谁会是 Zabbix 和 Nagios 的大众cc继任者者

Nagios zabbix ganglia的相关参数分析和优缺点介绍
我的图书馆
Nagios zabbix ganglia的相关参数分析和优缺点介绍
最近开始接触Linux运维,安装了Nagios,zabbix和ganglia三种监控工具并使用了一周后总结出一些优缺点,这里贴出来以供参考Nagios是一款开源的企业级监控系统,能够实现对系统CPU、磁盘、网络等方面参数的基本系统监控,以及SMTP,POP3,HTTP,NNTP等各种基本的服务类型。另外通过安装插件和编写监控脚本,用户可以实现应用监控,并针对大量的监控主机和多个对象部署层次化监控架构。&Nagios本身不提供监控功能,所有的监控功能都要考插件实现,需要在被监控机上安装agent插件&Nagios的功能添加设备通过修改监控节点的配置文件自动发现设备不能,需要添加新插件发现接口通过修改配置文件数据库不支持SQL数据库服务端口通过插件支持TCP和UDP协议发现应用通过ssh 或插件是否支持NRPE支持SNMP协议版本支持V1,2,3&Nagios的优缺点优点缺点稳定的系统管理代码没有自动发现的功能,需要修改配置文件可以监控很多协议,邮件和短信通知只能在终端进行配置,不方便扩展,可读性比较差使用命令加载配置文件,不用打扰Nagios的运行没有简单的方法接收解释SNMP陷阱良好的文档库没有MIB编译器或浏览器出错的服务器,应用和设置会自动重启,自动日志滚动时间控制台功能弱,插件易用性差配置灵活,可以自定义shell脚本,通过分布式监控模式对性能和流量等指标的处理比较差&没有历史数据,只能实时报警,出错后难以追查故障原因&Nagios的界面比较清爽,功能主要靠插件实现,所以安装比较容易,可以根据自己的需求定制插件的安装,但在被监听端要进行配置,而且出错后没有历史记录可以查看很难找到出错原因,所以不推荐使用&&Zabbix的主要特点:优点:l&&安装与配置简单,学习成本低?- 支持多语言(包括中文) & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & &&l&&自动发现服务器与网络设备l&&分布式监视以及WEB集中管理功能l&&可以无agent监视 &l&&用户安全认证和柔软的授权方式 &&&&&&&&&&&&&l&&通过WEB界面设置或查看监视结果 &&&l&&smtp等报警通知功能l&&支持多平台的企业级分布式开源监控软件l&&安装部署简单,多种数据采集插件灵活集成l&&功能强大,可实现复杂多条件告警,l&&自带画图功能,得到的数据可以绘成图形l&&提供多种API接口,支持调用脚本l&&出现问题时可自动远程执行命令(需对agent设置执行权限)&&&&缺点:l&项目批量修改不方便l&社区虽然成熟,但是中文资料相对较少,服务支持有限;l&入门容易,能实现基础的监控,但是深层次需求需要非常熟悉Zabbix并进行大量的二次定制开发,难度较大;l&系统级别报警设置相对比较多,如果不筛选的话报警邮件会很多;并且自定义的项目报警需要自己设置,过程比较繁琐;l&缺少数据汇总功能,如无法查看一组服务器平均值,需进行二次开发;l&数据报表需要特殊二次开发定义;&Ganglia&&简介·&ganglia是UC Berkeley发起的一个开源集群监视项目,设计用于测量和监控数以千计的节点。·&主要是用来监控系统性能,如:cpu&、内存、硬盘使用率,&I/O负载、网络流量情况等。·&ganglia采用多层次的结构模式,使得&ganglia&可以实现良好的扩展。·&gmond&带来的系统负载非常少,这使得它成为在集群中各台计算机上运行的一段代码,而不会影响用户性能。&ganglia 监控套件主要包括三个主要部分:— gmond— gmetad— web接口,通常被称为ganglia-web。&&(1)&gmond :运行在每一个需要监测的节点上,收集监测统计,发送和接收在同一个组播或单播通道上的统计信息。如果作为发送者(mute=no),会收集基本指标,比如系统负载(load_one),CPU利用率。如果作为接收者(deaf=no),会聚合所有从别的主机上发来的指标,并把它们都保存在内存缓冲区中。