![image](https://img2022.cnblogs.com/blog/2222036/202205/2222036-20220530004609017-323802324.png)
1:环境
10.0.0.10 |
ElasticSearch_master kafka+zk-1 |
10.0.0.11 |
ElasticSearch_node_1 kafka+zk-1 |
10.0.0.12 |
ElasticSearch_node_2 kafka+zk-2 |
10.0.0.13 |
logstash kibana cerebro |
服务 |
版本 |
java |
1.8.0_221 |
elasticsearch |
7.10.1 |
filebeat |
7.10.1 |
kibana |
7.10.1 |
logstash |
7.10.1 |
cerebro |
0.9.2-1 |
kafka |
2.12-2.3.0 |
zookeeper |
3.5.6 |
2:参数配置
ES节点执行(三个)
# 配置主机名
hostnamectl set-hostname elk_master_1
hostnamectl set-hostname elk_node_1
hostnamectl set-hostname elk_node_2
# 增加文件描述符
cat << eof>> /etc/security/limits.conf
* soft nofile 65536
* hard nofile 65536
* soft nproc 65536
* hard nproc 65536
* hard memlock unlimited
* soft memlock unlimited
eof
# 修改默认限制内存
cat << eof>>/etc/systemd/system.conf
DefaultLimitNOFILE=65536
DefaultLimitNPROC=32000
DefaultLimitMEMLOCK=infinity
eof
# 优化内核,对es支持
cat << eof>>/etc/sysctl.conf
# 关闭交换内存
vm.swappiness =0
# 影响java线程数量,建议修改为262144或者更高
vm.max_map_count= 262144
# 优化内核listen连接
net.core.somaxconn=65535
# 最大打开文件描述符数,建议修改为655360或者更高
fs.file-max=655360
# 开启ipv4转发
net.ipv4.ip_forward= 1
eof
# 修改hosts配置文件
cat << eof>> /etc/hosts
10.0.0.10 elk_master_1
10.0.0.11 elk_node_1
10.0.0.12 elk_node_2
eof
# 重启服务器
3:部署Zookeeper
ES节点执行(三个节点)
# 创建Zookeeper项目目录
mkdir -p /usr/local/zk/{log,data}
# 下载ZK并解压
wget https://dlcdn.apache.org/zookeeper/zookeeper-3.5.9/apache-zookeeper-3.5.9-bin.tar.gz
tar xf apache-zookeeper-3.5.9-bin.tar.gz
mv apache-zookeeper-3.5.9-bin zookeeper
# 修改配置文件
[root@elk_master_1 conf]# pwd
/root/zookeeper/conf
[root@elk_master_1 conf]# cat zoo.cfg
tickTime=2000
syncLimit=5
dataDir=/usr/local/zk/data
dataLogDir=/usr/lcoal/zk/log
maxClientCnxns=60
initLimit=10
clientPort=2181
server.1=10.0.0.10:2888:3888
server.2=10.0.0.11:2888:3888
server.3=10.0.0.12:2888:3888
# 写入节点标记
# 分别在三个节点/usr/local/zk/myid写入节点标记
[root@elk_master_1 ~]# echo "1" > /usr/local/zk/data/myid
[root@elk_node_1 ~]# echo "2" > /usr/local/zk/data/myid
[root@elk_node_2 ~]# echo "3" > /usr/local/zk/data/myid
# 安装java
yum install -y java-1.8.0-openjdk java-1.8.0-openjdk-devel
# 启动zookeeper集群
cd zookeeper/bin/
./zkServer.sh start
/usr/bin/java
ZooKeeper JMX enabled by default
Using config: /root/zookeeper/bin/../conf/zoo.cfg
Starting zookeeper ... STARTED
# 检查集群状态
./zkServer.sh status
/usr/bin/java
ZooKeeper JMX enabled by default
Using config: /root/zookeeper/bin/../conf/zoo.cfg
Client port found: 2181. Client address: localhost. Client SSL: false.
