Advanced features of EFS are that it keeps multiple inode tables in close proximity to data blocks rather than a single inode table, and it uses a bitmap to keep track of free blocks instead of a list of free blocks.
Inodes are created when an EFS filesystem is created, not when files are created. When a file is created, an inode is allocated to that file. Thus, the maximum number of files in a filesystem is limited by the number of inodes in that filesystem. By default, the number of inodes created is a function of the size of the partition or logical volume. Typically one inode is created for every 4K bytes in the partition or logical volume. You can specify the number of inodes with the -n option to the filesystem creation command, mkfs. Inodes use disk space, so there is a tradeoff between the number of inodes and the amount of disk space available for files.
The first block of an EFS filesystem is not used. Information about the filesystem is stored in the second block of the filesystem (block 1), called the superblock. This information includes:
An inode contains addresses for 12 extents, which can hold a combined 1536 blocks, or 786,432 bytes. If a file is large enough that it cannot fit in the 12 extents, each extent is then loaded with the address of up to 148 indirect extents. The indirect extents then contain the actual data that makes up the file. Because EFS uses indirect extents, you can create files up to 2 GB, assuming you have that much disk space available in your filesystem.
The last block of the filesystem is a duplicate of the filesystem superblock. This is a safety precaution that provides a backup of the critical information stored in the superblock.
EFS filesystems can become fragmented over time. Fragmented filesystems have small contiguous blocks of free space and files with poor layouts of the file extents. The fsr command reorganizes filesystems to improve file extent layout and compact the filesystem free space. By default, fsr is run once a week automatically from crontab.