ELK——安装 logstash 2.2.0、elasticsearch 2.2.0 和 Kibana 3.0

时间:2024-01-12 15:22:32

本文内容

  • Elasticsearch
  • logstash
  • Kibana
  • 参考资料

迁移到:http://www.bdata-cap.com/newsinfo/1712695.html

本文介绍安装 logstash 2.2.0 和 elasticsearch 2.2.0,操作系统环境版本是 CentOS/Linux 2.6.32-504.23.4.el6.x86_64。

安装 JDK 是必须的,一般操作系统都会有,只是版本的问题,后面会提到。

而 Kibana 只是一个用纯 JavaScript 写的前端 UI。一定要注意 Kibana 的版本,它会要求 ES 的版本。比如 Kibana 3 要求 Elasticsearch 至少 0.9.9 或更高;Kibana 4.5 要求 ES 至少 2.3.0 或更高。

假设,ELK 都位于 10.1.8.166 机器上。

Elasticsearch

Elasticsearch主页

Elasticsearch(简称,ES)提供 ZIP、TAR、DEB 和 RPM 包。但 Github 上提供了一个针对中文环境的 Elasticsearch-RTF,RTF 即 Ready To Fly,它是一个继承了基本插件(如服务封装、中文分词、mapper-attachments、transport-thrift、tools.carrot2 等)的并带有示例程序的可直接上手的简易工程版本,换句话说,帮你入门的。本文针对 Elasticsearch-RTF 为例。基本上,elasticsearch 解压后就能使用。

ELK——安装 logstash 2.2.0、elasticsearch 2.2.0 和 Kibana 3.0

假设你已经从 Github 上下载 elasticsearch-rtf,名为 elasticsearch-master.zip,并上传到你的 Linux 服务器 /usr/local/elasticsearch目录(如果没有,就用 mkdir 命令创建一个)。

  • 现在,解压,并重新命名文件夹:
[root@vcyber local]# cd /usr/local/elasticsearch

 

[root@vcyber local]# unzip elasticsearch-master.zip

 

[root@vcyber elasticsearch]# ls

 

elasticsearch-master  elasticsearch-master.zip

 

[root@vcyber local]# mv elasticsearch-master elasticsearch

 

[root@vcyber elasticsearch]# ls

 

elasticsearch  elasticsearch-master.zip

  • 尝试运行 elasticsearch:

Linux 环境:

[root@vcyber elasticsearch]# pwd

 

/usr/local/elasticsearch/elasticsearch

 

[root@vcyber elasticsearch]# bin/elasticsearch

Windows 环境,执行相应的 .bat 文件,即 elasticsearch.bat。

但报错了:

[root@vcyber elasticsearch]# bin/elasticsearch

 

Exception in thread "main" java.lang.RuntimeException: Java version: Oracle Cooration 1.7.0_51 [Java HotSpot(TM) 64-Bit Server VM 24.51-b03] suffers from crical bug https://bugs.openjdk.java.net/browse/JDK-8024830 which can cause dataorruption.

 

Please upgrade the JVM, see http://www.elastic.co/guide/en/elasticsearch/referce/current/_installation.html for current recommendations.

 

If you absolutely cannot upgrade, please add -XX:-UseSuperWord to the JAVA_OPT environment variable.

 

Upgrading is preferred, this workaround will result in degraded performance.

 

        at org.elasticsearch.bootstrap.JVMCheck.check(JVMCheck.java:123)

 

        at org.elasticsearch.bootstrap.Bootstrap.init(Bootstrap.java:283)

 

        at org.elasticsearch.bootstrap.Elasticsearch.main(Elasticsearch.java:3

 

Refer to the log for complete error details.

