ClickHouse 最近跟Es杠上了，日志场景谁更适合

yum install java-1.8.0-openjdk-devel.x86_64
/etc/profile 配置环境变量
更新系统时间
yum install  ntpdate
ntpdate asia.pool.ntp.org

mkdir zookeeper
mkdir ./zookeeper/data
mkdir ./zookeeper/logs

wget  --no-check-certificate https://mirrors.tuna.tsinghua.edu.cn/apache/zookeeper/zookeeper-3.7.1/apache-zookeeper-3.7.1-bin.tar.gz
tar -zvxf apache-zookeeper-3.7.1-bin.tar.gz -C /usr/zookeeper

export ZOOKEEPER_HOME=/usr/zookeeper/apache-zookeeper-3.7.1-bin
export PATH=$ZOOKEEPER_HOME/bin:$PATH

进入ZooKeeper配置目录
cd $ZOOKEEPER_HOME/conf

新建配置文件
vi zoo.cfg

tickTime=2000
initLimit=10
syncLimit=5
dataDir=/usr/zookeeper/data
dataLogDir=/usr/zookeeper/logs
clientPort=2181
server.1=zk1:2888:3888
server.2=zk2:2888:3888
server.3=zk3:2888:3888

在每台服务器上执行，给zookeeper创建myid
echo "1" > /usr/zookeeper/data/myid
echo "2" > /usr/zookeeper/data/myid
echo "3" > /usr/zookeeper/data/myid

进入ZooKeeper bin目录
cd $ZOOKEEPER_HOME/bin
sh zkServer.sh start

mkdir -p /usr/kafka
chmod 777 -R /usr/kafka
wget  --no-check-certificate https://mirrors.tuna.tsinghua.edu.cn/apache/kafka/3.2.0/kafka_2.12-3.2.0.tgz
tar -zvxf kafka_2.12-3.2.0.tgz -C /usr/kafka


不同的broker Id 设置不一样，比如 1,2,3
broker.id=1
listeners=PLAINTEXT://ip:9092
socket.send.buffer.bytes=102400
socket.receive.buffer.bytes=102400
socket.request.max.bytes=104857600
log.dir=/usr/kafka/logs
num.partitions=5
num.recovery.threads.per.data.dir=3
offsets.topic.replication.factor=2
transaction.state.log.replication.factor=3
transaction.state.log.min.isr=3
log.retention.hours=168
log.segment.bytes=1073741824
log.retention.check.interval.ms=300000
zookeeper.connect=zk1:2181,zk2:2181,zk3:2181
zookeeper.connection.timeout.ms=30000
group.initial.rebalance.delay.ms=0

后台常驻进程启动kafka
nohup /usr/kafka/kafka_2.12-3.2.0/bin/kafka-server-start.sh /usr/kafka/kafka_2.12-3.2.0/config/server.properties   >/usr/kafka/logs/kafka.log >&1 &

/usr/kafka/kafka_2.12-3.2.0/bin/kafka-server-stop.sh

$KAFKA_HOME/bin/kafka-topics.sh --list --bootstrap-server  ip:9092

$KAFKA_HOME/bin/kafka-console-consumer.sh --bootstrap-server ip:9092 --topic test --from-beginning

$KAFKA_HOME/bin/kafka-topics.sh  --create --bootstrap-server  ip:9092  --replication-factor 2 --partitions 3 --topic xxx_data

sudo rpm --import https://packages.elastic.co/GPG-KEY-elasticsearch

Create a file with a .repo extension (for example, elastic.repo) in your /etc/yum.repos.d/ directory and add the following lines:
在/etc/yum.repos.d/ 目录下创建elastic.repo

[elastic-8.x]
name=Elastic repository for 8.x packages
baseurl=https://artifacts.elastic.co/packages/8.x/yum
gpgcheck=1
gpgkey=https://artifacts.elastic.co/GPG-KEY-elasticsearch
enabled=1
autorefresh=1
type=rpm-md

yum install filebeat
systemctl enable filebeat
chkconfig --add filebeat

文件目录： /etc/filebeat/filebeat.yml

filebeat.inputs:
- type: log
  enabled: true
  paths:
    - /root/logs/xxx/inner/*.log
  json:  
如果不设置该索性，所有的数据都存储在message里面，这样设置以后数据会平铺。
       keys_under_root: true 
output.kafka:
  hosts: ["kafka1:9092", "kafka2:9092", "kafka3:9092"]
  topic: 'xxx_data_clickhouse'
  partition.round_robin:
            reachable_only: false
            required_acks: 1
            compression: gzip
processors: 
剔除filebeat 无效的字段数据
    - drop_fields:  
        fields: ["input", "agent", "ecs", "log", "metadata", "timestamp"]
        ignore_missing: false

nohup ./filebeat -e -c /etc/filebeat/filebeat.yml > /user/filebeat/filebeat.log & 
输出到filebeat.log文件中，方便排查