&(2)gmetad:运行在server端,它会定期检查gmond,从那里拉取数据,并将他们的指标存储在RRD存储引擎中。也可以查询多个集群并聚合指标。&(3) ganglia-web :一般安装在有gmetad运行的机器上,以便读取RRD文件。浏览web页面时调用php脚本,从RRDTool数据库中抓取信息,动态的生成各类性能监控图表。&&用户群:适用于大型服务器集群用户。(1)优点:1.&适合监控系统性能,通过曲线很容易见到每个节点的工作状态2.&可以自定义监控项,监控展示有表格和图像两种,支持手机版3.&部署方便,通过不同的分层管理上万台机器,无需逐个添加配置,有利于后期的大规模扩张。4. Ganglia的强大在于:ganglia服务端能够通过一台客户端收集到同一个网段的所有客户端的数据,ganglia集群服务端能够通过一台服务端收集到它下属的所有客户端数据。这个体系设计表示一台服务器能够通过不同的分层能够管理上万台机器。这个功能是其他mrtg,nagios,cacti所不能比拟。5.Ganglia相比zabbix的优势在于客户端收集agent(gmond)所带来的系统开销非常低,不会影响相关服务的性能。6.&Ganglia的扩展插件非常好写,无论用何种语言(shell,php,pyton)都可以写,只要把最终结果传给gmetric就可以,这样在web上就可以看到对应的数据。(2)缺点:1.&没有内置的消息通知系统2.&没有报警机制,出现问题不能够及时报警&&
TA的最新馆藏
喜欢该文的人也喜欢中国领先的IT技术网站
51CTO旗下网站
Zabbix,Nagios和Zenoss分析比较
Zabbix,Nagios和Zenoss分析比较,Zabbix是一个基于WEB界面的提供分布式系统监视以及网络监视功能的企业级的开源解决方案。zabbix能监视各种网络参数,保证服务器系统的安全运营;本文讲述的是Zabbix的安装和基本配置。
作者:佚名来源:csdn| 13:25
Zabbix,Nagios和Zenoss分析比较
1.后端,何种实现,用了哪些技术,整体框架是怎么样的,理解的难易度,安装和配置,使用的难易度,数据存储技术
2.前端,同样.
目前为止,我主要研究了zabbix,nagios和zenoss三套系统.下面我一条条来分析.
1.比较Installation
我最先安装的zabbix系统,我在ubuntu和fedora上都安装了这套系统,zabbix系统并有没有一体化的安装包,它需要安装者去自己下载程序,然后下载好对应的依赖库,或者组件,然后来自己利用automake(这也可以说是linux上的标准了)来完成编译和安装.
相比之下Zabbix Network Monitoring这本书给的帮助更多一些.在安装过程中让我确确实实地感受到了不同的linux的distribution的不同,有些命令在小细节上在不同的平台是很不一样的.比如说chkconfig.还有书中的强调用CheckInstall来完成安装,其想法是因为以包的形式来进行安装,这样在发布和管理上就可有更加简单和合理,这一点对我挺有教育意义的,我觉的很对.但是checkinstall的开发早就停滞了,而且在和redhat的rpmbuild的配合工作时,似乎遇到了一些问题,所以我就不得不在使用make install在fedora平台上,这样会埋下uninstall的问题的伏笔.
在安装这个之前,我在国内的文章里看了一看,都是讨论这个nagios难安装难配置,还让我有点望而却步.可是我查看过了nagios的官方网站的quickstart后,我发现nagios的installation并没有什么负责的,看来国内的技术跟进实在是让人失望. 还是要跟到技术的发源地去.
在安装过程中参考这个,基本上就没有什么问题,我在Ubuntu和Fedora上都安装通过了,不过在Fedora中要小心SeLinux可能带来的问题,我不太了解Selinux,在整个过程中我就是直接用setenforce 0关掉了整个Selinux系统的.
安装做得最专业的,在目前我看过的这3个中,就要算这个zenoss了.在zenoss的官方网站提供了各个平台的一体化的安装包,只要你下载了安装包,对安装包直接实施安装指令就会自动完成整个安装和你配合的配置过程了.