大意是:Java 运行时异常,本机版本 JDK 有 bug……让升级 JVM。如果实在不能升级,就向 JAVA_OPT 环境变量添加 -XX:-UseSuperWord 选项。

于是,看一下本机的Java 版本:

[root@vcyber elasticsearch]# java -version

 

java version "1.7.0_51"

 

Java(TM) SE Runtime Environment (build 1.7.0_51-b13)

 

Java HotSpot(TM) 64-Bit Server VM (build 24.51-b03, mixed mode)

 

[root@vcyber elasticsearch]# echo $JAVA_HOME

 

/usr/java/jdk1.7.0_51

 

[root@vcyber elasticsearch]#

 

版本是 1.7.0_51。再在官网查了一下,说:“Elasticsearch requires at least Java 7. Specifically as of this writing, it is recommended that you use the Oracle JDK version 1.8.0_72. Java installation varies from platform to platform so we won’t go into those details here. Oracle’s recommended installation documentation can be found on Oracle’s website. Suffice to say, before you install Elasticsearch, please check your Java version first by running (and then install/upgrade accordingly if needed):”,大意是,ES 至少要求 7,推荐使用 1.8.0_72。

  • 那就删除之前的版本,按个新的吧。先删掉之前的 JDK,然后再用 yum 按个新的:
[root@vcyber elasticsearch]# yum list installed | grep java

 

[root@vcyber elasticsearch]# yum list installed | grep jdk

 

jdk.x86_64                           2000:1.7.0_51-fcs                installed

 

[root@vcyber elasticsearch]# yum -y remove jdk.x86_64

 

……

 

[root@vcyber elasticsearch]#yum -y install java-1.8.0-openjdk*

 

……

注意:java-1.8.0-openjdk*”,后面有个星号,即安装 java 全部相关的东西~

  • 安装完成后,设置 JDK 的环境变量:
[root@vcyber elasticsearch]# export JAVA_HOME=/usr/lib/jvm/java-1.8.0

 

[root@vcyber elasticsearch]# export PATH=$JAVA_HOME/bin:$PATH

 

[root@vcyber elasticsearch]# export CLASSPATH=.:$JAVA_HOME/lib/dt.jar:$JAVA_HOME/lib/tools.jar

 

[root@vcyber elasticsearch]# java -version

 

openjdk version "1.8.0_71"

 

OpenJDK Runtime Environment (build 1.8.0_71-b15)

 

OpenJDK 64-Bit Server VM (build 25.71-b15, mixed mode)

 

[root@vcyber elasticsearch]# echo $JAVA_HOME

 

/usr/lib/jvm/java-1.8.0

 

[root@vcyber elasticsearch]#

另外,JDK 安装在了我机器的 /usr/lib/jvm 目录下,自己确认一下你的路径。你可以把环境变量的设置放到 profile 里,一劳永逸。

  • 再次运行:
[root@vcyber elasticsearch]# bin/elasticsearch

 

Exception in thread "main" java.lang.RuntimeException: don't run elasticsearch as root.

 

        at org.elasticsearch.bootstrap.Bootstrap.initializeNatives(Bootstrap.java:93)

 

        at org.elasticsearch.bootstrap.Bootstrap.setup(Bootstrap.java:144)

 

        at org.elasticsearch.bootstrap.Bootstrap.init(Bootstrap.java:285)

 

        at org.elasticsearch.bootstrap.Elasticsearch.main(Elasticsearch.java:35)

 

Refer to the log for complete error details.

 

[root@vcyber elasticsearch]#

  • 又报错,elasticsearch 不能用 root 用户运行,那就建立一个:
[root@vcyber elasticsearch]# groupadd es

 

[root@vcyber elasticsearch]# useradd -g es es

 

[root@vcyber elasticsearch]# passwd es

 

Changing password for user es.

 

New password:

 

BAD PASSWORD: it is WAY too short

 

BAD PASSWORD: is too simple

 

Retype new password:

 

passwd: all authentication tokens updated successfully.

 

[root@vcyber elasticsearch]#

 

[root@vcyber elasticsearch]# chown -R root .

 

[root@vcyber elasticsearch]# chown -R es .

 

[root@vcyber elasticsearch]# chgrp -R es .