检查当前CPU是否支持SSE 4.2，如果不支持，需要通过源代码编译构建
grep -q sse4_2 /proc/cpuinfo && echo "SSE 4.2 supported" || echo "SSE 4.2 not supported"
返回 "SSE 4.2 supported" 表示支持，返回 "SSE 4.2 not supported" 表示不支持

创建数据保存目录，将它创建到大容量磁盘挂载的路径
mkdir -p /data/clickhouse
修改/etc/hosts文件，添加clickhouse节点
举例：
10.190.85.92 bigdata-clickhouse-01
10.190.85.93 bigdata-clickhouse-02

服务器性能参数设置：
cpu频率调节，将CPU频率固定工作在其支持的最高运行频率上，而不动态调节，性能最好
echo 'performance' | tee /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor

内存调节，不要禁用 overcommit
echo 0 | tee /proc/sys/vm/overcommit_memory

始终禁用透明大页(transparent huge pages)。 它会干扰内存分配器，从而导致显着的性能下降
echo 'never' | tee /sys/kernel/mm/transparent_hugepage/enabled

首先，需要添加官方存储库：
yum install yum-utils
rpm --import <https://repo.clickhouse.tech/CLICKHOUSE-KEY.GPG>
yum-config-manager --add-repo <https://repo.clickhouse.tech/rpm/stable/x86_64>

查看clickhouse可安装的版本：
yum list | grep clickhouse
运行安装命令：
yum -y install clickhouse-server clickhouse-client

修改/etc/clickhouse-server/config.xml配置文件，修改日志级别为information，默认是trace
<level>information</level>
执行日志所在目录：

正常日志
/var/log/clickhouse-server/clickhouse-server.log
异常错误日志
/var/log/clickhouse-server/clickhouse-server.err.log

查看安装的clickhouse版本：
clickhouse-server --version
clickhouse-client --password

sudo clickhouse stop
sudo clickhouse tart
sudo clickhouse start

CREATE TABLE default.kafka_clickhouse_inner_log ON CLUSTER clickhouse_cluster (
    log_uuid   String ,
    date_partition   UInt32 ,
    event_name   String ,
    activity_name   String ,
    activity_type   String ,
    activity_id   UInt16 
) ENGINE = Kafka SETTINGS
    kafka_broker_list = 'kafka1:9092,kafka2:9092,kafka3:9092',
    kafka_topic_list = 'data_clickhouse',
    kafka_group_name = 'clickhouse_xxx',
    kafka_format = 'JSONEachRow',
    kafka_row_delimiter = '\n',
    kafka_num_consumers = 1;

 需要在clickhouse client 创建加上 --stream_like_engine_allow_direct_select 1

 clickhouse-client --stream_like_engine_allow_direct_select 1 --password xxxxx

创建本地表（使用复制去重表引擎）
create table default.bi_inner_log_local ON CLUSTER clickhouse_cluster (
    log_uuid   String ,
    date_partition   UInt32 ,
    event_name   String ,
    activity_name   String ,
    credits_bring   Int16 ,
    activity_type   String ,
    activity_id   UInt16 
) ENGINE = ReplicatedReplacingMergeTree('/clickhouse/tables/default/bi_inner_log_local/{shard}','{replica}')
  PARTITION BY date_partition
  ORDER BY (event_name,date_partition,log_uuid)
   SETTINGS index_granularity = 8192;

 <macros>
        <shard>01</shard>
        <replica>example01-01-1</replica>
    </macros>

CREATE TABLE default.bi_inner_log_all ON CLUSTER clickhouse_cluster AS default.bi_inner_log_local
ENGINE = Distributed(clickhouse_cluster, default, bi_inner_log_local, xxHash32(log_uuid));

 <!--分布式表配置-->
 <remote_servers>
       <clickhouse_cluster> <!--集群名称, 可以自定义, 后面在新建库、表的时候需要用到集群名称-->
     <shard>
    <!--内部复制(默认false), 开启后, 在分布式表引擎下, 数据写入时-->
                        <!--每个分片只会去寻找一个节点写, 并不是每个都写-->
                        <internal_replication>true</internal_replication>
                        <replica>
                            <host>ip1</host>
                            <port>9000</port>
                                    <user>default</user>
                                    <password>xxxx</password>
                        </replica>
                    </shard>
                    <shard>
                        <internal_replication>true</internal_replication>
                        <replica>
                            <host>ip2</host>
                            <port>9000</port>
                                    <user>default</user>
                                    <password>xxxx</password>
                        </replica>
                    </shard>
                </clickhouse_cluster>
        </remote_servers>

CREATE MATERIALIZED VIEW default.view_bi_inner_log ON CLUSTER clickhouse_cluster TO default.bi_inner_log_all AS 
SELECT 
    log_uuid ,
date_partition ,
event_name ,
activity_name ,
credits_bring ,
activity_type ,
activity_id 
FROM default.kafka_clickhouse_inner_log;

-------------  END  -------------扫描下方二维码，加入技术群。暗号：加群