2.比较Server Configuration and Service Configuration
zabbix的服务器运行的配置是基于几个以cfg结尾的text格式的文件来进行的,在初始的实验性配置时,使用源码包中的默认配置文件就可以启动服务了.zabbix的服务中用到的配置数据则是用数据库后端来存储,我使用的是mysql的数据库,这个在实际生产环境中,就要考虑到备份的问题和可能的数据库的性能问题了.
nagios的服务器运行配置,在使用的也是类似与zabbix一样的方式,不过他的配置文件就要多得多.特别是nagios的服务所用到的配置文件也是基于text格式的,其中用其特定的格式来进行定义的.这里就涉及到了对于这种配置语法和语言的学习问题, 不是很难,当然也不是那么容易.
zenoss由于是一体化的安装,所以我并没有看到太多,但我看到了它使用了内嵌的mysql服务器,所以我想它的数据应该是主要储存在了这个里面.
3.比较Front End
现在的Network Monitoring的软件,如果没有一个好的前端,来让SA或其他相关人员来查看,管理和使用的话,估计前途也不大.zabbix的web application是使用php来写的,他是构架于apache之上的,从我的观察和研究感觉,zabbix的web application不但是一个查看平台,它还是zabbix的service里的具体应用的配置前途,所以zabbix的web application会显得十分复杂, 在使用时也不是一日之功,有很多需要琢磨.
nagios也是使用php来实现其web application的,但是从我的使用观察上看,很明显,nagios的web application的功能要比zabbix的功能要少.也就是这样,使得nagios的界面就相比来收要简单明了,比较容易让初学者接受. 但是nagios的数据获得是不用数据库系统的,它的web application只是一个外部框架,真正的功能实现都是使用CGI的native linux 应用程序. 这点是跟zabbix很不同的.
zenoss没有使用apache作为其web服务器,而是自己内嵌了一个web server来进行web 服务的. zenoss的web前端,相比之下,感觉如果没有一本手册,是很难下手的,也就是说它确实挺复杂的,不过我还使用不深,还需要研究
以上就是Zabbix,Nagios和Zenoss的分析比较
【编辑推荐】
【责任编辑: TEL:(010)】
大家都在看猜你喜欢
头条头条热点头条头条
24H热文一周话题本月最赞
讲师:126560人学习过
讲师:207992人学习过
讲师:218608人学习过
精选博文论坛热帖下载排行
20多年以来,《软件工程:实践者的研究方法》一书是最受学生和行业专业人员欢迎的软件工程指南。它在全面而系统、概括而清晰地介绍软件工程...
订阅51CTO邮刊zabbix/cacti/nagios
ZABBIX使用手册
监控软件区别(php)
v&采集数据
v&保存数据
v&数据展示
v&数据分析及报警
l&Nagios(报警工具)
v&只关心正常与否的状态
v&依赖关系
l&Zabbix(数据采集&报警,综合cacti和nagios)
v&自动发现监控设备
v&支持分布式监控有代理
Zabbix监控方式:
or port check
Agent(硬件接口)
Monitoring
v&Database
Monitoring
v&Calculated
Monitoring(监控流量---&通过两次采集数据点来监控)
Command Monitoring(脚本)
Zabbix Agent可以监控的功能:
Zabbix 可以监控web的功能:
&O&响应时间
&O&下载速度
&O&响应代码
&O&响应内容中是否包含需要的内容
&O&基于用户登录、登出判断是否正常
&O&支持http和https
Zabbix的报警功能:
&O&SMS:短信
message:即时聊天工具
&O&Command
Execution:脚本
自动报警升级:
如果不得不在分布式环境中使用zabbix,尤其是我们多个机房,每个机房里面都有很多服务器,很显然如果每个客户端都跟server端通信,那很有可能被防火墙down掉。
&这个时候就需要Zabbix
在当地放一个zabbix Proxy 让他统计数据
而且基于tcp协议 &就只有一个端口,然后发送给server端。
Server:负责接收agent发送的报告信息的核心组件,所有配置、统计数据及操作数据均由其组织进行;
v&Database
Storage:专用于存储所有配置信息,以及由zabbix收集的数据;
interface:zabbix的GUI接口,通常与server运行在同一台主机上;
v&Proxy:可选组件,常用于分布式监控环境中,代理server收集部分监控端的监控数据并同意发往server端;
v&Agent:部署在监控主机上,负责收集本地数据并发送到server端或proxy端。
v&主机(host):要监控的网络设备,可由IP或DNS名称指定;
v&主机组(host
group):主机的逻辑容器,可以包涵主机和模板,但同一个组内的主机和模板不能互相连接;主机组通常再给用户或用户组指派监控权限时使用;
v&监控项(item):一个特定监控指标的相关数据,这些数据来自于被监控对象;对于item是zabbix进行数据收集的核心,没有item,将没有数据;相对某监控对象来说,每个item都由“key”进行标识;
v&触发器(trigger):一个表达式,用于评估某监控对象的某特定item内所收到的数据是否在合理范围内,即阈值;接收到的数据量大于阈值时,触发器状态从“OK”转变为“Problem”,当数据量再次回归到合理范围时,其状态将从“Problem”转换回“OK”;
