docker方式部署elk日志搜索平台
Docker部署ELKF操作文档
前提介绍
1、之前搭建elk+f+k使用原生系统软件安装方式,由于docker镜像日趋成熟,docker官网和elastic官网都有相关镜像和各自安装文档可供参考,各个版本也在定期更新,这次决定换用docker方式进行搭建安装。
Docker(elk)的hub网站链接及文档:
https://hub.docker.com/r/sebp/elk
https://elk-docker.readthedocs.io/
2、搭建前准备:
[由于公司资源有限,开一台虚拟机放置elk所有插件,若有足够资源,可考虑使用分布式部署及es集群方式]
空余主机一台(内存>=6G):
Linux Centos7.6
用到的主要软件有:
Elasticsearch7.0.0 (搜索引擎-server端)
Kibana7.0.0 (图形化web界面-server端)
Logstash7.0.0 (log的汇总与收集-server端)
Filebeats7.0.1 (log收集-client端) 或 Metricbeat-7.0.1(client端)
一、安装docker及elk相关软件
yum list | grep docker
yum makecache fast
###移除旧版本###
yum remove docker docker-client docker-client-latest docker-common docker-latest docker-latest-logrotate docker-logrotate docker-engine
yum install yum-utils device-mapper-persistent-data lvm2 -y
ls /etc/yum.repos.d/
yum-config-manager –add-repo https://download.docker.com/linux/centos/docker-ce.repo
ls /etc/yum.repos.d/docker-ce.repo
###只开启需要安装项###
yum-config-manager –disable docker-ce-nightly
yum-config-manager –disable docker-ce-test
yum-config-manager –enable docker-ce
yum install docker-ce docker-ce-cli containerd.io
systemctl status docker
systemctl start docker
systemctl enable docker
systemctl status docker
docker –version
docker info
###调整系统内核###
sysctl -w vm.max_map_count=262144
###搜索相关镜像###
docker search elk
docker search sebp/elk
docker pull sebp/elk:700
echo $?
docker ps
docker rmi b0e0bd2a140b #删除镜像
docker images
docker rm 30fe8fb61024 #删除容器
docker ps -a
docker run -p 5601:5601 -p 9200:9200 -p 5044:5044 -v /home/elk-data:/elk-data -it –name elk sebp/elk:700 –d(后台守护启动)
docker run -p 5601:5601 -p 9200:9200 -p 5044:5044 -v /home/elk-data:/elk-data -it -d –name elk sebp/elk:700
netstat -anptu|grep 5601
netstat -anptu|grep 9200
###进入后台运行容器###
docker exec -it elk /bin/bash
exit退出
二、采集端安装Filebeat
(这里是用的原生软件安装)
#curl -L -O https://artifacts.elastic.co/downloads/beats/filebeat/filebeat-7.0.1-linux-x86_64.tar.gz
tar -zxvf filebeat-7.0.1-linux-x86_64.tar.gz
mv filebeat-7.0.1-linux-x86_64 /home/
cd /home/
mv filebeat-7.0.1-linux-x86_64/ filebeat7.0
cd filebeat7.0/
cat filebeat.yml | grep -vE ‘^$|#’
#filebeat配置,主要收集nginx、php和php-slow日志
filebeat.inputs:
– type: log
paths:
– /home/logs/hao.log
fields:
software: nginx
logname: hao
servername: “server-81”
fields_under_root: true
– type: log
paths:
– /home/logs/zf.error.log
fields:
software: nginx-error
logname: zf-error
servername: “server-81”
fields_under_root: true
– type: log
paths:
– /usr/local/php/var/log/php-fpm7.log
fields:
software: php
logname: php-fpm
servername: “server-81”
fields_under_root: true
– type: log
paths:
– /usr/local/php/var/log/slow.log
multiline.pattern: ‘^\[[0-9]{2}’
multiline.negate: true
multiline.match: after
fields:
software: php
logname: php-slow
servername: “server-81”
fields_under_root: true
#默认
filebeat.config.modules:
path: ${path.config}/modules.d/*.yml
reload.enabled: false
setup.template.settings:
index.number_of_shards: 1
setup.kibana:
#数据输出到logstash[此ip为公司外网映射到内网elk的ip]
output.logstash:
hosts: [“222.111.11.22:5044”]
#默认
processors:
– add_host_metadata: ~
– add_cloud_metadata: ~
###启动filebeat###
./filebeat -c filebeat.yml &
ps -ef|grep filebeat
netstat -anptu |grep 5044
###统一nginx日志格式,其他php采用正则匹配###
vim /etc/nginx/nginx.conf #主配置格式#子配置access … main;
…
log_format main escape=json
‘{“nx_localtime@timestamp”:”$time_local”,’
‘”nx_host”:”$server_addr”,’
‘”nx_client_ip”:”$remote_addr”,’
‘”nx_body_size”:$body_bytes_sent,’
‘”nx_request_time”:$request_time,’
‘”nx_scheme”:”$scheme”,’
‘”nx_http_host”:”$host”,’
‘”nx_request_method”:”$request_method”,’
‘”nx_uri”:”$uri”,’
‘”nx_status”:$status,’
‘”nx_referer”:”$http_referer”,’
‘”nx_agent”:”$http_user_agent”,’
‘”nx_upstream_host”:”$upstream_addr”,’
‘”nx_upstream_time”:$upstream_response_time,’ ‘”nx_upstream_response_length”:$upstream_response_length,’
‘”nx_upstream_status”:$upstream_status,’
‘”nx_upstream_connect_time”:”$upstream_connect_time”}’;