Mode: leader
# 设置全局变量
cat << eof>>/etc/profile
export ZOOKEEPER_INSTALL=/root/zookeeper/
export PATH=\$PATH:\$ZOOKEEPER_INSTALL/bin
eof
# 这样就可以全局使用zkServer.sh命令了
4:部署 Kafka
ES节点执行(三个节点)
# 下载解压kafka压缩包
wget https://downloads.apache.org/kafka/2.8.1/kafka_2.13-2.8.1.tgz
tar xf kafka_2.13-2.8.1.tgz
mv kafka_2.13-2.8.1 kafka
# 创建数据目录
mkdir -p /usr/local/kafka/{data}
# 配置kafka
cd kafka/config/
cat server.properties
broker.id=1 # 此ID不能唯一
listeners=PLAINTEXT://:9092
num.network.threads=3
num.io.threads=8
socket.send.buffer.bytes=102400
socket.receive.buffer.bytes=102400
socket.request.max.bytes=104857600
log.dirs=/usr/local/kafka/data
num.partitions=3
num.recovery.threads.per.data.dir=1
log.retention.hours=168
log.segment.bytes=1073741824
log.retention.check.interval.ms=300000
zookeeper.connect=10.0.0.11,10.0.0.12,10.0.0.13
zookeeper.connection.timeout.ms=6000
############################# Server Basics #############################
# broker的id,值为整数,且必须唯一,在一个集群中不能重复
broker.id=1
############################# Socket Server Se:ttings #############################
# kafka默认监听的端口为9092 (默认与主机名进行连接)
listeners=PLAINTEXT://:9092
# 处理网络请求的线程数量,默认为3个
num.network.threads=3
# 执行磁盘IO操作的线程数量,默认为8个
num.io.threads=8
# socket服务发送数据的缓冲区大小,默认100KB
socket.send.buffer.bytes=102400
# socket服务接受数据的缓冲区大小,默认100KB
socket.receive.buffer.bytes=102400
# socket服务所能接受的一个请求的最大大小,默认为100M
socket.request.max.bytes=104857600
############################# Log Basics #############################
# kafka存储消息数据的目录
log.dirs=../kfkdata
# 每个topic默认的partition数量
num.partitions=3
# 在启动时恢复数据和关闭时刷新数据时每个数据目录的线程数量
num.recovery.threads.per.data.dir=1
############################# Log Flush Policy #############################
# 消息刷新到磁盘中的消息条数阈值
#log.flush.interval.messages=10000
# 消息刷新到磁盘中的最大时间间隔,1s
#log.flush.interval.ms=1000
############################# Log Retention Policy #############################
# 日志保留小时数,超时会自动删除,默认为7天
log.retention.hours=168
# 日志保留大小,超出大小会自动删除,默认为1G
#log.retention.bytes=1073741824
# 日志分片策略,单个日志文件的大小最大为1G,超出后则创建一个新的日志文件
log.segment.bytes=1073741824
# 每隔多长时间检测数据是否达到删除条件,300s
log.retention.check.interval.ms=300000
############################# Zookeeper #############################
# Zookeeper连接信息,如果是zookeeper集群,则以逗号隔开
zookeeper.connect=10.0.0.11,10.0.0.12,10.0.0.13
# 连接zookeeper的超时时间,6s
zookeeper.connection.timeout.ms=6000
# 启动kafka集群
cd kafka/bin/
./kafka-server-start.sh ../config/server.properties # 前台执行
./kafka-server-start.sh -daemon ../config/server.properties # 后台执行
# 任意节点测试
在创建topic在集群中的任意节点 发布消息订阅消息验证结果
./kafka-topics.sh --create --zookeeper 10.0.0.10:2181,10.0.0.11:2182,10.0.0.12:2181 --partitions 3 --replication-factor 1 --topic logs
5:部署elasticsearch
ES节点执行(三个节点)
# 下载安装elasticsearch
wget https://pan.cnsre.cn/d/Package/Linux/ELK/elasticsearch-7.10.1-x86_64.rpm # 可以从官方下载,我这里用的别人的盘
# 安装ES
rpm -ivh elasticsearch-7.10.1-x86_64.rpm
# 备份配置文件
cd /etc/elasticsearch/
