Wed, 30 Mar 2011 20:20:08 GMT
... can be found here.
I am currently using ext4+lvm, and it is ok so far. It works, is fast and can do everything I need.
I was using zfs-fuse for a long time, but it had major problems with file-locks and stuff (for example, man-db regularly freezed my system), which are, as far as I read, resolved meanwhile, but currently I do not have the time to set up a new system. However, I never had any data loss caused by zfs-fuse, no matter how often it crashed, after reboot everything worked again.
No comparism to btrfs which broke my partition and gave me data loss, after suspending and resuming (if it does not support suspending and resuming, then it should not allow the kernel to shut down - and zfs-fuse even crashed and did not damage anything, even though it is not even a kernel-module).
When seeing the list in that link, I wonder why there is still so much development done with btrfs: ZFS is only about one year older than Btrfs (according to Wikipedia), is stable, there are implementations for many OSes, there is a FUSE-Implementation, and its only problem is the license. ZFS and Btrfs belong to Oracle - why do they not just change the license of ZFS such that it can be added to the Linux-Kernel easily, and use the power of the Btrfs-Developers to enhance ZFS?
I am currently using ext4+lvm, and it is ok so far. It works, is fast and can do everything I need.
I was using zfs-fuse for a long time, but it had major problems with file-locks and stuff (for example, man-db regularly freezed my system), which are, as far as I read, resolved meanwhile, but currently I do not have the time to set up a new system. However, I never had any data loss caused by zfs-fuse, no matter how often it crashed, after reboot everything worked again.
No comparism to btrfs which broke my partition and gave me data loss, after suspending and resuming (if it does not support suspending and resuming, then it should not allow the kernel to shut down - and zfs-fuse even crashed and did not damage anything, even though it is not even a kernel-module).
When seeing the list in that link, I wonder why there is still so much development done with btrfs: ZFS is only about one year older than Btrfs (according to Wikipedia), is stable, there are implementations for many OSes, there is a FUSE-Implementation, and its only problem is the license. ZFS and Btrfs belong to Oracle - why do they not just change the license of ZFS such that it can be added to the Linux-Kernel easily, and use the power of the Btrfs-Developers to enhance ZFS?
Show comments (Requires JavaScript, loads external content and cookies from Disqus.com)