v&事件(event):即发生的一个值得关注的事情,例如触发器的状体转变,新的agent或重新上线的agent的自动注册等。
v&动作(action):只对于特定时间事先定义的处理方法,通过包含操作(如发送通知)和条件(何时执行操作)
v&报警升级(escalation):发送报警或执行远程命令的自定义方案,如每隔5分钟发送一次报警,共发5次;
v&媒介(media):发送通知的手段或通道,如Email、Jabber或SMS
v&通知(notification):通过选定的媒介向用户发送的有关某事件的信息;
v&远程命令(remote
command):预定义的命令,可在被监控的主机处于某特定的条件下自动执行。
v&模板(template):用于快速定义被监控主机的预设条目集合,通常包含了item、trigger、graph、screen、application、以及low-level
discovery rule;模板可以直接连接至单个主机;
v&应用程序(application):一组item的集合
v&Web场景(web
scennario):用于检测web站点的可用性的一个或多个HTTP请求
v&前段(frontend):Zabbix的web接口
宏(macros)
其实就是变量,只能包含大写字母和下划线或者点,开头必须为大写
宏替换次序
用户自定义参数:
Zabbix仅支持在触发器上定义依赖关系
Zabbix安装配置
源码安装:
下载地址:/download.php
# tar -zxvf zabbix-2.0.0.tar.gz
创建用户:
# groupadd zabbix
# useradd -g zabbix zabbix
注意:同时安装了server和agent的节点上,建议其运行用户不要相同。
创建数据库:
server和proxy的运行都依赖于数据库,agent则不需要。
以MySQL数据库为例:
mysql& CREATE DATABASE zabbix CHARACTER SET utf8 COLLATE
mysql& GRANT ALL ON zabbix.* TO zbuser@'%' IDENTIFIED BY
# 请按需要修改用户名和密码;
shell& mysql -u -p zabbix &
database/mysql/schema.sql
# 如果仅为proxy创建数据库,只导入schema.sql即可;否则,请继续下面的步骤;
shell& mysql -u -p zabbix &
database/mysql/images.sql
shell& mysql -u -p zabbix &
database/mysql/data.sql
数据库配置:
修改数据库配置:vim
/etc/zabbix/zabbix_server.conf
启动zabbix服务 service zabbix-server start
启动http服务
&service httpd
根据提示错误修改php配置文件
post_max_size = 16M
max_execution_time = 300
max_input_time = 300
刷新页面:
定义一次完整的监控:添加主机(主机组) --&
(内置的简单图形) --& trigger (event:OK--&Problem,Recovery) --& action
(通知信息(通知信息大量使用宏)condition(自定义触发条件),operation(step))
1、添加主机:
2、添加监控项item
配置触发器Tigger:
信息:action
操作:Operations(报警升级)
1、第一次发送给管理员(或者运维组)
2、第6次发送给主管
3、第10次发送给CTO
条件:Conditions
报警种类及配置:
邮件报警配置:
yum &&install
echo "zabbix test mail" |mail -s
修改/etc/postfix/mail.cf
myhostname = bbs
myorigin = $mydomain
inet_interfaces = all
mydestination = $myhostname, localhost.$mydomain, localhost,
$mydomain,
service postfix restart
Zabbix配置:
配置action
5、创建graph:图形化展示数据(可以将多个item放在一张图中显示)
6、创建Screen(将多个graph放在一个屏幕上展示)
高级用法:
宏:可以理解为变量
格式:UserParameter=,
vim /etc/zabbix/zabbix_agentd.conf
UserParameter=memory.free,/usr/bin/free |awk '/^Mem:/{print
Service zabbix-agent restart
zabbix_get -s 192.168.19.66 -k "memory.free"
添加为监控对象:
自动发现(discovery)
添加主机:定义接口,关联模板
自动发现配置:
Web监控方案
可以监控web的界面
Web监控配置
创建一个新的application
创建web senarios
已投稿到:
以上网友发言只代表其个人观点,不代表新浪网的观点或立场。11-2911-2911-2911-2911-2911-2911-2911-2911-2911-29最新范文01-0101-0101-0101-0101-0101-0101-0101-0101-0101-0101-0101-0101-0101-0101-01开源监控系统中 Zabbix 和 Nagios 哪个更好?-土地公问答
开源监控系统中 Zabbix 和 Nagios 哪个更好?
