通过ELK快速搭建一个你可能需要的集中化日志平台

在项目初期的时候，大家都是赶着上线，一般来说对日志没有过多的考虑，当然日志量也不大，所以用log4net就够了，随着应用的越来越多，日志散

落在各个服务器的logs文件夹下，确实有点不大方便，这个时候就想到了，在log4net中配置 mysql的数据源，不过这里面有一个坑，熟悉log4net的同学知

道写入mysql有一个batch的阈值，比如说batchcache中有100条，才写入mysql，这样的话，就有一个延迟的效果，而且如果batchcache中不满100条的话，

你在mysql中是看不到最新的100条日志。。。而且采用中心化的mysql，涉及到tcp传输，其中的性能大家也应该明白，而且mysql没有一个好的日志界面，

只能自己去写UI，所以还还得继续寻找其他的解决方案，也就是本篇的ELK。

一：ELK名字解释

　   ELK就是ElasticSearch + LogStash + Kibana，这三样搭配起来确实非常不错，先画张图给大家看一下。

1. LogStash

它可以流放到各自的服务器上收集Log日志，通过内置的ElasticSearch插件解析后输出到ES中。

2.ElasticSearch

这是一个基于Lucene的分布式全文搜索框架，可以对logs进行分布式存储，有点像hdfs哈。。。

3. Kibana

所有的log日志都到ElasticSearch之后，我们需要给他展示出来，对吧？ 这个时候Kibana就出手了，它可以多维度的展示es中的数据。这也解决了

用mysql存储带来了难以可视化的问题。

二：快速搭建

上面只是名词解释，为了演示，我只在一台centos上面搭建了。

1.  官方下载 :https://www.elastic.co/cn/products，在下面这张图上，我们找到对应的三个产品，进行下载就好了。

[root@slave1 myapp]# ls
elasticsearch               kafka_2.-1.0..tgz              nginx-1.13..tar.gz
elasticsearch-5.6.4.tar.gz  kibana                            node
elasticsearch-head          kibana-5.2.0-linux-x86_64.tar.gz  node-v8.9.1-linux-x64.tar.xz
images                      logstash                          portal
java                        logstash-5.6.3.tar.gz             service
jdk1.                      logstash-tutorial-dataset         sql
jdk-8u144-linux-x64.tar.gz  nginx
kafka                       nginx-1.13.
[root@slave1 myapp]# 

我这里下载的是elasticsearch 5.6.4，kibana5.2.0 ，logstash5.6.3三个版本。。。然后用 tar -xzvf解压一下。

2. logstash配置

解压完之后，我们到config目录中新建一个logstash.conf配置。

[root@slave1 config]# ls
jvm.options  log4j2.properties  logstash.conf  logstash.yml  startup.options
[root@slave1 config]# pwd
/usr/myapp/logstash/config
[root@slave1 config]# vim logstash.conf

然后做好input ，filter，output三大块， 其中input是吸取logs文件下的所有log后缀的日志文件，filter是一个过滤函数，这里不用配置，output配置了导入到

hosts为127.0.0.1:9200的elasticsearch中，每天一个索引。

input {
     file {
        type => "log"
        path => "/logs/*.log"
        start_position => "beginning"
    }
}

