Brief introduction of UNIX system performance monitoring and shell realization system statistical program of resource utilization
Abstract: This article briefly describes several command-line tools for system performance management in unix systems (mainly taking IBM's AIX as an example), and writes shell example programs, using the tools mentioned to achieve statistics of system resource utilization Procedures to provide a reference for doing unix system management and tuning.
1 Introduction
Performance optimization in Unix systems and determining performance bottlenecks in the system are one of the main tasks of system administrators. In a computer system, CPU, memory, hard disk, and network are the main factors that affect system performance. Therefore, the monitoring and adjustment of system performance is also mainly how to obtain a certain balance among these resources to meet people's expectations for system performance.
The IBM minicomputer RS6000 is widely used in enterprise applications. Its AIX operating system, like other UNIX systems, provides a very rich means for system administrators to monitor the system. Here are a few monitoring tools in the AIX system as an example, using the shell program and the scheduled operation program cron and the powerful awk tool to achieve statistical system resource utilization, you can count the daily average memory usage, daily peak CPU 1. The average daily utilization rate of CPU, and make the monthly statistics based on the average daily data in the month. The statistics of the previous month are generated on the first morning of each month, so that the system administrator can understand the utilization rate of the system ’s operating resources. Provide a good reference. The system tools Vmstat, svmon, awk, etc. mentioned in them are very powerful, so they are not limited to the length of the article. If you want to understand the usage of these commands more thoroughly, please refer to the relevant technical information or manual.
2. Performance monitoring tool in AIX
A system to provide users with continuous high availability services, not only requires that it can run normally, but also to make its performance in the best state. One of the responsibilities of the system administrator is to perform statistical analysis of the system performance, grasp the operating status and resource usage of the system, and then make the system operate healthily by rationally configuring system resources.
From the perspective of system management, performance management mainly focuses on the allocation and utilization of existing resources, including physical resources and logical resources. Physical resources include CPU, memory, and I / O devices, and logical devices include logical volume managers, virtual memory managers, system resource controllers, and file systems. Here highlights some commands and tools about system performance.
1.iostat
The iostat command mainly monitors the system input / output device load by observing the physical disk active time and their average transmission speed. According to the report generated by the iostat command, the user can determine whether a system configuration is balanced, and based on this, the input / output load can be better balanced between the physical disk and the adapter.
2.netpmon
The netpmon command can monitor system events and performance regarding network behavior and CPU consumption by network behavior. The netpmon command reports network behavior during the specified monitoring cycle.
3.ps tool
The ps command is the most common command in a UNIX system. It mainly displays statistics and status information about processes in the system, such as process ID, I / O behavior, and CPU utilization. Using the information provided by the ps command, you can determine how long a process has run, how much CPU time the process has used, and whether the process has been punished by the system. You can also use the ps command to determine how much memory the process uses, how much I / O is completed, the priority of the process, and who created the process.
4.vmstat
The vmstat command reports statistics on core threads, virtual memory, traps, disk, and CPU behavior. And each behavior report is more detailed with percentages to indicate user state, nuclear state, idle, and waiting for disk I / O.
The kernel maintains statistics on core threads, page breaks, and interrupt behavior, and the vmstat command accesses these data by using the knlist subroutine and / dev / kmen pseudo device driver. Disk input / output statistics are maintained by device drivers. For disks, the average transfer speed is determined by using the active time core to transfer the number of messages. The percentage of active time is calculated from the amount of time the drive is busy during the reporting period.
5.sar
The sar command reports CPU usage, I / O, and other system behavior. The sar command can collect, report, and save system behavior information. If no input file is specified, sar calls the sarc command to access system data.
Users can use the cron command to run two shell scripts (/ usr / lib / sa / sa1 and / usr / lib / sa2) to provide daily statistics and reports. Some sample sections are included in the crontab file / var / spool / cron / crontabs / adm to demonstrate when cron will run these shell scripts. The data collected in this way is useful for determining the time period characteristics of the system and determining the peak usage time.
However, it should be noted that the sar command will generate a considerable amount of reading and writing when it runs by itself. Therefore, it is best to run sar statistics without workload to see how much sar has an impact on the overall statistics.
6.topas
The topas command is used to monitor various system resources, such as CPU usage, CPU events and queues, memory and paging space usage, disk performance, network performance, and NFS statistics. It also reports the consumption of system resources by processes assigned to different WLM classes. It can also report the hottest processes in the system and the popular classes of Workload Manager (WLM). Information about the WLM category will only be displayed when WLM is activated. The topas command defines popular processes as those that use a lot of CPU time. The topas command has no log option, all information is real-time.
7.truss
The truss command tracks a system's system calls, signals received, and machine errors incurred. The application to be checked can be specified on the command line of the truss command, or the truss command can be hung on one or more already running processes.
8.svmon
The svmon command is used to display the current memory status. It can capture and analyze snapshot information of virtual memory. All statistics are performed in units of 4k memory pages.
3. System utilization statistics script
1. Introduction to statistical procedures
This system utilization statistics script program is composed of stat.sh and setup.sh, which has only been tested in AIX 4.3.3 and 5.2 environments, and does not yet support other UNIX servers. It is recommended to place the two scripts in the / stat directory to ensure that root has execute permissions on the two files (cd / stat; chmod 755 * .sh). You can run the setup.sh script, which will directly add an item to the root crontab. Run stat.sh at 23:55 every night. The script is output to the / tmp / stat directory by default, and result.lst is the average daily statistical value. The output format is "Date Memory Daily Average Utilization% CPU Daily Peak% CPU Daily Average Utilization%" Monthly statistics are based on The daily data is averaged, and the statistical data of the previous month is generated in the early morning of each month and output to report.lst. For instructions on crontab and awk and shell programming methods in the script, please refer to the relevant technical documents.