开源监控系统中 Zabbix 和 Nagios 哪个更好?
我比较看好zabbix这款监控软件,理由如下:1.分布式监控,天生具有的功能,适合于构建分布式监控系统,具有node,proxy2种分布式模式2.自动化功能,自动发现,自动注册主机,自动添加模板,自动添加分组,是天生的自动化运维利器的首选,当然于自动化运维工具搭配,puppet+zabbix,或者saltstack+zabbix,那是如鱼得水。3.自定义监控比较方便,自定义监控项非常简单,支持变量,支持low leveldiscovery,可以参考我写的文档自动化运维之监控篇---利用zabbix自动发现功能实现批量web url监控4.触发器,也就是报警条件有多重判断机制,当然,这个需要你去研究一下,这也是zabbix的精华之处,5.支持多种监控方式,agentd,snmp,ipmi,jmx,逻辑图如下6.提供api功能,二次开发方便,你可以选用zabbix来进行二次深度开发,结合cmdb资产管理系统,业务管理系统,从而使你的自动化运维系统达到新的高度。6.提供api功能,二次开发方便,你可以选用zabbix来进行二次深度开发,结合cmdb资产管理系统,业务管理系统,从而使你的自动化运维系统达到新的高度。7.当然zabbix还有很多其他功能,这里不一一介绍了。很多人说zabbix不简单,其实是zabbix的设计理念有些超前,当你都研究到一定程度,你不得不佩服zabbix的团队是非常强悍的,这种工作机制也是相对先进的。国内的大厂,都有一套自己的监控系统,自己设计,自己开发,其功能也能和zabbix一样,更能适合于自己的需求,但一般企业用,特别是中型互联网公司,还是极力推荐zabbix。另外附上我的文档Zabbix使用手册这里面有我的经验总结,以及一些使用心得与技巧最后建议大家多看官方文档新浪微盘下载地址:最新文档版本为Zabbix使用手册百度网盘下载地址:Zabbix使用手册免费高速下载同时提供zabbix的安装二次定制的RPM包,该项目地址为:https://ix-rpm/tree/master/zabbix-广告一下, 本人新书已经出版,欢迎围观 Zabbix企业级分布式监控系统
我最近在研究zabbix,和生产环境中的nagios+cacti相比较。如@wenx所说,nagios最大的亮点是轻量灵活,且报警机制很强,如果你只是需要监控服务器/服务是否在运行,nagios足矣。但是如果牵涉到画图方面,我通过这段时间的亲身体会,感觉nagios+cacti的结合是不如zabbix的all in one方式的。nagios和cacti的结合需要NPC插件,牵涉到npc在cacti中的安装、配置、图表模板的对接,还要牵涉到多个数据库的创建、建表,ndo2db组件的调试(这东西的默认sql居然有错误,直接导入会出错:http://ml),总之非常繁琐,整个环境搭起来可能需要耗费不少时间。而zabbix把报警、画图全包办了,自己就能搞定上面的一堆东西。另外zabbix的agentd在windows server下运行良好,甚至32/64位均有,不像nagios只能用一个半死不活的插件。综上,zabbix和nagios不存在谁好谁坏的问题,只存在你的需求不同需要选择哪个更好的问题:如果只为了报警,nagios吧,足够满足需求;如果需要图表,zabbix吧,别用cacti了……
esty有一句话,叫一切监控软件最后都发展成又一个nagios。所以你不管最后用不用,最好都先了解一下nagios。然后,楼上那些以为nagios分布式不好的,画图不好的,api不好的,都只是了解基础用法而已。mod_gearman插件做分布式,pnp插件做绘图,nagios-api做restful交互,都超级好用
zabbix的图表有点忒难看了点。。。
Zabbix维护相对困难,Nagios维护相对容易,Cacti也是一样。但是Cacti的数据存储方式不是太好,过去的数据太模糊。Nagios的监控相对独立(插件机制,每次执行一条命令),而Zabbix的监控相对集中。开箱即用的角度来看,当然是Zabbix了,Agent支持功能较多,不需要一个个费心去找,而Naiogs则需要你进行拼装。告警风暴的抑制方面个人还没有机会对比。-------------------------------更新告警风暴的抑制方面--------------------------------------------Nagios默认可以使用parent的方式来进行依赖关系的设置,从而实现抑制告警,一级一级设备进行,比较简单。而Zabbix的就需要在触发器层面上进行设置,比较复杂,尤其是在发现等规则上设置,相当痛苦。
这个问题老实说没有意义,只有最合适的,没有最好的。而且Nagios和zabbix属于不同的领域,前者是开源软件,后者是商业开源软件。在商业开源领域,zabbix的竞争者是hyperic和zenoss,以及groundworks。下面列出Gartner对这个领域的分析:The Commercial Open-Source Monitoring LandscapeAs enterprises have adopted open-source monitoring tools, there is growing market demand by enterprises that were not comfortable relying on community support channels. This created a business opportunity for commercial open-source vendors to enter the market providing support options, which evolved from basic support to advanced support. Some vendors offer multiple tiers of software coupled
these offerings include enhanced tools and support. The commercial open-source vendors positioned themselves as utilizing community development resources, but have since discovered they must fund and control the road map and development. All vendors discussed in this research have strong communities where users collaborate and share plug-ins and other enhancements, while the provider controls the release of software.Some sample vendors with commercial open-source offerings include GroundWork, Merethis, Nagios, VMware, Zabbix and Zenoss. These vendors are reporting growth driven by the demands of managed service providers, enterprises cutting software costs and simplification of monitoring needs.All products, aside from VMware's offering, allow for Nagios's community-developed plug-ins to be deployed in the product. Active developers from all these vendors are developing open-source and commercial open-source offerings, creating value for customers as and the community. There is additional dedicated staff for professional services, support, sales, marketing and additional functions.GroundWork: GroundWork Monitor EnterpriseBased in California and included in the Gartner Magic Quadrant for Event