output {
  stdout {
   codec => rubydebug { }
  }

  elasticsearch {
    hosts => "127.0.0.1"
    index => "log-%{+YYYY.MM.dd}"
  }
}

配置完了之后，我们就可以到bin目录下启动logstash了，配置文件设置为conf/logstash.conf，从下图中可以看到，当前开启的是9600端口。

[root@slave1 bin]# ls
cpdump             logstash      logstash.lib.sh  logstash-plugin.bat  setup.bat
ingest-convert.sh  logstash.bat  logstash-plugin  ruby                 system-install
[root@slave1 bin]# ./logstash -f ../config/logstash.conf
Sending Logstash's logs to /usr/myapp/logstash/logs which is now configured via log4j2.properties
[--28T17::,][INFO ][logstash.modules.scaffold] Initializing module {:module_name=>"fb_apache", :directory=>"/usr/myapp/logstash/modules/fb_apache/configuration"}
[--28T17::,][INFO ][logstash.modules.scaffold] Initializing module {:module_name=>"netflow", :directory=>"/usr/myapp/logstash/modules/netflow/configuration"}
[--28T17::,][INFO ][logstash.outputs.elasticsearch] Elasticsearch pool URLs updated {:changes=>{:removed=>[], :added=>[http://127.0.0.1:9200/]}}
[--28T17::,][INFO ][logstash.outputs.elasticsearch] Running health check to see if an Elasticsearch connection is working {:healthcheck_url=>http://127.0.0.1:9200/, :path=>"/"}
[--28T17::,][WARN ][logstash.outputs.elasticsearch] Restored connection to ES instance {:url=>"http://127.0.0.1:9200/"}
[--28T17::,][INFO ][logstash.outputs.elasticsearch] Using mapping template from {:path=>nil}
[--28T17::,][INFO ][logstash.outputs.elasticsearch] Attempting to install template {:manage_template=>{"template"=>"logstash-*", "version"=>, "settings"=>{"index.refresh_interval"=>"5s"}, "mappings"=>{"_default_"=>{"_all"=>{"enabled"=>true, "norms"=>false}, "dynamic_templates"=>[{"message_field"=>{"path_match"=>"message", "match_mapping_type"=>"string", "mapping"=>{"type"=>"text", "norms"=>false}}}, {"string_fields"=>{"match"=>"*", "match_mapping_type"=>"string", "mapping"=>{"type"=>"text", "norms"=>false, "fields"=>{"keyword"=>{"type"=>"keyword", "ignore_above"=>}}}}}], "properties"=>{"@timestamp"=>{"type"=>"date", "include_in_all"=>false}, "@version"=>{"type"=>"keyword", "include_in_all"=>false}, "geoip"=>{"dynamic"=>true, "properties"=>{"ip"=>{"type"=>"ip"}, "location"=>{"type"=>"geo_point"}, "latitude"=>{"type"=>"half_float"}, "longitude"=>{"type"=>"half_float"}}}}}}}}
[--28T17::,][INFO ][logstash.outputs.elasticsearch] New Elasticsearch output {:class=>"LogStash::Outputs::ElasticSearch", :hosts=>["//127.0.0.1"]}
[--28T17::,][INFO ][logstash.pipeline        ] Starting pipeline {"id"=>"main", "pipeline.workers"=>, "pipeline.batch.size"=>, "pipeline.batch.delay"=>, "pipeline.max_inflight"=>}
[--28T17::,][INFO ][logstash.pipeline        ] Pipeline main started
[--28T17::,][INFO ][logstash.agent           ] Successfully started Logstash API endpoint {:port=>}

3. elasticSearch

这个其实也是ELK中的核心，启动的时候一定要注意，因为es不可以进行root账户启动，所以你还需要开启一个elsearch账户。

groupadd elsearch                   #新建elsearch组
useradd elsearch -g elsearch -p elasticsearch  #新建一个elsearch用户
chown -R elsearch:elsearch  ./elasticsearch    #指定elasticsearch所属elsearch组