cp elasticsearch.yml elasticsearch.yml.bak
# 创建目录
mkdir -p /usr/local/es/{data,snapshot,logs}
# 修改配置文件
#集群名
cluster.name: elk-cluster
#node名 不可重复
node.name: elk-1
#数据存储路径
path.data: /usr/local/es/data
#数据快照路径
path.repo: /usr/local/es/snapshot
#日志存储路径
path.logs: /usr/local/es/logs
#es绑定的ip地址,根据自己机器ip进行修改
network.host: 0.0.0.0
#服务端口
http.port: 9200
#集群master需要和node名设置一致
discovery.seed_hosts: ["10.0.0.10", "10.0.0.11", "10.0.0.12"]
cluster.initial_master_nodes: ["10.0.0.10", "10.0.0.11", "10.0.0.12"]
#允许跨域请求
http.cors.enabled: true
#* 表示支持所有域名
http.cors.allow-origin: "*"
#添加请求header
http.cors.allow-headers: Authorization,X-Requested-With,Content-Length,Content-Type
#生产必须为true,内存锁定检查,目的是内存地址直接映射,减少一次copy时间
bootstrap.memory_lock: true
#系统过滤检查,防止数据损坏,考虑集群安全,生产设置成false
bootstrap.system_call_filter: false
#xpack配置
xpack.security.enabled: true
xpack.security.transport.ssl.enabled: true
xpack.security.transport.ssl.verification_mode: certificate
xpack.security.transport.ssl.keystore.path: /etc/elasticsearch/elastic-certificates.p12
xpack.security.transport.ssl.truststore.path: /etc/elasticsearch/elastic-certificates.p12
eof
# 精简
cluster.name: elk-cluster
node.name: elk-1 # 此名称必须唯一
path.data: /usr/local/es/data
path.repo: /usr/local/es/snapshot
path.logs: /usr/local/es/logs
network.host: 0.0.0.0
http.port: 9200
discovery.seed_hosts: ["10.0.0.10", "10.0.0.11", "10.0.0.12"]
cluster.initial_master_nodes: ["10.0.0.10", "10.0.0.11", "10.0.0.12"]
http.cors.enabled: true
http.cors.allow-origin: "*"
http.cors.allow-headers: Authorization,X-Requested-With,Content-Length,Content-Type
bootstrap.memory_lock: true
bootstrap.system_call_filter: false
xpack.security.enabled: false # 从这里往下都没开
xpack.security.transport.ssl.enabled: false
xpack.security.transport.ssl.verification_mode: certificate
xpack.security.transport.ssl.keystore.path: /etc/elasticsearch/elastic-certificates.p12
xpack.security.transport.ssl.truststore.path: /etc/elasticsearch/elastic-certificates.p12
# 修改JVM
[root@elk_master_1 elasticsearch]# cat jvm.options | grep 1g
-Xms1g
-Xmx1g
将jvm.options文件中22-23行的1g设置为你的服务内存的一半
# 分配权限
chown -R elasticsearch:elasticsearch /usr/local/es/
# 启动服务
# 三个节点全部启动并加入开机启动
systemctl start elasticsearch.service
systemctl enable elasticsearch.service
# 使用xpack进行安全认证
# xpack的安全功能
1:TLS 功能。 可对通信进行加密
2:文件和原生 Realm。 可用于创建和管理用户
3:基于角色的访问控制。 可用于控制用户对集群 API 和索引的访问权限
4:通过针对 Kibana Spaces 的安全功能,还可允许在Kibana 中实现多租户。
# 但我这里并没有开启认证,但是可以操作一下
# 在任意一台机器操作
/usr/share/elasticsearch/bin/elasticsearch-certutil ca
/usr/share/elasticsearch/bin/elasticsearch-certutil cert --ca elastic-stack-ca.p12
# 两条命令均一路回车即可,不需要给秘钥再添加密码
# 证书创建完成之后,默认在es的数据目录。
# 将证书拷贝到etc下,并给上权限。
ll /usr/share/elasticsearch/elastic-*
-rw------- 1 root root 3451 May 29 14:38 /usr/share/elasticsearch/elastic-certificates.p12
-rw------- 1 root root 2527 May 29 14:38 /usr/share/elasticsearch/elastic-stack-ca.p12
cp /usr/share/elasticsearch/elastic-* /etc/elasticsearch/
