1. Delete 0 byte file
find -type f -size 0 -exec rm -rf {} \;2. View process
Sort by memory size
PS -e -o "%C : %p : %z : %a"|sort -k5 -nr
3. Sorted by CPU utilization from large to small
ps -e -o "%C : %p : %z : %a"|sort -nr
4. Print the URL in the cache
grep -r -a jpg /data/cache/* | strings | grep "http:" | awk -F'http:' '{print "http:"$2;}'5. To view the number of concurrent http requests and their TCP connection status:
netstat -n | awk '/^tcp/ {++S[$NF]} END {for(a in S) print a, S[a]}'6, sed -i '/Root/s/no/yes/' /etc/ssh/sshd_config Sed in this article, the line of root matches the line of root and replaces no with yes.
7. How to kill MySQL process
ps aux |grep mysql |grep -v grep |awk '{print $2}' |xargs kill -9 (Learn from it awk Use of)killall -TERM mysqld
kill -9 `cat /usr/local/apache2/logs/httpd.pid` Try the killing process PID
8. Show services enabled at run level 3:
ls /etc/rc3.d/S* |cut -c 15- (Learn from it cut Purpose of intercepting data)
9. How to display multiple information when writing SHELL and use EOF
cat << EOF +--------------------------------------------------------------+ | === Welcome to Tunoff services === | +--------------------------------------------------------------+ EOF
10. Clever use of for (such as building soft links to MySQL)
cd /usr/local/mysql/bin for i in * do ln /usr/local/mysql/bin/$i /usr/bin/$i done
11. Get IP address
ifconfig eth0 |grep "inet addr:" |awk '{print $2}'| cut -c 6- perhaps
ifconfig | grep 'inet addr:'| grep -v '127.0.0.1' | cut -d: -f2 | awk '{ print $1}'12. Size of memory
free -m |grep "Mem" | awk '{print $2}'13. View port 80 to establish a connection
netstat -an -t | grep ":80" | grep ESTABLISHED | awk '{printf "%s %s\n",$5,$6}' | sort14. View the number of concurrent requests from Apache and its TCP connection status
netstat -n | awk '/^tcp/ {++S[$NF]} END {for(a in S) print a, S[a]}'15. Because my colleague wanted to count the size of all jpg files under the server, he wrote a SHELL for him to count. It was originally implemented with xargs, but he handled part at a time. There are multiple summations... The following commands can be solved.
find / -name *.jpg -exec wc -c {} \;|awk '{print $1}'|awk '{a+=$1}END{print a}'The more CPUs (multiple CPUs, cat /proc/cpuinfo |grep -c processor), the lower the system load and the more requests can be processed per second.
16. CPU load
cat /proc/loadavg
Check whether the first three output values exceed 4 times of the system logical CPU.
17, CPU load
mpstat 1 1
Check if% idle is too low (e.g. less than 5%).
18. Memory space
free
Check whether the free value is too low. You can also use # cat /proc/meminfo
19. SWAP space
free
Check whether the swap used value is too high. If the swap used value is too high, further check whether the swap action is frequent:
vmstat 1 5
Observe whether the si and so values are large
20. Disk space
df -h
Check whether the partition utilization (Use%) is too high (e.g. more than 90%). If you find that the space of a partition is nearly exhausted, you can enter the mount point of the partition and Use the following command to find the file or directory that occupies the most space:
du -cks * | sort -rn | head -n 10
21. Disk I/O load
iostat -x 1 2
Check if I/O utilization (% util) exceeds 100%
22. Network load
sar -n DEV
Check whether the network traffic (rxbyt/s, txbyt/s) is too high
23. Network error
netstat -i
Check whether there is a network error (drop fifo colls carrier), or use the command: # cat /proc/net/dev
24. Number of network connections
netstat -an | grep -E "^(tcp)" | cut -c 68- | sort | uniq -c | sort -n
25. Total number of processes
ps aux | wc -l
Check whether the number of processes is normal (for example, more than 250)
26. Number of runnable processes
vmwtat 1 5
Column shows the number of runnable processes. Check whether they exceed 4 times the logical CPU of the system
27. Process
top -id 1
Observe whether there are abnormal processes.
28. Users
who | wc -l
Check whether there are too many login users (e.g. more than 50) You can also use the command: # uptime.
29. System log
# cat /var/log/rflogview/*errors
Check whether there are abnormal error records You can also search for some exception keywords, such as:
grep -i error /var/log/messages grep -i fail /var/log/messages
30. Core log
dmesg
Check whether there are abnormal error records.
31. System time
date
Check whether the system time is correct.
32. Number of open files
lsof | wc -l
Check whether the total number of open files is too high.
33. Log
# logwatch –print
Configure / etc/log.d/logwatch.conf, set Mailto as your email address, and start the mail service (sendmail or postfix), so that you can receive log reports every day.
The default logwatch only reports yesterday's logs. You can use # logwatch – print – range all to obtain all log analysis results.
You can use # logwatch – print – detail high to get more specific log analysis results (not just error logs).
34. Kill the processes related to port 80
lsof -i :80|grep -v "ID"|awk '{print "kill -9",$2}'|sh35. Clear the dead process
ps -eal | awk '{ if ($2 == "Z") {print $4}}' | kill -936. tcpdump packet capture is used to prevent port 80 from being attacked by people. It can analyze data
tcpdump -c 10000 -i eth0 -n dst port 80 > /root/pkts
37. Then check the number of IP duplicates and sort them from small to large. Note "- t\ +0" There are two spaces in the middle
# less pkts | awk {'printf $3"\n"'} | cut -d. -f 1-4 | sort | uniq -c | awk {'printf $1" "$2"\n"'} | sort -n -t\ +038. Check how many PHP CGI processes are active
netstat -anp | grep php-cgi | grep ^tcp | wc -l
39. View the service started by the system
chkconfig --list | awk '{if ($5=="3:on") print $1}'40. kudzu check the network card model
kudzu --probe --class=networ