Linux Inode runs out of no space

A case of an application of the inode is that when the file is created, there is not enough space to see through DF to find that actually used is only 20%, and here is why Du and DF see the same disk when the size is inconsistent (you can try it definitely inconsistent)

First, the inode exhausted the poor is certainly the inode table is not free, how to solve???

Find./-name "*.log*" |xargs-l rm can be deleted if too many small files may error arglist too long

Find./-name "*.log*"-exec rm-rf {} \; can also be deleted, all are said to delete the file one by one as parameters passed

And if you want to not format the hard disk (MKFS), you can delete a portion and then reassign the connection to another file system.

Ln-s/newcache/opt/newcache This access is the I node that uses the new file system

If you are involved in the early consideration of the size of the hard disk storage data, you can consider the following two points: if there are many small files, when creating a hard disk can be adjusted block smaller. If the file is particularly large, such as video, you can adjust the block larger.

How to increase the number of Inode only umount the file system and then Mkfsmkfs ext3/dev/sda1 [-N number-of-inodes] directly specify the required inode number or mkfs ext3/dev/sda1 [-I How to calculate the number of Bytes-per-inode]inode: typically hard disk size/block (i.e., a sector)

Linux Inode runs out of no space