再帮我看看这一段,老外写的东西,就是好。(50分)

  • 主题发起人 主题发起人 huangbaili
  • 开始时间 开始时间
H

huangbaili

Unregistered / Unconfirmed
GUEST, unregistred user!
Some administratorsdo
not understand the processor Utilization statistic in
MONITOR.NLM, and they believe that the higher the percentage, the worse
NetWare’s performance will be. This is entirely false. If you watch the
utilization figure in monitor for only a short while you抣l realize that this
number spikes to 100% quite often. Even numbers averaging 60's or 70's may be
OK in some environments. However, high utilization that is a problem will
usually display these conditions: (1) Performance deteriorates noticeably;
some examples are listed here, (2) Utilization will peg at 100% for periods 15
to 20 minutes or longer, sometimes causing users to be dropped by the server,
(3) Logins take an extremely long time, (4) Applications fail or discontinue to
work, (5) File copies take too long to complete, and (6) Switching screens at
the file server console takes a long time. If you are not seeing these
conditions, your utilization may be normal. Baseline your server. Know what is
normal, and know the difference between a cosmetic problem and a true
performance issue.
Server utilization is defined as the amount of work the server isdo
ing with
respect to the maximum amount it could bedo
ing. Server utilization in NetWare
is measured as an amount of processor utilization. Processor utilization is
measured as a percentage of the time the CPU is currently active. NetWare
calculates the utilization value by taking the inverse of the percentage of
time that the CPU sits idle.
An example of a high utilization situation usually happens in larger
configurations at login time. When multiple logins are being processed by the
server, the decryption of the users passwords is very CPU intensive and a
utilization spike can be readily seen. Large mathematical formulas and
operations must occur to decrypt a password to allow the users to authen
ticate
into the network. If too many login processes are running at the same time then

the processor can fall behind. Logins will eventually complete, but the server
may experience
100% utilization while processing the requests to decrypt the password.
 
翻译得不太好,凑合看看吧。
一些管理员并不了解MONITOR。NLM中的
处理器利用状况的统计数字。他们相信百分比越高,NETWARE的性能就越差。这是完全错误的。如果你在监视器前呆上一会,看看利用率的数字,你会发现它经常上窜至100%。在某些环境下,平均起来有60-70的利用率也能正常工作。但是,出现下列情况的高利用率则可能有问题:
1 性能明显降低。
2 利用率在15-20分钟,甚至更长的一段时间内居高不下,有时会使用户被踢出服务器。
3 登录花费太长的时间
4 应用程序不能正常工作
5 文件复制花费太长的时间
6 在文件服务器上切换屏幕花费太长的时间
如果你没发现上述情况,则工作正常。
知道什么是正常的,明确表面问题与真正的性能问题之间的区别是你对你的服务器了解程度的基线。
未完待续。。。。
 
服务器的利用率是它实际支出与其能力的比率,在NetWare中实际由CPU的利用率测得.CPU的
利用率又由其活动时间比率测得,在NetWare中其实求的是CPU idle的反.
一个利用率出现峰值的例子会是登陆时的复杂配置产生,当服务器(Server)
要处理各个用户的登陆密码时需要大量的数学计算,是很占用CPU的,所以当很多
用户登陆时,CPU很快吃不消,峰值很快达到,当然,过后,一切恢复了平静,只是你看
到啦Server的利用图好像很吓人,没什么大不了啦,哈哈.
 
服务器利用率被定义为这台服务器正在做的工作量,与它能做的最大工作量有关。NETWARE上的服务器利用率是用处理器利用率来量度的。处理器利用率是用CPU
处于活动状态的时间的百分比来量度的。NETWARE通过将CPU处于空闲状态的时间的百分比求反来获得利用率。
举一个例子,高利用率情况经常发生在登录时的大量配置时。当许多登录请求被服务器处理时,很容易发现利用率突然上升。必须进行大量的数学演算来为口令解码以让用户鉴别。如果在同一时间有太多的登录过程,处理器就会落后。最终登录会完成,但处理器在为口令解码时会有100%的利用率。
 
接受答案了.
 
后退
顶部