This post has already been read 3164 times!
File systems are integral part of any operating system and is used to store and organize data on computers. Any place that a computer stores data is by engaging some type of file system. A file system can be thought of as an index or database containing the physical location of every piece of data on a hard drive. Some example we can think of would be FAT, FAT32, NTFS, XFS, EXT, etc…
Interestingly in the world of file systems we have come across many and all are having its own interesting features. One such File System which I would like to mention is “Gluster File System”. Gluster is a distributed scale out enterprise ready file system. It’s primary feature is that it allows rapid provisioning of additional storage based on the storage consumption. Secondly, it supports automatic failover as a feature. You can configure failover automatically, so that if a server goes down, you don’t lose access to the data. Importantly, there are no manual steps required for failover after fixing the server that failed and to bring it back online. Further, you don’t have to do anything to get the data back except to wait. In the mean time, the most current copy of your data keeps getting served from the node that was still running.
If you feel that there is a need of a Metadata Server to keep all this background information, No.. It is done without a centralized metadata server. Using Gluster File System, it is very easy to provision our own storage backend Network Attached Storage or Network Attached Storages using most of the hardware available . You can add minimum amount initially and as you want more, adding takes just a few steps. You can build a clustered file system in a very short span of time. Advantage is that It runs on any standard hardware.
One more important feature is that, there is no need to configure kernels or fine tune the OS. GFS runs on most Unix File Systems, with XFS and ext4 present. Only recommendations are for more heavily utilized arrays, but these are simple to implement and you probably have some of these configured already anyway. Gluster data can be accessed from just about anywhere by using traditional NFS, SMB/CIFS for Windows clients, or native GlusterFS. If you require more information, please visit http://gluster.org |