 

[root@vcyber elasticsearch]# ls -l

 

total 4

 

drwxr-xr-x 7 es es 4096 Mar  1 03:07 elasticsearch

 

[root@vcyber elasticsearch]#

ES 2.* 版本已经不允许用 root 用户运行了,但是 1.* 版本无所谓。

  • 重新打开一个终端,用 es 用户登录,并运行 elasticsearch:
[root@vcyber ~]$ cd /usr/local/elasticsearch/elasticsearch

 

[root@vcyber elasticsearch]$ bin/elasticsearch

 

[2016-03-01 05:11:48,413][WARN ][bootstrap                ] unable to install syscall filter: seccomp unavailable: CONFIG_SECCOMP not compiled into kernel, CONFIG_SECCOMP and CONFIG_SECCOMP_FILTER are needed

 

[2016-03-01 05:11:48,750][INFO ][node                     ] [Googam] version[2.1.1], pid[15042], build[40e2c53/2015-12-15T13:05:55Z]

 

[2016-03-01 05:11:48,750][INFO ][node                     ] [Googam] initializing ...

 

[2016-03-01 05:11:49,088][INFO ][plugins                  ] [Googam] loaded [elasticsearch-analysis-ik, elasticsearch-analysis-mmseg, elasticsearch-analysis-stconvert, elasticsearch-analysis-pinyin], sites []

 

[2016-03-01 05:11:49,121][INFO ][env                      ] [Googam] using [1] data paths, mounts [[/ (/dev/mapper/vg_vcyber-lv_root)]], net usable_space [26.1gb], net total_space [34.8gb], spins? [possibly], types [ext4]

 

[2016-03-01 05:11:51,119][INFO ][mmseg-analyzer           ] [Dict Loading] chars loaded time=42ms, line=12638, on file=chars.dic

 

……

  • 此时,在另一个终端,访问 elasticsearch:
[root@vcyber elasticsearch]# curl -X GET http://localhost:9200

 

{

 

  "name" : "Captain Savage",

 

  "cluster_name" : "elasticsearch",

 

  "version" : {

 

    "number" : "2.1.1",

 

    "build_hash" : "40e2c53a6b6c2972b3d13846e450e66f4375bd71",

 

    "build_timestamp" : "2015-12-15T13:05:55Z",

 

    "build_snapshot" : false,

 

    "lucene_version" : "5.3.1"

 

  },

 

  "tagline" : "You Know, for Search"

 

}

 

[root@vcyber elasticsearch]#

 

ES 已经安装成功。

  • 配置 ES

但此时,ES 不能通过IP访问,最好修改 config/elasticsearch.yml,配置一下 ES。

首先,找到“network.host”行,添加一行:

network.host: your id address

就能通过IP,或浏览器访问。

再找到“http.port”行,添加一行:

http.port: 9200

否则,ES 每次启动时,端口可能会变(端口被占用,ES 自己会改端口)~

如果还不能正常启动 ES,并提示端口被占用,就查看一下什么程序占用 9200 端口,kill 掉,重启 ES 就行。

我遇到的,Java 把 9200 端口占用了。

  • 安装 Head 插件

Head 是一个用来监控 ES 状态的客户端插件,可以为初学用户提供很多便利,例如,使用 Head 提供的 HTTP 客户端,通过 HTTP 方式来操作 ES。

先查看你的 ES 都有哪些插件:

[root@vcyber ~]# cd /usr/local/elasticsearch/elasticsearch

[root@cyber elasticsearch]# bin/plugin list

Installed plugins in /usr/local/elasticsearch/elasticsearch/plugins:

    - elasticsearch-analysis-mmseg-1.7.0

    - elasticsearch-analysis-stconvert-1.6.1

    - elasticsearch-analysis-pinyin-1.5.2

    - elasticsearch-analysis-ik-1.7.0

[root@vcyber elasticsearch]#

大部分是关于中文分词的,没有 Head 插件。

ES 支持在线和本地安装 Head。本地安装时,从 Github 上下载 Head 插件,然后上传到你的 ES 服务器,比如,Elasticsearch/plugins 目录。

下面是在线安装:

[root@vcyber ~]# cd /usr/local/elasticsearch/elasticsearch/

[root@vcyber elasticsearch]# bin/plugin install mobz/elasticsearch-head

-> Installing mobz/elasticsearch-head...

Trying https://github.com/mobz/elasticsearch-head/archive/master.zip ...