接下来我们默认启动就好了，什么也不用配置，然后在日志中大概可以看到开启了9200，9300端口。

[elsearch@slave1 bin]$ ./elasticsearch
[--28T17::,][INFO ][o.e.n.Node               ] [] initializing ...
[--28T17::,][INFO ][o.e.e.NodeEnvironment    ] [0bC8MSi] using [] data paths, mounts [[/ (rootfs)]], net usable_space [.9gb], net total_space [.6gb], spins? [unknown], types [rootfs]
[--28T17::,][INFO ][o.e.e.NodeEnvironment    ] [0bC8MSi] heap size [.9gb], compressed ordinary object pointers [true]
[--28T17::,][INFO ][o.e.n.Node               ] node name [0bC8MSi] derived from node ID [0bC8MSi_SUywaqz_Zl-MFA]; set [node.name] to override
[--28T17::,][INFO ][o.e.n.Node               ] version[5.6.], pid[], build[8bbedf5/--31T18::.105Z], OS[Linux/3.10.-.el7.x86_64/amd64], JVM[Oracle Corporation/Java HotSpot(TM) -Bit Server VM/1.8.0_144/25.144-b01]
[--28T17::,][INFO ][o.e.n.Node               ] JVM arguments [-Xms2g, -Xmx2g, -XX:+UseConcMarkSweepGC, -XX:CMSInitiatingOccupancyFraction=, -XX:+UseCMSInitiatingOccupancyOnly, -XX:+AlwaysPreTouch, -Xss1m, -Djava.awt.headless=true, -Dfile.encoding=UTF-, -Djna.nosys=true, -Djdk.io.permissionsUseCanonicalPath=true, -Dio.netty.noUnsafe=true, -Dio.netty.noKeySetOptimization=true, -Dio.netty.recycler.maxCapacityPerThread=, -Dlog4j.shutdownHookEnabled=false, -Dlog4j2.disable.jmx=true, -Dlog4j.skipJansi=true, -XX:+HeapDumpOnOutOfMemoryError, -Des.path.home=/usr/myapp/elasticsearch]
[--28T17::,][INFO ][o.e.p.PluginsService     ] [0bC8MSi] loaded module [aggs-matrix-stats]
[--28T17::,][INFO ][o.e.p.PluginsService     ] [0bC8MSi] loaded module [ingest-common]
[--28T17::,][INFO ][o.e.p.PluginsService     ] [0bC8MSi] loaded module [lang-expression]
[--28T17::,][INFO ][o.e.p.PluginsService     ] [0bC8MSi] loaded module [lang-groovy]
[--28T17::,][INFO ][o.e.p.PluginsService     ] [0bC8MSi] loaded module [lang-mustache]
[--28T17::,][INFO ][o.e.p.PluginsService     ] [0bC8MSi] loaded module [lang-painless]
[--28T17::,][INFO ][o.e.p.PluginsService     ] [0bC8MSi] loaded module [parent-join]
[--28T17::,][INFO ][o.e.p.PluginsService     ] [0bC8MSi] loaded module [percolator]
[--28T17::,][INFO ][o.e.p.PluginsService     ] [0bC8MSi] loaded module [reindex]
[--28T17::,][INFO ][o.e.p.PluginsService     ] [0bC8MSi] loaded module [transport-netty3]
[--28T17::,][INFO ][o.e.p.PluginsService     ] [0bC8MSi] loaded module [transport-netty4]
[--28T17::,][INFO ][o.e.p.PluginsService     ] [0bC8MSi] no plugins loaded
[--28T17::,][INFO ][o.e.d.DiscoveryModule    ] [0bC8MSi] using discovery type [zen]
[--28T17::,][INFO ][o.e.n.Node               ] initialized
[--28T17::,][INFO ][o.e.n.Node               ] [0bC8MSi] starting ...
[--28T17::,][INFO ][o.e.t.TransportService   ] [0bC8MSi] publish_address {192.168.23.151:}, bound_addresses {[::]:}
[--28T17::,][INFO ][o.e.b.BootstrapChecks    ] [0bC8MSi] bound or publishing to a non-loopback or non-link-local address, enforcing bootstrap checks
[--28T17::,][INFO ][o.e.c.s.ClusterService   ] [0bC8MSi] new_master {0bC8MSi}{0bC8MSi_SUywaqz_Zl-MFA}{xcbC53RVSHajdLop7sdhpA}{192.168.23.151}{192.168.23.151:}, reason: zen-disco-elected-as-master ([] nodes joined)
[--28T17::,][INFO ][o.e.h.n.Netty4HttpServerTransport] [0bC8MSi] publish_address {192.168.23.151:}, bound_addresses {[::]:}
[--28T17::,][INFO ][o.e.n.Node               ] [0bC8MSi] started
[--28T17::,][INFO ][o.e.g.GatewayService     ] [0bC8MSi] recovered [] indices into cluster_state
[--28T17::,][INFO ][o.e.c.r.a.AllocationService] [0bC8MSi] Cluster health status changed from [RED] to [YELLOW] (reason: [shards started [[.kibana][]] ...]).

4. kibana

它的配置也非常简单，你需要在kibana.yml文件中指定一下你需要读取的elasticSearch地址和可供外网访问的bind地址就可以了。

[root@slave1 config]# pwd
/usr/myapp/kibana/config

[root@slave1 config]# vim kibana.yml

elasticsearch.url: "http://localhost:9200"
server.host: 0.0.0.0

然后就是启动，从日志中可以看出，当前开了5601端口。

[root@slave1 kibana]# cd bin
[root@slave1 bin]# ls
kibana  kibana-plugin  nohup.out
[root@slave1 bin]# ./kibana
  log   [::27.650] [info][status][plugin:kibana@5.2.] Status changed from uninitialized to green - Ready
  log   [::27.748] [info][status][plugin:elasticsearch@5.2.] Status changed from uninitialized to yellow - Waiting for Elasticsearch
  log   [::27.786] [info][status][plugin:console@5.2.] Status changed from uninitialized to green - Ready
  log   [::27.794] [warning] You're running Kibana 5.2.0 with some different versions of Elasticsearch. Update Kibana or Elasticsearch to the same version to prevent compatibility issues: v5.6.4 @ 192.168.23.151:9200 (192.168.23.151)
  log   [::27.811] [info][status][plugin:elasticsearch@5.2.] Status changed from yellow to green - Kibana index ready
  log   [::28.250] [info][status][plugin:timelion@5.2.] Status changed from uninitialized to green - Ready
  log   [01:23:28.255] [info][listening] Server running at http://0.0.0.0:5601
  log   [::28.259] [info][status][ui settings] Status changed from uninitialized to green - Ready

5. 浏览器中输入：http://192.168.23.151:5601/ 你就可以打开kibana页面了，，默认让我指定一个查看的Index。

接下来我们在本机的/logs文件夹下创建一个简单的1.log文件，内容为“hello world”，然后在kibana上将logstash-*  改成 log* ，Create按钮就会自动出来。

[root@slave1 logs]# echo 'hello world' > .log

进入之后，点击Discover，你就可以找到你输入的内容啦~~~~ 是不是很帅气。。。

如果你装了head安装包，你还可以看到它确实带了日期模式的Index索引，还自带5个默认分片数。

好了，本篇就说这么多吧，希望对你有帮助。