OpenSUSE 13.2 Released
MasterPatricko writes The latest version of the openSUSE distribution, 13.2, has been officially released. Key features include integrated support for filesystem snapshots, enabled by a switch to btrfs as the default file system, a new network manager (Wicked), as well as the usual version updates. This release includes seven supported desktop environments (KDE 4.14, GNOME 3.14, Xfce, LXDE, Enlightenment 19, Mate and Awesome) and even preview packages of Plasma 5.1, all presented with a unified openSUSE theme. Download LiveUSB and DVD images now from software.opensuse.org/132.
I am using OpenSUSE 13.1 right now with ext4 partitions and I am pondering migrating to OpenSUSE 13.2 with btrfs or simply updating the distro with ''zypper dup'' and keeping my ext4 fs.
If you are using btrfs, what has been your experience? Better performance? As stable as ext4?
The right to offend is far more important than the right not to be offended. (Rowan Atkinson)
And those of us who RTFS already knew that.
General Relativity: Space-time tells matter where to go; Matter tells space-time what shape to be.
The idea is to use the BTRFS for the system partition (mostly for being able to do OS snapshots), and have /home and whatever other work folders as XFS or ext4. I saw some phoronix article (can't remember where, so I cannot provide the link here), they ran some tests and concluded that BTRFS and XFS are generally slightly faster than ext4, but the difference was not significant, so you would probably not notice it unless for some special scenarios. Myself I had no issue with either of the 3 (on opensuse and xubuntu) and neither did I notice any visible performance differences
Systemd?
I used xfs for years, most of the time it was fine. The problems I had came from one assumption: xfs was written for high quality SGI hardware with UPS, not consumer PCs. It is not as hardware-failure tolerant as - for example - ext3 or ext4.
Mielipiteet omiani - Opinions personal, facts suspect.
Useradd has a pretty hostile interface, but it has "always" AFAIK had the required capability.
If you use "useradd -g joe joe", then group "joe" must already exist. Crap.
But if you use "useradd -U joe", or you just use "useradd joe" AND if USERGROUPS_ENAB=yes in /etc/login.defs, then a group will be created with the same name as the user, and that group will be set as the user's primary group.
"Of course" it's more complicated than that; a number of config files and switches are involved in the exact behavior, but if the distro sets up the default config "correctly", then the group-per-user "just works".
I have had things fail and XFS performed like a champ and all data was recovered. I also love the xfsdump/xfsrestore tools and xfs_fsr which will defragment a mounted filesystem (though XFS is excellent at not fragmenting in the first place).
If you want to do a very large filesystem then XFS is the obvious choice since EXT4 can't scale beyond about 16TB. My only complaint, and this is not with XFS itself, is that the tools like gparted do not play nice with growing XFS partitions. They balk that the partition is mounted then complain because they don't know how to deal with XFS. The thing is that XFS has its own tools for growing the filesystem.
One of my friends was actually one of the original authors of XFS at SGI and it was interesting talking to him. A lot of work went into the real-time part for handling live video streams when hardware was much slower.
This post is encrypted twice with ROT-13. Documenting or attempting to crack this encryption is illegal.