條數據，因此應該有數據了
select * from test.device_thing_data;
#查詢分佈式表,也有數據了
select * from image_label_all;
#在cs03-02上查詢複製表的數據（因爲cs03-02是cs01-01的副本，因此數據被自動同步過來了），因此應該有數據了
select * from test.device_thing_data;
#在cs03-01上插入一條數據，此時應該會把數據同步到cs02-02上
INSERT INTO test.device_thing_data (user_id) VALUES ('2')
#再次查詢分佈式表，此時應該查到cs03-01和cs01-01上的兩條數據
select * from image_label_all;
博主：測試生財（一個不爲996而996的測開碼農）
座右銘：專一測試開發與自動化運維，努力讀書思考寫做，爲內卷的人生奠基財務自由。
內容範疇：技術提高，職場雜談，事業發展，閱讀寫做，投資理財，健康人生。
csdn：https://blog.csdn.net/ccgshigao
博客園：https://www.cnblogs.com/qa-freeroad/
51cto：https://blog.51cto.com/14900374
在碼農的苦悶歲月裏，期待咱們一塊兒成長，歡迎關注，感謝拍磚！
相關文章
- 1. 3分鐘快速搭建Greenplum集羣
- 2. 搭建mongodb集羣（副本集+分片）
- 3. MongoDB3.6集羣搭建（分片+副本集）
- 4. MongoDB 分片副本集集羣搭建
- 5. mongodb3.6集羣搭建:分片+副本集
- 6. mongodb集羣搭建（分片+副本）
- 7. 快速搭建kafka集羣3臺
- 8. Linux Centos MongoDB分片集羣搭建：分片+副本集
- 9. 使用Docker Toolbox快速搭建3節點的shipyard集羣
- 10. ClickHouse 副本和分片
- 更多相關文章...
- • Swarm 集羣管理 - Docker教程
- • SQL 快速參考 - SQL 教程
- • Docker容器實戰(八) - 漫談 Kubernetes 的本質
- • Docker 清理命令
相關標籤/搜索
每日一句
-
每一个你不满意的现在，都有一个你没有努力的曾经。
歡迎關注本站公眾號,獲取更多信息
