IOPS & MPS Workload for Oracle

 


使用下面SQL,在两个时间点上分别采集数据,之后按下述公式计算
可以得到数据库的IOPS和MBPS。

SELECT 'Number of Small Reads :' ||
       sum(decode(name, 'physical read total IO requests', value, 0) -
           decode(name,
                  'physical read total multi block requests',
                  value,
                  0)),
       'Number of Small Writes:' ||
       sum(decode(name, 'physical write total IO requests', value, 0) -
           decode(name,
                  'physical write total multi block requests',
                  value,
                  0)),
       'Number of Large Reads :' ||
       sum(decode(name,
                  'physical read total multi block requests',
                  value,
                  0)),
       'Number of Large Writes:' ||
       sum(decode(name,
                  'physical write total multi block requests',
                  value,
                  0)),
       'Total Bytes Read :' ||
       sum(decode(name, 'physical read total bytes', value, 0)),
       'Total Bytes Written :' ||
       sum(decode(name, 'physical write total bytes', value, 0)),sysdate
  FROM gv$sysstat


First collect (TA)              First collect (TB)
---------------------------     -----------------------------
Number of Small Reads : (A1)    Number of Small Reads  : (B1) 
Number of Small Writes: (A2)    Number of Small Writes : (B2) 
Number of Large Reads : (A3)    Number of Large Reads  : (B3) 
Number of Large Writes: (A4)    Number of Large Writes : (B4) 
Total Bytes Read    : (A5)    Total Bytes Read     : (B5) 
Total Bytes Written  : (A6)    Total Bytes Written   : (B6)

Small Read IOPS =(B1-A1)/(TB-TA) 
Small Write IOPS =(B2-A2)/(TB-TA) 
Total Small IOPS =(Small Read IOPS + Small Write IOPS)/(TB-TA)
I/O Percentage of Reads to Writes = Small Read IOPS : Small Write IOPS

Large Read IOPS =(B3-A3) /(TB-TA)
Large Write IOPS =(B4-A4)/(TB-TA)
Total Large IOPS =(Large Read IOPS + Large Write IOPS) / (TB-TA) = 179 IOPS
I/O Percentage of Reads to Writes = Large Read IOPS : Large Write IOPS

Total MBPS Read =((B5-A5) /(TB-TA))/(1024*1024)
Total MBPS Written =((B6-A6)/(TB-TA))/(1024*1024)
Total MBPS =((Total MBPS Read+Total MBPS Written)/(TB-TA))/(1024*1024)


data collect at 2010-10-27 15:35:03      data collect at 2010-10-27 15:45:03   
-------------------------------------    ------------------------------------- 
Number of Small Reads   :64164492        Number of Small Reads   :64245677     
Number of Small Writes  :76697185        Number of Small Writes  :76797034     
Number of Large Reads   :8480670         Number of Large Reads   :8481086      
Number of Large Writes  :135162266       Number of Large Writes  :135262659    
Total Bytes Read     :2262090980352   Total Bytes Read     :2263096722944
Total Bytes Written   :3110310104576   Total Bytes Written    :3112703476736 

 

 

Small Read IOPS =(64245677-64164492)/(10*60)  = 135 IOPS
Small Write IOPS =(76797034-76697185)/(10*60) = 166 IOPS
Total Small IOPS                = 301 IOPS
I/O Percentage of Reads to Writes             = 81.3%

Large Read IOPS =(8481086-8480670) /(10*60)     = 0.69 IOPS
Large Write IOPS =(135262659-135162266)/(10*60) = 167 IOPS
Total Large IOPS                 = 168  IOPS
I/O Percentage of Reads to Writes               = 1:240

Total MBPS Read =((2263096722944-2262090980352) /(10*60))/(1024*1024)   = 1.6 MBPS
Total MBPS Written =((3112703476736-3110310104576)/(10*60))/(1024*1024) = 3.8 MBPS
Total MBPS =((3576699760+18044017953)/(10*60))/1048576= 5.4 MBPS

 

来自 “ ITPUB博客 ” ,链接:http://blog.itpub.net/611609/viewspace-692643/,如需转载,请注明出处,否则将追究法律责任。

转载于:http://blog.itpub.net/611609/viewspace-692643/

【电能质量扰动】基于ML和DWT的电能质量扰动分类方法研究(Matlab实现)内容概要:本文研究了一种基于机器学习(ML)和离散小波变换(DWT)的电能质量扰动分类方法,并提供了Matlab实现方案。首先利用DWT对电能质量信号进行多尺度分解,提取信号的时频域特征,有效捕捉电压暂降、暂升、中断、谐波、闪变等常见扰动的关键信息;随后结合机器学习分类器(如SVM、BP神经网络等)对提取的特征进行训练与分类,实现对不同类型扰动的自动识别与准确区分。该方法充分发挥DWT在信号去噪与特征提取方面的优势,结合ML强大的模式识别能力,提升了分类精度与鲁棒性,具有较强的实用价值。; 适合人群:电气工程、自动化、电力系统及其自动化等相关专业的研究生、科研人员及从事电能质量监测与分析的工程技术人员;具备一定的信号处理基础和Matlab编程能力者更佳。; 使用场景及目标:①应用于智能电网中的电能质量在线监测系统,实现扰动类型的自动识别;②作为高校或科研机构在信号处理、模式识别、电力系统分析等课程的教学案例或科研实验平台;③目标是提高电能质量扰动分类的准确性与效率,为后续的电能治理与设备保护提供决策依据。; 阅读建议:建议读者结合Matlab代码深入理解DWT的实现过程与特征提取步骤,重点关注小波基选择、分解层数设定及特征向量构造对分类性能的影响,并尝试对比不同机器学习模型的分类效果,以全面掌握该方法的核心技术要点。
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值