Correlation and Analysis, GroundWork packages many open-source products into a tested bundled release with support offerings. The commercial offering includes features such as better visualization, integration with many ticketing systems, public cloud provider monitoring (Eucalyptus, Amazon Elastic Compute Cloud [Amazon EC2]), multitenancy for managed services providers, Java application monitoring via Java Management Extensions (JMX) and increased reporting capability across these tools. GroundWork has optimized the platform for increased scalability. The product also includes a publicly documented software development kit (SDK) consisting of SOAP/XML Web services that allows the platform to be extended with simple development efforts. The modifications done to existing open-source tools are contributed back to the open-source community, as well as being packaged into a community-supported open-source release called GroundWork Monitor Community Edition.Nagios: Nagios XIMinnesota is the headquarters for the commercial offering by the Nagios development team, which is at the core of the company's open-source product. The product adds additional dashboarding, customization, performance graphing and additional features. The commercial offering also includes support for a low yearly fee. Nagios has spawned several commercial open-source solutions, including GroundWork Monitor and Merethis Centreon.VMware: VMware vFabric HypericBased in California, VMware offers Hyperic, which competes differently than many similar options, the differentiator being the depth monitoring of Java applications, which leverage the open-source Spring Framework. VMware acquired SpringSource in 2009, which had multiple products, including Spring Framework, as well as the Hyperic product offering. The reasoning behind the acquisition was to launch VMware's cloud offering. Hyperic understands and manages the details of Spring-based applications via JMX with greater granularity than other monitoring applications. It has a large number of supported technologies from modern middleware, nonrelational databases, caching layers and testing frameworks. Additionally, Hyperic has close ties with VMware's vSphere platform, yet falls short on its support for packaged enterprise applications. The two offerings include Hyperic Open Source and VMware vFabric Hyperic. The commercial open-source vFabric edition includes support and adds enterprise-class features such as templates, increased manageability and an enhanced alert management system. Hyperic is an agent-based solution, aside from specific technologies that do not support agents (Merethis: CentreonHeadquartered in Paris, France, Merethis created a Nagios-based monitoring platform that includes open-source and commercial modules. The additional functionality includes a new interface, browser-based configuration and setup, and has advanced alerting rules and integration to many ticketing systems. There have been scalability improvements to the core engine that are released as the open-source product Centreon-Core. The scalability improvements to Nagios are released under the open-source projects Centreon Engine and Centreon Broker. These modifications are backward-compatible with Nagios open-source releases. Commercially released add-ons include additional dashboarding, map-based views and business intelligence reporting. Merethis also offers multiple support levels and professional services to implement and customize the solution for specific customer needs.Zabbix: ZabbixBased in Riga, Latvia, Zabbix sells support packages for the software and custom development of additional functionalities or integration. Zabbix is deployed on an existing OS, or as a virtual appliance. The product has a comprehensive set of