Downloading ...................................................................................................................................................................................................................................................................................................................................................................................................................................DONE

Verifying https://github.com/mobz/elasticsearch-head/archive/master.zip checksums if available ...

NOTE: Unable to verify checksum for downloaded plugin (unable to find .sha1 or .md5 file to verify)

Installed head into /usr/local/elasticsearch/elasticsearch/plugins/head

如果命令使用的是“mobz/elasticsearch-head”,那么 ES 将自己联网从 Github 下载再安装。

但是报错了,说校验和有问题。加上“-v”选项,这次换本地安装,而且是 zip 压缩包(从 Github 上下载的),即“file:plugins/elasticsearch-head-master.zip”,再执行一下:

[root@vcyber elasticsearch]# bin/plugin install -v file:plugins/elasticsearch-head-master.zip

-> Installing from file:plugins/elasticsearch-head-master.zip...

Trying file:plugins/elasticsearch-head-master.zip ...

Downloading .........DONE

Verifying file:plugins/elasticsearch-head-master.zip checksums if available ...

NOTE: Unable to verify checksum for downloaded plugin (unable to find .sha1 or .md5 file to verify)

- Plugin information:

Name: head

Description: head - A web front end for an elastic search cluster

Site: true

Version: master

JVM: false

 

Installed head into /usr/local/elasticsearch/elasticsearch/plugins/head

注意:install 选项,ES 的 2.* 版本,都不带“-”横线选项,即“-install”。

注意:若是本地安装,而且,你把 Head 压缩包放到了 Elasticsearch/plugins 目录下,安装后,一定要将 Elasticsearch-head zip 压缩包删掉,否则启动 ES 时会报“不能初始化插件”错误。

[2016-03-02 07:06:16,547][WARN ][bootstrap                ] unable to install syscall filter: seccomp unavailable: CONFIG_SECCOMP not compiled into kernel, CONFIG_SECCOMP and CONFIG_SECCOMP_FILTER are needed

[2016-03-02 07:06:16,866][INFO ][node                     ] [MODAM] version[2.1.1], pid[19446], build[40e2c53/2015-12-15T13:05:55Z]

[2016-03-02 07:06:16,866][INFO ][node                     ] [MODAM] initializing...

Exception in thread "main" java.lang.IllegalStateException: Unable to initialize plugins

Likely root cause: java.nio.file.FileSystemException: /usr/local/elasticsearch/elasticsearch/plugins/elasticsearch-head-master.zip/plugin-descriptor.properties: Not a directory

之后用浏览器访问 http://your ip adress:9200/_plugin/head,你就会看到如下界面:

ELK——安装 logstash 2.2.0、elasticsearch 2.2.0 和 Kibana 3.0

myfirestidx 是我之前通过命令行自己建的。

Elasticsearch Head 就成功安装了。

  • 安装 Marvel 图形化监控插件

[root@vcyber elasticsearch]# bin/plugin install -v elasticsearch/marvel/latest

-> Installing elasticsearch/marvel/latest...

Trying http://download.elasticsearch.org/elasticsearch/marvel/marvel-latest.zip...

Downloading ................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................DONE

Installed elasticsearch/marvel/latest into /usr/local/elasticsearch/elasticsearch-1.7.1/plugins/marvel

[root@vcyber elasticsearch-1.7.1]#

ELK——安装 logstash 2.2.0、elasticsearch 2.2.0 和 Kibana 3.0

我去啊~此插件收费~

  • 安装 kopf 网络插件
[root@vcyber elasticsearch]# bin/plugin install -v lmenezes/elasticsearch-kopf

-> Installing lmenezes/elasticsearch-kopf...

Trying https://github.com/lmenezes/elasticsearch-kopf/archive/master.zip...

