I never really got this setup working quite right. Use the information here however you like, but don't bother asking me questions about it. I can't answer them.
I'm the resident Linux guru at my job -- a mid-sized local company with a decent sized IT department. We like to install servers in clusters to improve our fault tolerance. Being the Linux guy in a shop where Windows servers outnumber Unix server about 8:1, I wanted to do a one up on Windows' active-passive (high availability, or HA) cluster setup by doing a 2-node active-active (load balanced, or LB) cluster using the Linux Virtual Server (LVS) system. Our Linux distribution of choice is RedHat Enterprise Linux 4 (RHEL 4), and CentOS is the most compatible free clone thereof. Version 4 of these distros uses the Linux 2.6 kernel.
本日志由 flyinweb 于 2009-11-19 11:26:57 发表到 Linux 中,目前已经被浏览 4105 次,评论 0 次;
作者添加了以下标签: Ultra Monkey,LVS;
首页只显示了部分日志内容,要查看日志的全部内容请阅读全文;
而且直接配置文件是效率最高的,通过其它驱动效率都相对较低,BDB
这个测试不太准确,看官方的测试结果:http://bind-dlz.sourceforg
为什么使用BDB时QPS这么低? 我在bind版本基本相似的环境中测试的
It is quite useful and interesting too.
VIRT 的上限是64G,也就是36位, cat /proc/cpuinfo的结果是:addre
昨天要准备用线程重写webbench,试验了下Fedora Linux 2.6.35.14
不明白您的具体的意思是什么?
已经发送到你QQ邮箱