I had a freshly installed opensuse 13.2 64 bit, kernel 3.16.6-2-desktop. Almost all partitions, including the ones that contained the OS and users' data, were Btrfs.

I put two hard drives of the computer under heavy load.

After several hours, the OS froze. In some way, I was able to read a message stating that a kernel bug was discovered, in a file from the Btrfs code. The error is reported in bugzilla:


That's all - I gave up on Btrfs. Nothing like that had happened to me with ext3, reiserfs and ext4.

What a shame! After years of reasoning, development, showcasing an array of useful features, big corporations' support and recommendation by opensuse, we have a complete, catastrophic failure! Did anybody actually bother to deploy and test?

It is interesting that the wiki page of Btrfs says that

"Stability status:
The filesystem disk format is no longer unstable"

Shall the users decide whether "not unstable" is the same as "stable"? Couldn't they do it themselves?

It is also interesting that there is no list of authors or contributors immediately available on the wiki. I also didn't dig deep enough to find out who has written such a powerful phrase, "not unstable". (Probably someone not undereducated in languages?)

In the discussion section of the wiki, there is a question posted in 2013: "AM I GOING TO LOSE MY FILES???" The discussion there continues: "That question is not answered here, but is what one intuitively is looking for. Could that question/answer be entered here, and made prominent, please?"

Well, I am answering now: yes, files may be lost in a routine copy operation.

I wonder whether you are using or going to use Btrfs. Have you heard of big deployments refusing Btrfs? Is Btrfs going the way of reiserfs?