Post AOGmqxtd6D7H7saCqO by RL_Dane@fosstodon.org
 (DIR) More posts by RL_Dane@fosstodon.org
 (DIR) Post #AOGmdP9Uxv6PbYXVJo by dfloyd888@fosstodon.org
       2022-10-05T21:23:48Z
       
       0 likes, 0 repeats
       
       Latest small project:  Changing a Raspberry Pi's / filesystem from ext4 to btrfs, for that bit-rot detecting goodness.
       
 (DIR) Post #AOGmdPgoy1l1Gu87G4 by joel@fosstodon.org
       2022-10-05T21:29:39Z
       
       0 likes, 0 repeats
       
       @dfloyd888 I don't understand btrfs yet, what do you mean with detecting bitrot? Why would ext4 not do it? I am quite a noob in this regard
       
 (DIR) Post #AOGmqxtd6D7H7saCqO by RL_Dane@fosstodon.org
       2022-10-05T21:32:09Z
       
       0 likes, 0 repeats
       
       @joel Not sure about bitrot *detection*, but btrfs has a lot of advanced features like deduplication (copy-on-write transparent hardlinks) and snapshots that makes it desirable for serious/enterprise users.I've found there is a performance hit to some of the fancy stuff, so watch out for that. Also, some proprietary software like Dropbox just don't know how to deal with it (that's the only one I know of, there may be others)@dfloyd888
       
 (DIR) Post #AOGqFGWPnj5nhl0HQW by dfloyd888@fosstodon.org
       2022-10-05T22:11:24Z
       
       0 likes, 0 repeats
       
       @joel Active bitrot detection is something that only a relative few filesystems have.  It takes some engineering because every block of every file has to have a checksum, and that checksum made part of the file checksum.Another trait of these filesystems that do bitrot checking, like ZFS and btrfs is the fact that they are always copy-on-write.  When a file is modified, the old data stays put, and the new data is thrown on new sectors/pages.  This is used instead of a journal (like ext4 uses.)