Changing the block size to a higher number than the default (1024) may yield significant performance in raw read speed because of a reduced number of seeks and also because the fsck session during boot may be faster, because there may be less file fragmentation, and for other reasons However, increasing the block size blindly (that is, without knowing the average file size) can result in wasted space If the average file size is 2010 bytes on a system with 4096 byte blocks, each file will waste on an average 2086 bytes (4096 2010)! So know your file size before messing with the block size

The default ext2 file system can be tuned using the e2fsprogs suite In this section, I will discuss how you can install this suite of programs and use them in tuning, performing routine checks, and repairing your disks

To tune the ext2 file system, you need to install the e2fsprogs utility package as follows: 1 Download the e2fsprogs-versionsrcrpm (replace version with the latest version number) source distribution from wwwrpmfindnet I downloaded the e2fsprogs-119-0srcrpm package You can also get the source from the e2fsprogs project site at http://e2fsprogssourceforgenet su to root 2 Run the rpm -ivh e2fsprogs-versionsrcrpm command to extract the source into a /usr/src/redhat/SOURCES/ directory The source RPM drops a e2fsprogs-versiontargz file that needs to be extracted using the tar xvzf e2fsprogs-versiontargz command This creates a subdirectory called e2fsprogs-version 3 Change to the new subdirectory 4 Run mkdir build to create a new subdirectory and then change to that directory 5 Run /configure script to configure the source tree 6 Run the make utility to create the binaries 7 Run the make check command to ensure that everything is built correctly 8 Run the make install command to install the binaries After you have installed the e2fsprogs utilities you can start using them as discussed in the following sections

You should never apply the ext2 utilities on a mounted ext2 and you should always back up your data whenever you are modifying anything belonging to a file system

The following section discusses how to use the tune2fs utility (part of the e2fsprogs package) to tune an unmounted ext2 file system called /dev/hda7 If you one or more of the settings discussed below, don t forget to change the partition name (/dev/hda7) to the appropriate name First, let s look at what tune2fs shows as the current settings for the unmounted /dev/hda7 by running the following command:

The output should be similar to this:

tune2fs 119, 13-Jul-2000 for EXT2 FS 05b, 95/08/09 Filesystem volume name: <none> Last mounted on: <not available>

Filesystem UUID: Filesystem magic number: Filesystem revision #: Filesystem features: Filesystem state: Errors behavior: Filesystem OS type: Inode count: Block count: Reserved block count: Free blocks: Free inodes: First block: Block size: Fragment size: Blocks per group: Fragments per group: Inodes per group: Inode blocks per group: Last mount time: Last write time: Mount count: Maximum mount count: Last checked: Check interval: Next check after: Reserved blocks uid: Reserved blocks gid: First inode: Inode size:

5d06c65b-dd11-4df4-9230-a10f2da783f8 0xEF53 1 (dynamic) filetype sparse_super clean Continue Linux 1684480 13470471 673523 13225778 1674469 1 1024 1024 8192 8192 1024 128 Thu Feb 15 17:51:19 2001 Thu Feb 15 17:51:51 2001 1 20 Thu Feb 15 17:50:23 2001 15552000 (6 months) Tue Aug 14 18:50:23 2001 0 (user root) 0 (group root) 11 128

The settings that I discuss in the list below are bolded above: Errors behavior: This setting dictates how the kernel behaves when errors are detected in the file system There are three possible values for this setting continue, remount-ro (readonly), panic The default setting is to continue running even if there is an error Mount count: The number of time you have mounted this file system Maximum mount count: This means that after the maximum number of read/write mode mounts the file system will be subject to a fsck checking session during the next boot cycle Last checked: This setting shows the last date on which a fsck check was performed The check interval for two consecutive fsck sessions Check interval: This is only used if the maximum read/write mount count is not reached during the interval In other words, if you don t unmount the file system for 6 months, then even though the mount count will be only 2, the Fsck check will be forced because the file system exceeded the check interval Next check after: This is the next fsck check date