Linked by Thom Holwerda on Wed 22nd Apr 2009 17:11 UTC, submitted by linuxmag
Linux "Btrfs holds the promise of giving Linux many enterprise class file system features similar to ZFS but with even more features and better performance. In fact, many Linux experts think that btrfs is one of the keys to the future of Linux. While btrfs is not quite ready to be your only file system, it is in the kernel ready for testing and is still undergoing very heavy development. In this article we will introduce the key features of btrfs find out how it compares to existing file systems."
Thread beginning with comment 360040
To read all comments associated with this story, please click here.
Benchmarking a FS ...
by pica on Thu 23rd Apr 2009 08:29 UTC
Member since:

... is more than just measuring sequential write and read speed.

There are a damn lot of aspects to consider

* How long does it take to create/delete files/directories? It could be quite annoying deleting a large directory tree. Take a gcc 4.x.x directory tree how many minutes do you wait until it is deleted?

* How much does the performance degrade if more than one actions are performed simultaniously? I assume, BTRFS is designed with server workloads in mind.

* How does the performance degrade over the time? Does the FS fragment? Does reusing a former allocated block imply a serious performance penalty? Are FS maintainance operations (e.g. defragmentation) automatically performed during idle times?

* Does the FS guarantee QoS (quality of service)? Streaming applications depend on guaranteed transfer rates.



Reply Score: 3