Downloading ..................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................DONE

Installed lmenezes/elasticsearch-kopf into /usr/local/elasticsearch/elasticsearch-1.7.1/plugins/kopf

[root@vcyber elasticsearch]#

ELK——安装 logstash 2.2.0、elasticsearch 2.2.0 和 Kibana 3.0

虽然安装成功了,但此版本 kopf 不支持我的 ES 版本。

ES 插件地址


logstash

logstash主页

logstash 提供各种安装包,包括 tar.gz,ZIP,DEB 和 RPM。另外,又提供了一个包含所有插件的压缩包——logstash-all-plugins-2.2.0.tar.gz 。本文以它为例。解压后,配置 logstash,运行即可。

ELK——安装 logstash 2.2.0、elasticsearch 2.2.0 和 Kibana 3.0

ELK——安装 logstash 2.2.0、elasticsearch 2.2.0 和 Kibana 3.0

  • 在 /usr/local 下创建目录 logstash:
[root@vcyber local]# cd /usr/local

 

[root@vcyber local]# pwd

 

/usr/local

 

[root@vcyber local]# mkdir logstash

 

[root@vcyber local]#

  • 上传 logstash 到该目录。然后,解压并配置 logstash 配置文件:
[root@vcyber local]# cd logstash

 

[root@vcyber logstash]# ls

 

logstash-all-plugins-2.2.0.tar.gz

 

[root@vcyber logstash]# tar zxf logstash-all-plugins-2.2.0.tar.gz

 

logstash-2.2.0  logstash-all-plugins-2.2.0.tar.gz

 

[root@vcyber logstash]# cd logstash-2.2.0

 

[root@vcyber logstash-2.2.0]# vi logstash.conf

 

[root@vcyber logstash-2.2.0]#

文件内容如下:

input{

 

    stdin{}

 

}

 

output{

 

    stdout{}

 

}

这个配置是最简单,输入是终端命令行,输出也是终端命令行。

  • 运行 logstash:
[root@vcyber logstash-2.2.0]# bin/logstash agent -f logstash.conf

 

hello world

 

hello world 2

 

Settings: Default pipeline workers: 2

 

Logstash startup completed

 

2016-02-28T22:37:06.130Z vcyber hello world

 

2016-02-28T22:37:06.132Z vcyber hello world 2

 

logst 已经成功安装。其中,agent 表示运行Agent模式,-f 表示指定配置文件,-p 表示端口,命令行参数可以参考logstash command-lines flags

  • 另外,你可以查看 logstash 都安装了哪些插件。
[root@vcyber logstash-2.2.0]# bin/plugin list

 

logstash-codec-avro

 

logstash-codec-cef

 

logstash-codec-cloudfront

 

logstash-codec-cloudtrail

 

logstash-codec-collectd

 

logstash-codec-compress_spooler

 

logstash-codec-dots

 

logstash-codec-edn

 

logstash-codec-edn_lines

 

logstash-codec-es_bulk

 

logstash-codec-fluent

 

logstash-codec-graphite

 

logstash-codec-gzip_lines

 

logstash-codec-json

 

logstash-codec-json_lines

 

logstash-codec-line

 

logstash-codec-msgpack

 

logstash-codec-multiline

 

logstash-codec-netflow

 

logstash-codec-nmap

 

logstash-codec-oldlogstashjson

 

logstash-codec-plain

 

logstash-codec-rubydebug

 

logstash-codec-s3plain

 

logstash-codec-spool

 

logstash-filter-aggregate

 

logstash-filter-alter

 

logstash-filter-anonymize

 

logstash-filter-checksum

 

logstash-filter-cidr

 

logstash-filter-cipher

 

logstash-filter-clone

 

logstash-filter-collate

 

logstash-filter-csv

 

logstash-filter-date

 

logstash-filter-de_dot

 

logstash-filter-dns

 

logstash-filter-drop

 

logstash-filter-elapsed

 

logstash-filter-elasticsearch

 

logstash-filter-environment

 

logstash-filter-extractnumbers

 

logstash-filter-fingerprint

 

logstash-filter-geoip

 

logstash-filter-grok

 

logstash-filter-i18n

 

logstash-filter-json

 

logstash-filter-json_encode

 

logstash-filter-kv

 

logstash-filter-metaevent

 

logstash-filter-metricize

 

logstash-filter-metrics

 

logstash-filter-multiline

 

logstash-filter-mutate

 

logstash-filter-oui

 

logstash-filter-prune

 

