Linux large files have been deleted, but df does not reduce the used space.

Linux large files have been deleted, but df's viewing space used has not been reduced to solve the Baidu problem: In Linux, df found that/partition used more than 80%, and then deleted most of the files, use df to check the usage and the displayed content remains unchanged. why? I checked it online. it seems that it is the inode reason, and it hasn't been deleted yet ,... linux large files have been deleted, but df's viewing space used has not been reduced to solve the Baidu problem: In Linux, df found that/partition used more than 80%, and then deleted most of the files, use df to check the usage and the displayed content remains unchanged. why? I checked it online and it seems that it was the inode reason, but it hasn't been deleted yet, but I know why, and I didn't say a solution. I want to know how to solve it, show df the actual usage of the current disk! Solution: www.2cto.com lsof | grep deleted process no. fd mfsmount 13124 root txt REG 160760 1368696/var/tmp/portage/sys-cluster/moosefs-1.6.26/image/usr/bin/mfsmount (deleted) syslog-ng 14142 root 7 w REG 192380776055 7104 21550/var/log/messages (deleted) python2.6 root 0u CHR 136,0 0t0 2/dev/pts/0 (deleted) php-fpm 30315 root 3u REG 659057, 3 0/tmp /. xcache.0.0.1718328763.lock (deleted) php-fpm 30 315 root 4u REG 1236916, 3 0/tmp /. xcache.0.1.20.168404.lock (deleted) php-fpm 30315 root 5u REG 104,3 0 1368672/tmp /. xcache.0.2.300103481.lock (deleted) php-fpm 30316 nobody 3u REG 104,3 0 659057/tmp /. xcache.0.0.1718328763.lock (deleted) php-fpm 30316 nobody 4u REG 104,3 0 1236916/tmp /. xcache.0.1.20.168404.lock (deleted) php-fpm 30316 nobody 5u REG 1368672, 3 0/tmp /. xcache.0.2.300103481.loc K (deleted) php-fpm 30318 nobody 3u REG 104,3 0 659057/tmp /. xcache.0.0.1718328763.lock (deleted) php-fpm 30318 nobody 4u REG 104,3 0 1236916/tmp /. xcache.0.1.20.168404.lock (deleted) php-fpm 30318 nobody 5u REG 1368672, 3 0/tmp /. xcache.0.2.300103481.lock (deleted) php-fpm 30319 nobody 3u REG 104,3 0 659057/tmp /. xcache.0.0.1718328763.lock (deleted) here we can see that syslog-ng occupies GB of disk space, and we have put/var/log/mes Sages is deleted, but the process is not finished. I first used kill to achieve this effect, and finally remembered that it was best to restart the syslog-ng service. #/Etc/init. d/syslog-ng restart if you restart the service and find that the df Usage value has not been reduced, kill it! # Kill-9 14142