bundled plug-ins in the packaged release, discovery of elements, an enhanced user interface, dashboard customization and enhanced graphing and charting features. The product is compatible with Nagios, allowing customers to leverage existing plug-ins. There is a wide array of support options, and they may be customized based on customer request.Zenoss: Service DynamicsZenoss is based Maryland and is included in the Gartner 2010 Magic Quadrant for Event Correlation and Analysis. The Zenoss Core offering is released as open-source and is community-supported, while the enterprise offering, Service Dynamics, includes scalability enhancements, performance improvements, premium support and additional features built around the publicly released Core Edition. The subscription is yearly and includes licenses and support. Zenoss is one of the only platforms to incorporate configuration management database-like device properties, which are modeled as the device is discovered. The product also includes service impact, event management and analytics. In the service impact modules, services and dependencies are configured, providing visibility into groups of components used to deliver a service or application. Zenoss is continuing to invest in richer event enrichment and metric analytics, virtualization monitoring and infrastructure as a service (IaaS) monitoring. Additional monitoring for impact analysis and analytics does carry an additional license fee, but provides value in root cause analysis. Zenoss provides integration with many ticketing systems, monitoring of cloud providers (VMware vCloud, Amazon Web Services and Google App Engine), as well as multitenancy required for implementations by managed services providers. Zenoss has a large set of community-developed ZenPacks, in addition to the company-supported packages.总结一下,在商业开源领域,除了Hyperic HQ之外,其他的都支持Nagios的插件;除了Hyperic和Zenoss、Zabbix,其他都是和Nagios 兼容,也就是说,其实整个monitoring这个领域一共也没有多少东西。按照设计理念,这些软件里面最牛的是Hyperic HQ ,按照未来的商业走向,未来发展最好的会是Zenoss 。在纯开源领域,分析如下:Some of the most common open-source packages include:Big Brother: This product was popular before Nagios was released and stabilized. The product was introduced in 1996, and still retains much of its dated user interface and limited feature set. The product is still in use at some enterprises in pockets. Big Brother is not free for commercial use, but the original author created Xymon under an open-source license, which has an active commu however, the dated interface still exists. Quest Software acquired the rights to the commercial Big Brother product in 2001, and renamed the commercial offering Big Brother Professional, which has an improved interface and additional features.MRTG, Cacti, Munin and Ganglia: These tools are graphing and polling engines, whose data is stored in a time series database (RRDtool). The core graphing and storage technology is a part of other open-source monitoring products, ranging from Centreon, Ntop, op5, OpenNMS, pfSense, SmokePing, Xymon, Zenoss and others. The polling is agentless (SNMP, WMI), but can leverage many APIs to poll additional hardware and software devices. MRTG, Cacti, Munin and Ganglia are not able to handle events, but there is a road map to correct this limitation in Cacti. Based on threshold settings, events may be dispatched to other event management systems (see "The Evolving Role of a Manager of Managers in IT Operations" ), such as Nagios. Notification management is nonexistent in the products, so events must move to an upper-level management solution. MRTG, Cacti, Munin and Ganglia have hundreds of plug-ins available on the Internet, which enable monitoring of other software and hardware devices. MRTG, Cacti and Munin are generic polling engines, while Ganglia is geared toward