2.setup.sh script
Execute setup.sh on the system command line, add /stat/stat.sh to the root user's cron scheduled job item, and add the file name as / var / spool / cron / crontabs / root. Otherwise, it is reported that the scheduled job already exists. code show as below:
#! / usr / bin / ksh
FNAME = / stat / stat.sh
CPATH = / var / spool / cron / crontabs / root
[-f $ FNAME] || (echo "FILE: $ FNAME NOT FOUND!"; exit 1)
if ["` grep $ FNAME $ CPATH` "=" "]; then
echo "55 23 * * * $ FNAME> / dev / null 2> & 1" >> $ CPATH
else
echo "The entry exists in $ CPATH!"
fi
3.stat.sh script
This script uses two important system performance tools provided by AIX, vmstat and svmon, to collect the entire day's running statistics of cpu and memory, respectively. In order not to affect the operation of the system, it is set to be executed every 60 seconds in the scheduled job cron, run 24 hours a day, and save the data to the cpu. Date and mem. Date files to facilitate the operation at 23:55 Then use the awk tool for one-day statistical calculations.
As we all know, awk text processing is processed line by line, so only need to set the counter cnt variable for accumulation to facilitate the calculation of the average utilization of cpu and memory. We use the algorithm of average memory utilization to illustrate, and the generation of other algorithms is similar to it.
The algorithm for average memory utilization is:
Set the cnt, delta, tbase, and base variables to zero initially, cnt is the row count variable, delta holds the memory usage change, tbase is the memory usage at the cron time point, and base is the total memory.
Starting from the first line of the mem. Date file, if base is 0, the values ​​in the $ 2 and $ 3 fields are assigned to base and tbase. Then the next line of cnt ++ is processed, that is, the memory usage at the new time point is subtracted from the memory usage at the previous time point, and the result is assigned to the delta variable to add:
delta + = $ 3-tbase
Until the last line, perform 100 * (tbase + delta / cnt) / base to get the average memory usage rate of the day.
The code of the program is as follows:
#! / usr / bin / ksh
DDIR = / tmp / stat / data
[-d $ DDIR] || mkdir -p $ DDIR
RFILE = / tmp / stat / result.lst
PFILE = / tmp / stat / report.lst
# get old date
OFILE = `date +% y% m% d`
OMON = `date +% y% m`
# produce report line for yestoday: DATE MEMAVG CPUPEAK CPUAVG
if [-f $ {DDIR} / mem. $ {OFILE} -a -f $ {DDIR} / cpu. $ {OFILE}]; then
awk 'BEGIN {cnt = 0; delta = 0; tbase = 0; base = 0} / ^ memory / {if (base == 0) {base = $ 2; tbase = $ 3}; cnt ++; delta + = $ 3-tbase} END {if (cnt> 0) printf ("%. 1f", 100 * (tbase + delta / cnt) / base)} '$ {DDIR} / mem. $ {OFILE} | read MEMAVG
awk 'BEGIN {cnt = 0; cidmin = 100; cidsum = 0} / [0-9] $ / {if ($ 16
echo $ OFILE "" $ MEMAVG "" $ CPUPEAK "" $ CPUAVG >> $ RFILE
fi
# wait TIll tomorrow
NFILE = `date +% y% m% d`
while ["$ OFILE" = "$ NFILE"]
do
sleep 60
NFILE = `date +% y% m% d`
done
# clear the old stat data file
find / tmp / stat / data -type f -cTIme +31 | rm -f &
# new stat
vmstat 60 1430> $ {DDIR} / cpu. $ {NFILE} &
svmon -G -i 60 1430> $ {DDIR} / mem. $ {NFILE} &
# produce the month report
[-f $ PFILE] || echo "#MONTHMEMAVGCPUMAXCPUAVG"> $ PFILE
if [`expr $ NFILE-$ OFILE` -gt 1 -a -f $ RFILE]; then
grep ^ $ OMON $ RFILE | awk 'BEGIN {cnt = 0; masum = 0; cmsum = 0; casum = 0} {cnt ++; masum + = $ 2; cmsum + = $ 3; casum + = $ 4} END {if (cnt> 0) printf ("%. 1f %%%. 1f %%%. 1f %%", masum / cnt, cmsum / cnt, casum / cnt)} '| read MAVG CMAX CAVG
echo $ OMON "" $ MAVG "" $ CMAX "" $ CAVG >> $ PFILE
fi
4. Summary
In summary, in the unix system, provides a very powerful system monitoring tool. You can use these tools and combine the convenience of shell scripts to compile very convenient and practical monitoring programs, and get timely and detailed system performance statistics. Based on these data, system administrators can make use of modulation tools for system performance, such as fdpr, schedtune, vmtune, etc., to adjust logical resources of the system's kernel parameters or upgrade hardware resources to ensure safe and stable operation of application systems
Srisung Technology Co.,Ltd.,provides various brands feeder,it is not only original new but also used.It can be repaired and tested if it is original and used.In adddtion, any other spare parts for SMT machine also can be found in this company.
Fuji Feeder is in stock for SMT machine for FUJI NXT CP6 CP7 XP242 XP243
FUJI Feeder
Yamaha Smt Feeder
Yamaha Feeder
Smt Feeder For Yamaha
Siemens Smt Feeder
SIEMENS Feeder
Samsung Smt Feeder
FUJI Feeder
Pneumatic Feeder,Fuji Feeder,Fuji Smt Tape Feeder,Smt Fuji Feeder
Shenzhen Srisung Technology Co.,Limited , https://www.sr-smt.com