logstash-filter-punct

 

logstash-filter-range

 

logstash-filter-ruby

 

logstash-filter-sleep

 

logstash-filter-split

 

logstash-filter-syslog_pri

 

logstash-filter-throttle

 

logstash-filter-tld

 

logstash-filter-translate

 

logstash-filter-unique

 

logstash-filter-urldecode

 

logstash-filter-useragent

 

logstash-filter-uuid

 

logstash-filter-xml

 

logstash-filter-zeromq

 

logstash-input-beats

 

logstash-input-cloudwatch

 

logstash-input-couchdb_changes

 

logstash-input-elasticsearch

 

logstash-input-eventlog

 

logstash-input-exec

 

logstash-input-file

 

logstash-input-fluentd

 

logstash-input-ganglia

 

logstash-input-gelf

 

logstash-input-gemfire

 

logstash-input-generator

 

logstash-input-github

 

logstash-input-graphite

 

logstash-input-heartbeat

 

logstash-input-http

 

logstash-input-http_poller

 

logstash-input-imap

 

logstash-input-irc

 

logstash-input-jdbc

 

logstash-input-jmx

 

logstash-input-kafka

 

logstash-input-log4j

 

logstash-input-lumberjack

 

logstash-input-meetup

 

logstash-input-pipe

 

logstash-input-puppet_facter

 

logstash-input-rabbitmq

 

logstash-input-redis

 

logstash-input-relp

 

logstash-input-rss

 

logstash-input-s3

 

logstash-input-salesforce

 

logstash-input-snmptrap

 

logstash-input-sqlite

 

logstash-input-sqs

 

logstash-input-stdin

 

logstash-input-stomp

 

logstash-input-syslog

 

logstash-input-tcp

 

logstash-input-twitter

 

logstash-input-udp

 

logstash-input-unix

 

logstash-input-varnishlog

 

logstash-input-websocket

 

logstash-input-wmi

 

logstash-input-xmpp

 

logstash-input-zenoss

 

logstash-input-zeromq

 

logstash-output-boundary

 

logstash-output-circonus

 

logstash-output-cloudwatch

 

logstash-output-csv

 

logstash-output-datadog

 

logstash-output-datadog_metrics

 

logstash-output-elasticsearch

 

logstash-output-elasticsearch-ec2

 

logstash-output-elasticsearch_http

 

logstash-output-elasticsearch_java

 

logstash-output-email

 

logstash-output-exec

 

logstash-output-file

 

logstash-output-ganglia

 

logstash-output-gelf

 

logstash-output-google_bigquery

 

logstash-output-google_cloud_storage

 

logstash-output-graphite

 

logstash-output-graphtastic

 

logstash-output-hipchat

 

logstash-output-http

 

logstash-output-influxdb

 

logstash-output-irc

 

logstash-output-juggernaut

 

logstash-output-kafka

 

logstash-output-librato

 

logstash-output-loggly

 

logstash-output-lumberjack

 

logstash-output-metriccatcher

 

logstash-output-mongodb

 

logstash-output-nagios

 

logstash-output-nagios_nsca

 

logstash-output-null

 

logstash-output-opentsdb

 

logstash-output-pagerduty

 

logstash-output-pipe

 

logstash-output-rabbitmq

 

logstash-output-redis

 

logstash-output-redmine

 

logstash-output-riemann

 

logstash-output-s3

 

logstash-output-sns

 

logstash-output-solr_http

 

logstash-output-sqs

 

logstash-output-statsd

 

logstash-output-stdout

 

logstash-output-stomp

 

logstash-output-syslog

 

logstash-output-tcp

 

logstash-output-udp

 

logstash-output-websocket

 

logstash-output-xmpp

 

logstash-output-zabbix

 

logstash-output-zeromq

 

logstash-patterns-core

 

[root@vcyber logstash-2.2.0]#


Kibana

Kibana 主页

chenryn 增强版的 Kibana

你得搞一个 Kibana 3.0 并上传你服务器的 /usr/local/Kibana 目录下,没有建立一个,然后用 unzip 解压(github 下载的是 zip 文件)。