large grid or large cluster monitoring. MRTG has been largely replaced by Cacti, but MRTG is still in use in many enterprises.Nagios: The most popular of the open-source monitoring solutions, Nagios, has been successful in its penetration into IT operations organizations of any size. Due to its widespread usage, it has spawned many enhanced commercial and open-source software packages, such as Centreon, GroundWork Monitor, Icinga, Nagios XI, op5 and Zabbix. The product was released in 1999, and due to its age, the open-source offering has a dated user interface, scalability issues due to much of it written in interpreted languages, and poorly optimized contributed plug-ins. The software can monitor agentlessly via SNMP and WMI, or can leverage agent technology. There are nearly 2,000 downloadable plug- most written in Perl and Python. There is a lack of robust event management or notification management, and there are plug-ins developed to provide additional options for both use cases.Ntop: This product takes a different ap it uses a packet capture engine or collects flow data via NetFlow or Internet Protocol Flow Information Export (IPFIX). When operating in packet capture mode, it's typically connected to a span or mirror port on the network. The product builds profiles of the systems on the network. It does not provide any alerting capabilities, but is used for diagnostics and debugging.OpenNMS: The OpenNMS project is an open-source network-monitoring platform with capabilities around basic event management, polling network devices for configuration and performance data, and processing syslog log streams. Agentless polling and data collection is supported by using SNMP, HTTP, WMI, DNS and JDBC. The product scales well and provides good basic network monitoring capabilities, including discovery, event management and polling.Wireshark: This open-source packet capture and protocol analysis tool was previously called Ethereal, it is cross-platform and leverages the pcap capture library. It's intended for troubleshooting network behavior and understanding network-borne security issues. Data is captured from a network interface, or captured remotely and analyzed with the tool. Over 1,000 protocols can be decoded and analyzed. There is an active community, adding to the extensive feature list that includes tools for telephony analysis, such as voice over IP (VoIP), flow graphing and statistical analysis.在现在这个时间,基础组件监控领域是一个比较乱的领域,可以参见http://进行的调查:What we learnt talking to 60 companies about monitoring
没有更好,只有更适合吧。我一直在用 Nagios,Zabbix 很强大,很全面,但相对来说,Nagios 要更轻量、灵活,特别是它的插件机制,你可以用自己熟悉的语言实现几乎任何自己想要实现的监控。Nagios 在图表方面相对较弱,如果只是希望实现监控及报警,Nagios 是个不错的选择,至于图表,你可以交给 cacti,或者直接用 rrdtool 实现。
saltstack+zabbix用的这个,自动化运维还是很轻松的~
绘图可以用Graphite
Zabbix和Nagios是否好用,看运维人员熟悉程度。一般大型的企业内部,从基础设施、虚拟化、应用监控、用户体验监控都分别用不同的监控系统,所谓是“分布式监控”+“集中告警".监控解决发现故障问题,集中告警解决故障通知和人员协作问题110云告警定位一个平台集中处理所有告警,从公有云阿里云、腾讯云,私有云vsphere,各种开源监控zabbix,nagios,hyperic等,都支持告警集中接入。并免费提供微信和短信,移动APP提醒。110云告警 集成Zabbix
有兴趣试一试小米开源的 Open-Falcon· GitHub,虽然手机很烂,不过这个监控系统还行,23333。
all in one 和各个组件的争论喝vim和emacs争论一样从未停止,也从来争论不出来答案,所以选择自己合适的把
该说的楼上都说了,补充一个,部署成本。另外引申一个,一般企业如果在国内有跨网多节点监控需求,ngs会让人奔溃的。
两个都用,都有些问题的,最好的一般是自己做的,我一般是改改zabbix,修修改改才是自己的
学习了,其实两者都很不错!
zabbix是趋势,nagios机器少满足需求,机器数量太多就显得有心无力,nagios配置相对灵活,zabbix不容易上手,研究彻底之后也就没啥,cacti,没仔细用过,现在基本上就用花流量图
如果是使用第三方的命令统计数据,然后制成监控脚本。nagios和zabbix和cacti哪个比较好啊?
nagios配置非常繁琐,而且需要许多依赖库,尤其是结合rrdtool和gd的图形化显示,老是显示不出来。。。
四年前的提问,放在今天来看,zabbix是碾压nagios的
如果不加任何限制条件,这个问题确实没有意义,其争论的结果时:偏好N群的人仍然喜欢Nagios,偏向Z群的仍然喜欢Zabbix.而大多提出问题的人其实是刚入门的人群,在这种喋喋不休的争论中而显得更加迷茫。与其哪些"和稀泥“的”看你怎么用“的答案,我个人更欣赏那些提出对N或Z偏好而能提出一大堆理由的人的答案,最起码在你选择使用哪种产品的时候,先问问,你的需求和这些参考意见中提出的产品优势相吻合吗?
其它类似问题
其它人正在问的问题}

我要回帖

更多关于 穆加贝继任者宣誓就职 的文章

更多推荐

版权声明:文章内容来源于网络,版权归原作者所有,如有侵权请点击这里与我们联系,我们将及时删除。

点击添加站长微信