chown elasticsearch:elasticsearch /etc/elasticsearch/elastic-*
# 做完之后,将证书拷贝到其他节点
scp /etc/elasticsearch/elastic-certificates.p12 root@elk_node_1:/etc/elasticsearch/
scp /etc/elasticsearch/elastic-stack-ca.p12 root@elk_node_1:/etc/elasticsearch/
scp /etc/elasticsearch/elastic-certificates.p12 root@elk_node_2:/etc/elasticsearch/
scp /etc/elasticsearch/elastic-stack-ca.p12 root@elk_node_2:/etc/elasticsearch/
# 然后上面的xpack就可以打开了
---
xpack.security.enabled: true
xpack.security.transport.ssl.enabled: true
xpack.security.transport.ssl.verification_mode: certificate
xpack.security.transport.ssl.keystore.path: /etc/elasticsearch/elastic-certificates.p12
xpack.security.transport.ssl.truststore.path: /etc/elasticsearch/elastic-certificates.p12
# 如果重启有问题,记得查一下拷贝过去证书的属主
# 为内置账号添加密码
ES中内置了几个管理其他集成组件的账号apm_system, beats_system, elastic, kibana, logstash_system, remote_monitoring_user使用之前,首先需要设置下密码。
/usr/share/elasticsearch/bin/elasticsearch-setup-passwords interactive
# 这里统一设置为 devopsdu 生产千万别这么玩
[root@elk_master_1 elasticsearch]# /usr/share/elasticsearch/bin/elasticsearch-setup-passwords interactive
Initiating the setup of passwords for reserved users elastic,apm_system,kibana,kibana_system,logstash_system,beats_system,remote_monitoring_user.
You will be prompted to enter passwords as the process progresses.
Please confirm that you would like to continue [y/N]y
Enter password for [elastic]:
Reenter password for [elastic]:
Enter password for [apm_system]:
Reenter password for [apm_system]:
Enter password for [kibana_system]:
Reenter password for [kibana_system]:
Enter password for [logstash_system]:
Reenter password for [logstash_system]:
Enter password for [beats_system]:
Reenter password for [beats_system]:
Enter password for [remote_monitoring_user]:
Reenter password for [remote_monitoring_user]:
Changed password for user [apm_system]
Changed password for user [kibana_system]
Changed password for user [kibana]
Changed password for user [logstash_system]
Changed password for user [beats_system]
Changed password for user [remote_monitoring_user]
Changed password for user [elastic]
6:部署 Cerebro
# 下载安装 (Kibana主机部署)
wget https://pan.cnsre.cn/d/Package/Linux/ELK/cerebro-0.9.2-1.noarch.rpm
rpm -ivh cerebro-0.9.2-1.noarch.rpm
# 修改配置文件
# 修改/etc/cerebro/application.conf配置文件
# 找到对应配置修改为以下内容
data.path: "/var/lib/cerebro/cerebro.db"
# data.path = "./cerebro.db"
hosts = [
{
host = "http://10.0.0.10:9200"
name = "elk-cluster"
auth = {
username = "elstic"
password = "devops"
}
}
]
# 安装 java
yum install -y java-1.8.0-openjdk java-1.8.0-openjdk-devel
# 启动
systemctl status cerebro && systemctl enable cerebro
netstat -nplt | grep 9000
# 访问测试
![image](https://img2022.cnblogs.com/blog/2222036/202205/2222036-20220531005553133-425413897.png)
![image](https://img2022.cnblogs.com/blog/2222036/202205/2222036-20220531005651949-1858987696.png)
9:部署Kibana
Kibana 节点安装
