云计算平台管理的三大利器Nagios、Ganglia和Splunk

云计算早已不是停留在概念阶段了,各大公司都购买了大量的机器,开始正式的部署和运营。而动辄上百台的性能强劲的服务器,为运营管理带来了巨大的挑战。

<ul style="margin:5px 0px 15px;padding:0px 0px 0px 20px;border:0px;outline:0px;font-size:13.63636302947998px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">如果没有方便的监控报警平台,对于管理员而言犹如噩梦,每天都将如救火队员一样,飞快地敲击键盘,用原始的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Unix</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">命令在多台机器中疲于奔命。</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">如果没有好的日志管理平台,对于开发者</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Troubleshooting</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">更是一件泪流满面的事情。</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">而如果你是运维团队的总负责人,简洁清晰的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Report</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">则非常重要。</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Stakeholder</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">们动不动就可能问起系统的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">SLA</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">、机器的利用率等诸多问题,毕竟,公司为此投入了巨大的资金和人力。</span>
    </li>
</ul>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">朋友们,当我们管理起公司寄予厚望的云计算平台时,当我们面对如此多充满挑战的实际问题时,该怎么办?</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <strong style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;"><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;color:#3366FF;">概述</span></strong>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">我们在搭建趋势云计算平台时,遇到了很多的问题和挑战。开始搭建时,第一次来了那么多性能强劲的机器,我们在感到兴奋的同时,也不免有些顾虑。大家坐在一起讨论,问题就列了满满一白板。</span>
</p>
<ul style="margin:5px 0px 15px;padding:0px 0px 0px 20px;border:0px;outline:0px;font-size:13.63636302947998px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">出了问题怎么办,有没有预警机制?</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">有没有可视化的管理界面?</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">管理平台需要自己开发吗?开发难度有多大?</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">有没有开源的管理工具?</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">那么多日志分布在各个机器上,有没有更有效的方法管理?</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">能否生成好的报表?</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">机器宕机,管理员能否收到短信通知?</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">如何做性能调优?</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">扩容升级时,能否给出依据?</span>
    </li>
</ul>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">带着这些问题,我们开始了自己的云计算平台管理和运营之旅,一路走来,收获颇丰。现在基本上形成了如图</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">1</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">所示的一整套云计算平台监控体系。</span>
</p>
<div id="attachment_11483" class="wp-caption aligncenter" style="margin:10px auto 1em;padding:4px;border:0px;outline:0px;font-size:13.63636302947998px;border-top-left-radius:3px;border-top-right-radius:3px;border-bottom-right-radius:3px;border-bottom-left-radius:3px;background-color:#FFFFFF;text-align:center;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;width:321px;">
    <img class=" wp-image-11483 " title="图1 云计算平台监控架构" src="http://www.programmer.com.cn/wp-content/uploads/2012/04/0011.jpg" alt="" width="311" height="145" style="margin:5px 0px 0px;padding:0px;border-style:none;outline:0px;font-size:13.63636302947998px;" />
    <p class="wp-caption-text" style="margin-top:0px;margin-bottom:0px;padding:6px 3px 2px;border:0px;outline:0px;font-size:10px;line-height:16px;text-indent:28px;">
        <span style="font-size:12px;">图1 云计算平台监控架构</span>
    </p>
</div>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">在这个系统中,我们综合利用了</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Nagios</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">、</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Ganglia</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">和</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Splunk</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">,搭建起云计算平台监控体系,使其具备错误报警、性能调优、问题追踪和自动生成运维报表的功能。有了这套系统,我们终于能够轻松管理</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Hadoop/HBase</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">云计算平台了。接下来将简单介绍它们的特点和功能。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;color:#3366FF;"><strong style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">Nagios:云计算平台的智能报警器</strong></span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">总不能天天盯着机器看吧,因此我们首先关心的是机器的监控与报警。最理想的境界是:如果机器出故障了,我能第一时间处理;如果机器没有问题(最好永远没有问题),我能去喝茶、钓鱼和睡大觉。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">发现机器有没有问题,对我们而言不是什么难事。写个脚本,</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Ping</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">一下</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">IP</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">,</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Telnet</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">每台机器的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Service</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">端口,如果增加了新机器就改改配置即可。但这样也太原始了吧,可视化效果差,不好维护,没有层次,不好管理,出不来报表,总不能老是用</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Excel</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">人工写报表吧。有没有更好的方法呢?</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">有,你可以用</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Nagios</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">Nagios</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">是一个可运行在</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Linux/Unix</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">平台之上的开源监视系统,可以用来监视系统运行状态和网络信息。</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Nagios</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">可以监视所指定的本地或远程主机以及服务,同时提供异常通知功能。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">Nagios</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">可以提供以下几种监控功能。</span>
</p>
