 | |||||||||||
| |||||||||||
 |
|||||||||||
| |
| ||||||||||||||||
| ShaoLin HA Cluster 1.0 Installation Manual and Operations Guide | ||
|---|---|---|
| Prev | Chapter 9. Tuning the clusters | Next |
Using journaling filesystems can greatly improve data consistency during a failure. It ensures file system meta data is in a consistent state. This can eliminate the need to go through lengthly fsck checks during mount time. There are several types of journaling filesystems in Linux.
It is compatible with ext2 filesystem. It is developed by Redhat and aimed for backward compatibility with ext2 with added journaling features. You can convert existing ext2 filesystem to ext3, please see more details on Section 9.1.1.
Reiserfs is based on fast balanced trees. It is a filesystems that is targeted for high performance on both small files and large files. Since Linux 2.4, Reiserfs is now come with the standard kernel. See Reiserfs homepage for more information.
IBM's journaled file system technology, currently used in IBM enterprise servers, is designed for high-throughput server environments, key to running intranet and other high-performance e-business file servers. JFS now comes with the stock Linux kernel since 2.4 . See more information on JFS for Linux homepage.
A filesystem originated from the SGI IRIX operating system. SGI has open-sourced the XFS and ported to Linux. This is known as the most powerful and scalable filesystem. However, not all Linux distribution include support for XFS. For more information, please see SGI XFS for Linux homepage.
If you have your Linux system already up an running. You want to install ShaoLin HA Cluster and migrate your existing system to a cluster enabled system. You don't want to format your filesystem as it already contains large amount of data. Since most of the old Linux box is using ext2 filesystem. It is possible to convert your existing ext2 file system to the new journaling ext3 file system without too much effort. A utility named tune2fs which comes with most of the up-to-date Linux distributions, allow you to convert existing unmounted ext2 volumes to ext3. Since ext3 is now a standard filesystem built-in to the Linux 2.4 series kernel. It is safe to use this filesystem as a primary filesystem for your clusters which provides both backward and future compatibility, also the performance and benefits or journaling filesystems.
First of all, you have to unmount your filesystem. Then run the command tune2fs -j [device] to convert your old ext2 volume to a ext3 volume. For more details, please read the manual pages of tune2fs. After you have converted the filesystem to ext3, remember to modify the /etc/fstab to change the mount entry from ext2 to ext3.