# 下载安装
wget https://artifacts.elastic.co/downloads/kibana/kibana-7.10.1-x86_64.rpm
rpm -ivh kibana-7.10.1-x86_64.rpm
# 修改备份配置文件
cd /etc/kibana/
cp kibana.yml kibana.yml.bak
[root@kibana kibana]# cat kibana.yml
server.port: 5601
server.host: 0.0.0.0
elasticsearch.hosts: ["http://10.0.0.10:9200/","http://10.0.0.11:9200/","http://10.0.0.12:9200/"] # ES主机
elasticsearch.username: "elastic" # 前面装ES配置的账号
elasticsearch.password: "devopsdu" # 配置的密码
i18n.locale: "zh-CN"
# 启动服务器
systemctl start kibana.service
systemctl enable kibana.service
# 检查服务
netstat -nplt | grep 5601
# 测试访问
![image](https://img2022.cnblogs.com/blog/2222036/202205/2222036-20220531010436557-1976132474.png)
![image](https://img2022.cnblogs.com/blog/2222036/202205/2222036-20220531010515159-1109714905.png)
前面使用elastic登录
账号:elastic
密码:devopsdu
10:部署filebeat
Kibana节点部署
# 下载filebeat
wget https://pan.cnsre.cn/d/Package/Linux/ELK/filebeat-7.10.1-x86_64.rpm
rpm -ivh filebeat-7.10.1-x86_64.rpm
# 备份配置并修改
cd /etc/filebeat/
cp filebeat.yml filebeat.yml.bak
# 修改配置
max_procs: 1 # 限制filebeat的进程数量,其实就是内核数,避免过多抢占业务资源
queue.mem.events: 256 # 存储于内存队列的事件数,排队发送 (默认4096)
queue.mem.flush.min_events: 128 # 小于queue.mem.events ,增加此值可提高吞吐量 (默认值2048)
filebeat.inputs: # inputs为复数,表名type可以有多个
- type: log # 输入类型
enable: true # 启用这个type配置
paths:
- /var/log/nginx/access.log # 收集nginx的access日志
json.keys_under_root: true # 默认Flase,还会将json解析的日志存储至messages字段
json.overwrite_keys: true # 覆盖默认的key,使用自定义json格式的key
max_bytes: 20480 # 单条日志的大小限制,建议限制(默认为10M,queue.mem.events * max_bytes 将是占有内存的一部)
fields: # 额外的字段
source: nginx-access # 自定义source字段,用于es建议索引(字段名小写,大写貌似不太行)
- type: log # 输入类型
enable: true # 启用这个type配置
paths:
- /var/log/nginx/error.log # 收集nginx的error日志
json.keys_under_root: true # 默认Flase,还会将json解析的日志存储至messages字段
json.overwrite_keys: true # 覆盖默认的key,使用自定义json格式的key
max_bytes: 20480 # 单条日志的大小限制,建议限制(默认为10M,queue.mem.events * max_bytes 将是占有内存的一部)
fields: # 额外的字段
source: nginx-error # 自定义source字段,用于es建议索引(字段名小写,大写貌似不太行)
setup.ilm.enabled: false # 自定义es的索引需要把ilm设置为false
output.kafka: # 输出到kafka
enabled: true # 该output配置是否启用
hosts: ["10.0.0.10:9092","10.0.0.11:9092","10.0.0.12:9092"] # kafka节点列表
topic: 'logstash-%{[fields.source]}' # kafka会创建该topic,然后logstash(可以过滤修改)会传给es作为索引名称
partition.hash:
reachable_only: true # 是否只发往可达分区
compression: gzip # 压缩
max_message_bytes: 1000000 # Event最大字节数。默认1000000。应小于等于kafka broker message.max.bytes值
required_acks: 1 # kafka ack等级
worker: 1 # kafka output的最大并发数
bulk_max_size: 2048 # 单次发往kafka的最大事件数
logging.to_files: true # 输出所有日志到file,默认true, 达到日志文件大小限制时,日志文件会自动限制替换
close_older: 30m # 如果文件在某个时间段内没有发生过更新,则关闭监控的文件handle。默认1h
force_close_files: false # 这个选项关闭一个文件,当文件名称的变化。只在window建议为true
close_inactive: 1m # 没有新日志采集后多长时间关闭文件句柄,默认5分钟,设置成1分钟,加快文件句柄关闭
close_timeout: 3h # 传输了3h后荏没有传输完成的话就强行关闭文件句柄,这个配置项是解决以上案例问题的key point
clean_inactive: 72h # 这个配置项也应该配置上,默认值是0表示不清理,不清理的意思是采集过的文件描述在registry文件里永不清理,在运行一段时间后,registry会变大,可能会带来问题
