How To Calculate File System Size With One Inode

Inodes, speculated to exist short for "index nodes," have been around since the introduction of the first UNIX file system around the late 1970s.

They were adopted into Linux in the 90s—and for practiced reason. They're an excellent way to keep rails of how your files are stored, and the method many systems are still based on today.

What is an inode?

An inode is a file information construction that stores information about any Linux file except its proper name and data.

What are inodes used for?

Information is stored on your disk in the form of fixed-size blocks. If you save a file that exceeds a standard block, your figurer will find the next available segment on which to store the rest of your file. Over time, that can get super disruptive.

That's where inodes come in. While they don't contain whatever of the file's actual data, it stores the file's metadata, including all the storage blocks on which the file'due south data can be found.

Data contained in an inode:

File size
Device on which the file is stored
User and group IDs associated with the file
Permissions needed to access the file
Cosmos, read, and write timestamps
Location of the data (though not the filepath)

Inodes are besides independent of filenames. That means you tin re-create a unmarried file, rename it, and still accept it point to the aforementioned inode as the original.

Where inodes can get weird

Much similar your fixed-size blocks of data, inodes are also uniform. Information technology is possible that yous have a file that exceeds the size of your inode. In this case, your original inode will have to point to another inode in order to fully embrace the file's metadata. This is called an indirect pointer.

Inode indirect pointer graph Yeah... weird. Source.

On summit of that, the number of inodes yous accept on your disk is static. That means if you have enough files, you may run out of inodes before you lot run out of disk space. If you run out of inodes, you'll have to rebuild your filesystem and include a larger limit for inodes.

Handy inode commands on Linux

Use the control df-i to pull basic data near your inode usage, including the file organisation on which the inodes are stored, your full inode count, how many are in use (in count and %), and how many remain.

                              [bluematador ~]$ df -i /dev/bluem1                
                Filesystem     Inodes IUsed  IFree IUse% Mounted on                
                /dev/bluem1    654972   516  654456   1% /kicking

Use -inum to find files associated with a certain inode. Inversely, use ls-i to get the inode number of a file.

              ls -i Control: brandish inode
$ls -i /dir/bluem_article
129792 /dir/bluem_article
129792 is the inode of /dir/bluem_article.

What can go wrong with inodes

Running out of inodes can look weird, because it'southward likely you'll still have free deejay space. And then how practice you know if you've run out of inodes?

From our inode troubleshooting page:

The moment a file system runs out of inodes, all new files and folders will exist rejected. Until that bespeak, no adverse effects will manifest. One time inodes are 100% used, you'll begin to discover:

Information loss

Applications crashing

OS restarting

Processes don't restart

Periodic tasks not firing

If you notice any of the above, use df-i to confirm that you've reached inode capacity.