A very frequent issue where all DBAs always comes across is filling up of server mount point. Once mount point utilization exhausts or becomes 100% full then your database or application will be hung.
In this post, we will learn different ways to filter out which files to clean-up.
1. Unix command to check the mount point status
[piyush@oraclecloud /p033 ]$
df -hP /p003
Filesystem Size Used
Avail Use% Mounted on
/p003
919G 902G 17G 99% /p003
2. Unix
command to check the top used size files under /p033 mount point
[piyush@oraclecloud /p033]$ ls
-ltr |head
total 40484
drwx------. 2
root root 4096 Nov 21 2016 test
-rwxrwxrwx. 1
root root 20569422 Nov 28 2016 finstall.zip
-rw-------. 1
root root 10145806 Nov 28 2016 installrpm.sh
-rw-------. 1
root root 10547066 Nov 28 2016 installdeb.sh
-rw-r--r--. 1
root root 0 Dec 7 2016 testfiles
3. Unix command to sort top 10 used files with size
du -hsx * | sort -rh | head -10
[piyush@oraclecloud /p033 ]$
du -hsx * | sort -rh | head -10
20G home
6.2G tmp
5.4G opt
5.1G usr
4.9G var
170M Agents
155M roots
21M etc
20M
install.sh
11M
installdeb.sh
4. Find
files that are older than 30 days and delete them.
Ex- delete files from /p003 mount
point
find /p003 -mtime +30
-exec rm {} \;
5.
Search numbers of files that are older than 30 days
find /p003 -ctime +30
| wc -l
Either you can manually delete the top consumed files or move them to another
mount point. Or you can use the find command to search and deleted files that
are older than 30 days.
Hope this helps. Happy Learning.
2 comments:
Adding a -type f param is a good practice
Also tee the logs to a file for reference and audits
Thats true. You can also add -type parameter.
Post a Comment