<ul style="margin:5px 0px 15px;padding:0px 0px 0px 20px;border:0px;outline:0px;font-size:13.63636302947998px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">监控网络服务(</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">SMTP</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">、</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">POP3</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">、</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">HTTP</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">、</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">NNTP</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">、</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Ping</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">等)。</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">监控主机资源(处理器负荷、磁盘利用率等)。</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">简单的插件设计使得用户可以方便地扩展自己服务的检测方法。</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">并行服务检查机制。</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">具备定义网络分层结构的能力,并使用“</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">parent</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">”主机定义来表达网络主机间的关系,这种关系可被用来发现和明晰主机宕机或不可达状态。</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">当服务或主机问题产生与解决时将告警发送给联系人(通过电子邮件、短信、用户定义方式)。</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">具备定义事件处理功能,可以在主机或服务的事件发生时获取更多问题定位。</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">自动的日志回滚。</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">可以支持并实现对主机的冗余监控。</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">可选的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Web</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">界面用于查看当前的网络状态、通知和故障历史、日志文件等。</span>
    </li>
</ul>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">Nagios</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">最好用的地方就是它将这些每天管理员做的工作自动化,你只需设定好要监听的端口即可,它会默默地工作,帮忙定时地去检测服务端口的状态,一旦发现问题,会及时发出报警。报警可以是电子邮件也可以是手机,从而使得管理员第一时间就能收到系统的状况。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">Nagios</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的报表功能也很强大。管理员可以很容易地得到每天、每周和每月的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Service</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">运行状况。</span>
</p>
<div id="attachment_11486" class="wp-caption aligncenter" style="margin:10px auto 1em;padding:4px;border:0px;outline:0px;font-size:13.63636302947998px;border-top-left-radius:3px;border-top-right-radius:3px;border-bottom-right-radius:3px;border-bottom-left-radius:3px;background-color:#FFFFFF;text-align:center;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;width:405px;">
    <img class=" wp-image-11486 " title="图2 SPN 后台运行的所有Service的当前状态" src="http://www.programmer.com.cn/wp-content/uploads/2012/04/21.jpg" alt="" width="395" height="285" style="margin:5px 0px 0px;padding:0px;border-style:none;outline:0px;font-size:13.63636302947998px;" />
    <p class="wp-caption-text" style="margin-top:0px;margin-bottom:0px;padding:6px 3px 2px;border:0px;outline:0px;font-size:10px;line-height:16px;text-indent:28px;">
        <span style="font-size:12px;">图2 SPN 后台运行的所有Service的当前状态</span>
    </p>
</div>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">&nbsp;</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">如图</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">2</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">所示,红色部分清楚地标注有问题的机器,点开链接,就可以得到有问题机器的情况。虽然在</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">HBase</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">中,几台</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Region&nbsp;Server</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">宕机不会对整体服务产生大的影响,但多少会影响到系统的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Performance</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">。而且,如果某几台</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Region&nbsp;Server</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">频繁宕机,对整个系统的稳定性也会产生不好的影响。有了</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Nagios</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">,我们可以快速定位有问题的机器,及时地将一些机器移除出</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">HBase</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">系统,待调整好了再上线运行,以保证系统的稳定性。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">现在,</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Nagios</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">已经成为了很多公司必备的监控工具。只需要简单地配置,就可以实现强大的功能,将管理员从日常烦琐的工作中解放出来。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">有了</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Nagios</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">,哪怕就是管理上千台机器,也不会手忙脚乱,而是有一种统领千军、运筹帷幄的感觉。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <strong style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;"><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;color:#3366FF;">Ganglia:看到云计算平台的方方面面</span></strong>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">Nagios</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的确不错,但你是不是真的可以喝茶、钓鱼、睡大觉呢?显然还不行。有了</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Nagios</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">,你基本上可以做个优秀的救火队员,能在事发第一时间到达现场、处理事故。但如何防患于未然,真正做到运筹帷幄、游刃有余呢?</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">我们需要更加精确的数据,能够看到云计算平台的方方面面,能根据这些数据,做出性能调整、升级、扩容等的决策,从而保证</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Service</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">能够满足不断增长的业务需求。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">这时候,你需要</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Ganglia</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">Ganglia是</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">UC&nbsp;Berkeley</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">发起的一个开源实时监视项目,用于测量数以千计的节点,为云计算系统提供系统静态数据以及重要的性能度量数据。</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Ganglia</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">系统基本包含以下三大部分。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">Gmond:Gmond</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">运行在每台计算机上,它主要监控每台机器上收集和发送度量数据(如处理器速度、内存使用量等)。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">Gmetad:Gmetad</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">运行在</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Cluster</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的一台主机上,作为</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Web&nbsp;Server</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">,或者用于与</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Web&nbsp;Server</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">进行沟通。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">Ganglia&nbsp;Web前端:Web</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">前端用于显示</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Ganglia</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Metrics</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">图表。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">Hadoop</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">和</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">HBase</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">本身对于</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Ganglia</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的支持非常好。