ignore_older: 70h # 设置了clean_inactive后就需要设置ignore_older,且要保证ignore_older < clean_inactive
# 启动服务
systemctl start filebeat.service && systemctl enable filebeat.service
11:部署logstash
Kibana 节点部署
# 下载安装
wget https://pan.cnsre.cn/d/Package/Linux/ELK/logstash-7.10.1-x86_64.rpm
rpm -ivh logstash-7.10.1-x86_64.rpm
# 备份并修改配置
cd /etc/logstash/
cp logstash.yml logstash.yml.bak
# 修改配置文件
[root@kibana logstash]# cat logstash.yml
http.host: "0.0.0.0"
# 指发送到Elasticsearch的批量请求的大小,值越大,处理则通常更高效,但增加了内存开销
pipeline.batch.size: 3000
# 指调整Logstash管道的延迟,过了该时间则logstash开始执行过滤器和输出
pipeline.batch.delay: 200
# 新增配置文件,从kafka获取日志
input { # 输入组件
kafka { # 从kafka消费数据
bootstrap_servers => ["10.0.11.172:9092,10.0.21.117:9092,10.0.11.208:9092"]
codec => "json" # 数据格式
#topics => ["3in1-topi"] # 使用kafka传过来的topic
topics_pattern => "logstash-.*" # 使用正则匹配topic
consumer_threads => 3 # 消费线程数量
decorate_events => true # 可向事件添加Kafka元数据,比如主题、消息大小的选项,这将向logstash事件中添加一个名为kafka的字段
auto_offset_reset => "latest" # 自动重置偏移量到最新的偏移量
#group_id => "logstash-node" # 消费组ID,多个有相同group_id的logstash实例为一个消费组
#client_id => "logstash1" # 客户端ID
fetch_max_wait_ms => "1000" # 指当没有足够的数据立即满足fetch_min_bytes时,服务器在回答fetch请求之前将阻塞的最长时间
}
}
filter{
# 当非业务字段时,无traceId则移除
#if ([message] =~ "traceId=null") { # 过滤组件,这里只是展示,无实际意义,根据自己的业务需求进行过滤
# drop {}
#}
mutate {
convert => ["Request time", "float"]
}
if [ip] != "-" {
geoip {
source => "ip"
target => "geoip"
# database => "/usr/share/GeoIP/GeoIPCity.dat"
add_field => [ "[geoip][coordinates]", "%{[geoip][longitude]}" ]
add_field => [ "[geoip][coordinates]", "%{[geoip][latitude]}" ]
}
mutate {
convert => [ "[geoip][coordinates]", "float"]
}
}
}
output { # 输出组件
elasticsearch { # Logstash输出到es
hosts => ["10.0.11.172:9200","10.0.21.117:9200","10.0.11.208:9200"]
index => "logstash-%{[fields][source]}-%{+YYYY-MM-dd}" # 直接在日志中匹配
#index => "%{[@metadata][topic]}-%{+YYYY-MM-dd}" # 以日期建索引
user => "elastic"
password => "123"
}
#stdout {
# codec => rubydebug
#}
}
input {
kafka {
bootstrap_servers => ["10.0.0.10:9092,10.0.0.11:9092,10.0.0.12:9092"]
codec => "json"
topics_pattern => "logstash-.*"
consumer_threads => 3
decorate_events => true
auto_offset_reset => "latest"
fetch_max_wait_ms => "1000"
}
}
mutate {
convert => ["Request time", "float"]
}
if [ip] != "-" {
geoip {
source => "ip"
target => "geoip"
add_field => [ "[geoip][coordinates]", "%{[geoip][longitude]}" ]
add_field => [ "[geoip][coordinates]", "%{[geoip][latitude]}" ]
}
mutate {
convert => [ "[geoip][coordinates]", "float"]
}
}
}
output {
elasticsearch {
hosts => ["10.0.0.10:9200","10.0.0.11:9200","10.0.0.12:9200"]
index => "logstash-%{[fields][source]}-%{+YYYY-MM-dd}"
user => "elastic"
password => "devopsdu"
}
}
# 测试接收日志
/usr/share/logstash/bin/logstash -f /etc/logstash/conf.d/get-kafka-logs.conf、
# 登录kibana创建索引
# 选择管理–索引模式–创建索引模式
![image](https://img2022.cnblogs.com/blog/2222036/202205/2222036-20220531020706352-1177113832.png)