[root@vcyber /]# cd /usr/local/kibana/

[root@vcyber kibana]# ls

exampledata  kibana-3.0  kibana-4.0  kibana-5.0

[root@vcyber kibana]#

你可以看到,我有三个版本,4.0 和 5.0 没有安装成功(总提示 node.js 报错)~Kibana 这三个主要版本差距很大~

简单介绍一下,你就知道,Kibana 的每个版本差距有多大。Logstash 早期曾经自带了一个特别简单的 Logstash-Web,用来查看 ES 数据。后来, Rashid Khan 用 PHP 写了一个更好的 Web,取名 Kibana,发布于 2011年12月11日。2012年8月19日,Rashid Khan 用 Ruby 重写了一遍,叫 Kibana 2,。因为,Logstash 也是用 Ruby 写的,所以就代替了 Logstash-Web。2014年2月,此人,用 AngularJS 又重写了一遍,(我靠,真是服了他了~),叫 Kibana 3。2014年4月,Kibana 3 停止开发,全面致力于 Kibana 4。到 2015年初,发布了用 JRuby 做后端的 beta 版,但3月正式推出时却使用 node.js 做后端。(我靠靠,服了~都被他用遍了~)

  • 到其 src 目录下:
[root@vcyber kibana]# cd kibana-3.0/src

[root@vcyber src]# ls

app  config.js  css  favicon.ico  font  img  index.html  vendor

[root@vcyber src]#

  • 修改其 config.js 文件,添加对 ES 的访问。无论是看文件内容,还是看后缀名,你能猜到,就是手动去改 JavaScript 文件而已:
elasticsearch: http://10.1.8.166:9200,

我 ES 服务器地址是 10.1.8.166,注意,屁股后面的逗号,绝对是有的。如果该文件配置错了,当你访问 Kibana 时,界面什么都看不到~

接下来,把 Kibana 搞成一个 Web 站点。tomcat、Python、Nginx 都行,我暂时用 Python。只要把 Kibana 的 src 目录发布出去就行。

  • 注意你当前的位置,执行如下命令,就能把 Kibana 做成一个 Web 站点:
[root@vcyber src]# pwd

/usr/local/kibana/kibana-3.0/src

[root@vcyber src]# python –m SimpleHTTPServer 8000

其中,SimpleHTTPServer 是 Python 模块名,区分大小写的;8000 是 Web 站点的端口。

这样,你就能通过 http://10.1.8.166:8000 访问 Kibana。

  • 但是,页面报错了:“Connection Failed”:

ELK——安装 logstash 2.2.0、elasticsearch 2.2.0 和 Kibana 3.0

提示,Kibana 不能连接 ES。至于原因,从 ES 1.4 版本开始,它增强了安全性。写得很清楚,照着做就行(事后才知道)。~但问题是,这些具体东西,在官网没找到(我是没找到),而且连 Kibana 3.0 的文档,都已经没有了~错误信息所说的:“http.cors” 相关配置,在 Elasticsearch.yml 中,也没给出~事实上,还真是写这两个配置。

  • 进入到 ES 的 config 目录:
[root@vcyber src]# cd /usr/local/elasticsearch/

[root@vcyber elasticsearch]# ls

elasticsearch-1.7.1  elasticsearch-2.2.0

[root@vcyber elasticsearch]# cd elasticsearch-1.7.1/config/

[root@vcyber config]# 

我有两个 ES 版本,Kibana 访问的是 1.7 那个版本。

  • 修改 Elasticsearch.yml 文件,添加如下内容:
http.cors.allow-origin: http://10.1.8.166:8000

http.cors.enabled: true

其中,http.cors.allow-origin 一行表明,谁可以访问 ES。若让所有域名都可以访问,写成 “*”~

本例写具体的 IP 地址和端口。

这样,Kibana 就可以访问 ES 了。如下图所示:

ELK——安装 logstash 2.2.0、elasticsearch 2.2.0 和 Kibana 3.0


参考资料

在 CentOS/Linux 把 Kibana 3.0 部署在 Nginx 1.9.12

为调试 Logstash Grok 表达式,安装 GrokDebuger 环境