通过简单的配置,我们可以将</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Hadoop</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">和</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">HBase</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的一些关键参数以图表的形式展现在</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Ganglia</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Web&nbsp;Console</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">上。这些对于我们洞悉</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Hadoop</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">和</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">HBase</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的内部系统状态有很大的帮助。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">在</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Hadoop</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">conf</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">文件夹下面,找到</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">hadoop-metrics.properties</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">,配置好</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Ganglia</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Server</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">即可。这里要注意,</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Ganglia&nbsp;3.0</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">和</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Ganglia&nbsp;3.1</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的区别,它们使用了不同的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">class</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">。</span>
</p>
<blockquote style="margin:0px 0px 1em 2.5em;padding:10px 15px;border:1px solid #DDDDDD;outline:0px;font-size:0.9em;background-color:#F7F7F7;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;">
    <p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:12.727272033691406px;text-indent:28px;">
        <strong style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12.727272033691406px;"><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;color:#3366FF;">dfs.class=org.apache.hadoop.metrics.ganglia.GangliaContext31</span></strong>
    </p>
    <p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:12.727272033691406px;text-indent:28px;">
        <strong style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12.727272033691406px;"><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;color:#3366FF;">dfs.period=10</span></strong>
    </p>
    <p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:12.727272033691406px;text-indent:28px;">
        <strong style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12.727272033691406px;"><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;color:#3366FF;">dfs.servers={Ganglia_Server}:8649</span></strong>
    </p>
</blockquote>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">有了这些图表,</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Hadoop</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">和</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">HBase</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">就不再是一个黑盒。无论是</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Hadoop</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Namenode</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">、</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Datanode</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">,还是</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">HBase</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">MasterServer</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">、</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">RegionServer</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">任何时刻的情况,都会一目了然。由于图标的跨度可以是小时、天、月甚至是年,这样,就可以非常方便地定期生成周报、月报和年报。同时,根据图中</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Metrics</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的状况,我们可以通过调整参数、增加内存和硬盘、增加机器等的方法调整单个机器或者整个</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Service</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的性能。</span>
</p>
<div id="attachment_11487" class="wp-caption aligncenter" style="margin:10px auto 1em;padding:4px;border:0px;outline:0px;font-size:13.63636302947998px;border-top-left-radius:3px;border-top-right-radius:3px;border-bottom-right-radius:3px;border-bottom-left-radius:3px;background-color:#FFFFFF;text-align:center;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;width:344px;">
    <img class=" wp-image-11487 " title="图3 Hadoop其中一个DataNode的Metrics" src="http://www.programmer.com.cn/wp-content/uploads/2012/04/31.jpg" alt="" width="334" height="154" style="margin:5px 0px 0px;padding:0px;border-style:none;outline:0px;font-size:13.63636302947998px;" />
    <p class="wp-caption-text" style="margin-top:0px;margin-bottom:0px;padding:6px 3px 2px;border:0px;outline:0px;font-size:10px;line-height:16px;text-indent:28px;">
        <span style="font-size:12px;">图3 Hadoop其中一个DataNode的Metrics</span>
    </p>
</div>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">&nbsp;</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">Nagios&nbsp;</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">最大的问题在于不能洞悉到</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Service</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">内部的状况。像</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Hadoop</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">、</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">HBase</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">这样的分布式系统,一个节点的故障并不等于整个</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Service</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的故障,影响的只是</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Service</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的性能。所以,在测定</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Service</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">SLA</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">时,我们不能以某一台机器的故障作为</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Service</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">故障的评判标准。比如在我们的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">HBase&nbsp;SLA</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的设定上,我们定义了</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">HBase&nbsp;Service</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">完全不能工作的评判标准如下。</span>