…
三、容器内部操作及配置文件
#docker exec –it elk /bin/bash
ls /elk-data/
cd opt/logstash/config/
cp logstash-sample.conf logstash.conf
mv /etc/logstash/*.conf /etc/logstash/*.conf.bak
vim logstash.conf
# Sample Logstash configuration for creating a simple
# Beats -> Logstash -> Elasticsearch pipeline.
input {
beats {
codec => json
port => 5044
host => “0.0.0.0”
client_inactivity_timeout => 300
ssl => false
}
}
filter {
if [software] == “nginx-error” {
grok {
match => { “message” => “(?<timestamp>%{YEAR}[./-]%{MONTHNUM}[./-]%{MONTHDAY}[- ]%{TIME}) \[%{LOGLEVEL:level}\] %{POSINT:pid}#%{NUMBER}: %{DATA:code} %{GREEDYDATA:event}(?:, client: (?<client_ip>%{IP}|%{HOSTNAME}))(?:, server: %{IPORHOST:server_website}?)(?:, request: \”%{WORD:http_method})? (?:%{NOTSPACE:request_uri})? (?:%{NOTSPACE:http_version}\”)?(?:, upstream: (?<upstream_ori>\”%{URI}\”|%{QS}))?(?:, host: %{QS:request_host})?(?:, referrer: \”%{URI:referrer_host}\”)?” }
remove_field => [“code”,”[prospector][type]”]
}
}
if [software] == “php” and [logname] == “php-fpm” {
grok {
match => [ “message”,”\[%{TIMESTAMP_ISO8601:logtime}\] %{WORD:env}\.(?<level>[A-Z]{4,5})\: %{GREEDYDATA:msg}}” ]
}
}
if [software] == “php” and [logname] == “php-slow” {
grok {
match => { “message” => “^\[%{DATA:time_local}\].*?script_filename\s+=\s+%{DATA:script_filename}\n%{GREEDYDATA:msg}” }
}
}
mutate {
remove_field => [“@version”,”tags”,”[ecs][version]”,”[agent][hostname]”,”[agent][ephemeral_id]”,”[agent][id]”,”[agent][type]”,”[agent][version]”,”[host][containerized]”,”[host][hostname]”,”[host][os][codename]”,”[host][os][family]”,”[host][os][platform]”]
#remove_field => “message”
}
}
-
- output {
#输出nginx的访问日志
if [software] == “nginx” {
elasticsearch {
hosts => [“http://localhost:9200”]
index => “nginx-%{logname}-%{+YYYY.MM.dd}”
}
}
#输出nginx的ERROR日志
if [software] == “nginx-error” {
elasticsearch {
hosts => [“http://localhost:9200”]
index => “nginx-%{logname}-%{+YYYY.MM.dd}”
}
}
#输出php慢日志
if [logname] == “php-slow” {
elasticsearch {
hosts => [“http://localhost:9200”]
index => “%{logname}-%{+YYYY.MM.dd}”
}
}
#输出level为ERROR的php日志
if [logname] == “php-fpm” and [level] == “ERROR” {
elasticsearch {
hosts => [“http://localhost:9200”]
index => “%{logname}-error-%{+YYYY.MM.dd}”
}
}
#输出非ERROR的php日志
if [logname] == “php-fpm” and [level] != “ERROR” {
elasticsearch {
hosts => [“http://localhost:9200”]
index => “%{logname}-%{+YYYY.MM.dd}”
}
}
#stdout { codec => rubydebug } #将此项打开屏幕输出模式
}
vim pipelines.yml
– pipeline.id: main
#path.config: “/etc/logstash/conf.d/*.conf”
path.config: “/opt/logstash/config/*.conf”
cd /opt/elasticsearch/config/
#由于不是集群,所以不需要太多配置
vim elasticsearch.yml
network.host: 0.0.0.0 127.0.0.1
cd /opt/kibana/config/
vim kibana.yml
server.host: “0.0.0.0”
#其他默认即可。
#启动ELK
/etc/init.d/elasticsearch start
/etc/init.d/kibana start
/etc/init.d/logstash start
#检测logstash配置语法是否错误
/opt/logstash/bin/logstash -f /opt/logstash/config/logstash.conf -t
#屏幕输出数据模式
/opt/logstash/bin/logstash -f /opt/logstash/config/logstash.conf
#查看状态正常,没有报错,可打开本机的5601端口访问浏览器界面。
四、API索引
<根据elasticsearch提供的API方式管理索引>
查询本机索引:
curl -XGET ‘localhost:9200/_cat/indices/?v’
删除索引:
curl -XDELETE ‘localhost:9200/nginx-test-2019.06*’
删除多个:
curl -DELETE ‘localhost:9200/_index1,_index2’
五、logstash-filter-grok插件(正则匹配)
# SYNTAX代表匹配值的类型,如NUMBER、WORD;SEMANTIC表示存储该值的一个变量名称
基础语法:%{SYNTAX:SEMANTIC}
# field_name表示存储该值的一个变量名称;后面跟上正则表达式;如:(?<queue_id>[0-9A-F]{10,11})
自定义语法:(?<field_name>the pattern here)
例如
配置:
filter {
grok {
match => {
“message” => “%{IP:client} %{WORD:method} %{URIPATHPARAM:request} %{NUMBER:bytes} %{NUMBER:duration}” }
}
}
源字段:
55.3.244.1 GET /index.html 15824 0.043
结果:
{
“method” => “GET”,
“message” => “58.23.56.101 GET /index.html 15824 0.043”,
“duration” => “0.043”,
“request” => “/index.html”,
“client” => “58.23.56.101”,
“bytes” => “15824”,
“@timestamp” => 2019-03-06T06:24:21.333Z
}
六、其他
<去掉 _id, _type, _index, _score, _source页面字段?>
Management -> Kibana -> Advanced Settings -> metafields, 去掉_id, _type, _index,_score, 留下_source即可,如果全部去掉则左侧无选项显示。