</p>
<ul style="margin:5px 0px 15px;padding:0px 0px 0px 20px;border:0px;outline:0px;font-size:13.63636302947998px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">Master&nbsp;Server&nbsp;</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">联系不上。</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">所有</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">RegionServer&nbsp;</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">都无法联系上。</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">-ROOT-&nbsp;</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">表无法访问。</span>
    </li>
    <li style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">
        <span style="font-size:12px;">.META.&nbsp;</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">表无法访问。</span>
        <p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;">
            <br />
        </p>
        <div id="attachment_11488" class="wp-caption aligncenter" style="margin:10px auto 1em;padding:4px;border:0px;outline:0px;font-size:13.63636302947998px;border-top-left-radius:3px;border-top-right-radius:3px;border-bottom-right-radius:3px;border-bottom-left-radius:3px;text-align:center;width:344px;">
            <img class=" wp-image-11488 " title="图4 Ganglia对Hadoop/HBase使用情况的监测" src="http://www.programmer.com.cn/wp-content/uploads/2012/04/4.jpg" alt="" width="334" height="154" style="margin:5px 0px 0px;padding:0px;border-style:none;outline:0px;font-size:13.63636302947998px;" />
            <p class="wp-caption-text" style="margin-top:0px;margin-bottom:0px;padding:6px 3px 2px;border:0px;outline:0px;font-size:10px;line-height:16px;text-indent:28px;">
                <span style="font-size:12px;">图4 Ganglia对Hadoop/HBase使用情况的监测</span>
            </p>
        </div>
    </li>
</ul>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">那么,我们就可以根据这个规则定义</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">SLA</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">,通过定期调用</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">HBaseAdmin</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">相应</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">API&nbsp;</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">,将测试的结果发给</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Ganglia</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">。采用同样的方法,我们还可以自定义一些规则,监视</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">HBase&nbsp;Master</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">、</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Zookeeper</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">等的情况。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">通过这些方法,我们完全能够针对</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Hadoop/HBase</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">使用的实际情况,做出</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Service</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">级别而不是机器级别的监控系统并生成报表。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">此外,</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Ganglia</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">还可以通过</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Server</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">反馈回来的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Load</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">信息,给出各个机器的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Load</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">情况,给我们做升级和扩容提供依据。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">如图</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">5</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">所示,</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Ganglia</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">分别会用不同颜色,标注出当前时刻的机器</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Load</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">分布情况。如果</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Load</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">过重,就应该检查机器的具体使用情况。</span>
</p>
<div id="attachment_11489" class="wp-caption aligncenter" style="margin:10px auto 1em;padding:4px;border:0px;outline:0px;font-size:13.63636302947998px;border-top-left-radius:3px;border-top-right-radius:3px;border-bottom-right-radius:3px;border-bottom-left-radius:3px;background-color:#FFFFFF;text-align:center;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;width:454px;">
    <img class=" wp-image-11489 " title="图5 HBase Cluster Load Metrics" src="http://www.programmer.com.cn/wp-content/uploads/2012/04/5.jpg" alt="" width="444" height="190" style="margin:5px 0px 0px;padding:0px;border-style:none;outline:0px;font-size:13.63636302947998px;" />
    <p class="wp-caption-text" style="margin-top:0px;margin-bottom:0px;padding:6px 3px 2px;border:0px;outline:0px;font-size:10px;line-height:16px;text-indent:28px;">
        <span style="font-size:12px;">图5 HBase Cluster Load Metrics</span>
    </p>
</div>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">Ganglia</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的安装配置,可以参考:</span><a href="http://www.spnguru.com/?p=604" target="_blank" style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;color:#0088CC;text-decoration:none;" rel="noopener"><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">http://www.spnguru.com/?p=604</span></a><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;color:#3366FF;"><strong style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">Splunk:像查Google一样查日志</strong></span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">有了</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Nagios</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">和</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Ganglia</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">,算是成功了一大半。作为一名优秀的管理员,我们需要具备一定的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Troubleshooting</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">能力,对一些常见的问题能给出解决方案。那么,对日志的分析就必不可少。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">但</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Hadoop/HBase</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的日志分布在各个机器上面,而日志之间关联性强。</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Client</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">端的错误有可能是</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Region&nbsp;Server</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">引起,而</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Region&nbsp;Server</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的错误有可能是</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Zookeeper</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">导致。有没有一个统一的日志管理平台呢?</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">众里寻它千百度,蓦然回首,我们找到了</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Splunk</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">——日志界的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Google</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">很遗憾,</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Splunk</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">不是开源的,但它的免费版本提供每天</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">500MB</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">日志索引。如果数据量较小,通过定义好</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Log</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的级别,基本上也能满足需求。但对于数据量较大的公司,就有些捉襟见肘。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">Splunk</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">支持</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">AdHoc</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的日志搜索,而且可以与</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Nagios</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">配合使用。比如</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Nagios</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">报警某台</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">RegionServer</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">端口不可达,我们收到</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Notification</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">后,登录</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Splunk</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">,直接搜索</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">shutdown</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">和</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">host</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">名称,找到</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">RegionServer</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">退出的日志。点击详细信息,分析日志,就能快速定位问题。如图</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">6</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">所示。</span>
</p>
<div id="attachment_11491" class="wp-caption aligncenter" style="margin:10px auto 1em;padding:4px;border:0px;outline:0px;font-size:13.63636302947998px;border-top-left-radius:3px;border-top-right-radius:3px;border-bottom-right-radius:3px;border-bottom-left-radius:3px;background-color:#FFFFFF;text-align:center;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;width:370px;">
    <img class=" wp-image-11491 " title="图6 Splunk与Nagios配合使用进行日志搜索" src="http://www.programmer.com.cn/wp-content/uploads/2012/04/6_%E5%89%AF%E6%9C%AC.jpg" alt="" width="360" height="211" style="margin:5px 0px 0px;padding:0px;border-style:none;outline:0px;font-size:13.63636302947998px;" />
    <p class="wp-caption-text" style="margin-top:0px;margin-bottom:0px;padding:6px 3px 2px;border:0px;outline:0px;font-size:10px;line-height:16px;text-indent:28px;">
        <span style="font-size:12px;">图6 Splunk与Nagios配合使用进行日志搜索</span>
    </p>
</div>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">&nbsp;</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">对</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Hadoop</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">和</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">HBase</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">有了进一步了解后,我们可以利用</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Splunk</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">实时检测日志中的关键字,定义关键字规则,如监控“</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">shutdown</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">”、“</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">quit</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">”、“</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">ERROR</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">”、“</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Zookeeper&nbsp;Session&nbsp;Expired</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">”等,一旦出现,利用</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Splunk</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Notification</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">功能,发出邮件通知管理员,管理员通过</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Splunk</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;font-family:宋体;"><span style="font-size:12px;">定位问题,就可以在系统真正出现问题之前,对系</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;">统进行调整,防患于未然。</span></span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">具体</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;"><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Splunk</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">的设置,可以参考:</span><a href="http://www.spnguru.com/?p=122" target="_blank" style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;color:#0088CC;text-decoration:none;" rel="noopener"><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">http://www.spnguru.com/?p=122</span></a><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">。</span></span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;"><strong style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">总结</strong></span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="font-size:12px;">搭建一套云计算平台,强大的监控管理系统是必不可少的。当然,任何工具都不是万能的,在实际维护过程中,我们也发现,</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Nagios</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">和</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:'Times New Roman';">Splunk</span><span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;font-family:宋体;">经常出现误报,如果规则定义得不好,大量的警报邮件如潮水一样涌来,反而掩盖了真正的问题。可以说,在云计算平台的运维管理上,没有一劳永逸的事情,随着规模的不断增大和应用的不断多样化,需要大家不断地实践和总结。</span>
</p>
<p style="margin-top:0px;margin-bottom:15px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;text-indent:28px;font-family:'Lucida Grande', 'Lucida Sans Unicode', Arial, Verdana, sans-serif;line-height:21.81818199157715px;white-space:normal;background-color:#FFFFFF;">
    <span style="margin:0px;padding:0px;border:0px;outline:0px;font-size:12px;color:#808080;"><strong style="margin:0px;padding:0px;border:0px;outline:0px;font-size:13.63636302947998px;">作者杨俊华,趋势科技研发中心资深开发工程师,2009年至今一直从事Hadoop和HBase开发和运维工作,关注Hadoop开源社区的发展。</strong></span>
</p>

发表回复

您的电子邮箱地址不会被公开。 必填项已用*标注