Announcement

Collapse
No announcement yet.

Bcachefs File-System Plans To Try Again To Land In Linux 6.6

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #11
    What I do not get so far is what I, as "Joe Average", can expect from it. I am using BTRFS since some years and I am quite happy with it. I enjoy the features like checksumming, snapshots etc. BTRFS was a huge step forward to me, coming from ext4. So which aspect of Bcachefs is supposed to make me migrate?

    Comment


    • #12
      This is something else here: https://evilpiepirate.org/git/bcache...fs/bkey.c#n727

      Not sure it should get merged

      Comment


      • #13
        Originally posted by Joe2021 View Post
        What I do not get so far is what I, as "Joe Average", can expect from it. I am using BTRFS since some years and I am quite happy with it. I enjoy the features like checksumming, snapshots etc. BTRFS was a huge step forward to me, coming from ext4. So which aspect of Bcachefs is supposed to make me migrate?
        as a btrfs user you get: tiered storage and erasure coding (raid5/6 not yet stable though).
        you can set parity level on a subvol basis: eg a very huge folder is raid5 but another smaller folder is mirrored.

        afaik all features of btrfs are already available in bcachefs too

        i am really looking forward how good it performs with vm's. my btrfs experience with vm's was horrible (and yes i know i can disable cow. but i want to use cow features).

        Comment


        • #14
          Originally posted by flower View Post

          as a btrfs user you get: tiered storage and erasure coding (raid5/6 not yet stable though).
          you can set parity level on a subvol basis: eg a very huge folder is raid5 but another smaller folder is mirrored.

          afaik all features of btrfs are already available in bcachefs too

          i am really looking forward how good it performs with vm's. my btrfs experience with vm's was horrible (and yes i know i can disable cow. but i want to use cow features).
          I really don't understand... you criticize the alleged instability of Btrfs, and then you are looking forward to using an experimental fs, which as of today hasn't even been accepted into the kernel tree.
          Strange, isn't it?​

          Comment


          • #15
            Originally posted by woddy View Post

            I really don't understand... you criticize the alleged instability of Btrfs, and then you are looking forward to using an experimental fs, which as of today hasn't even been accepted into the kernel tree.
            Strange, isn't it?​
            well i used btrfs when everybody said it is stable and it wasn't. i lost data multiple times.
            bcachefs is marked experimental. i know that it can loose data and i can prepare for that.

            my biggest problem was vm performance though. since i switched to zfs i don't have that problem anymore.

            btw: yes in the past i critized btrfs instablility. but not in your quote. in that quote i only listed features of bcachefs. i can't comment on btrfs stability anymore because the last time i used it was about five years ago. i really just dont care about btfs anymore.

            Comment


            • #16
              Originally posted by flower View Post
              as a btrfs user you get: tiered storage and erasure coding (raid5/6 not yet stable though).
              you can set parity level on a subvol basis: eg a very huge folder is raid5 but another smaller folder is mirrored.
              I’m using BTRFS on top of dm-crypt on top of bcache on top of dm-raid. I guess the only feature I’m missing is per-subvol/file settings, and combining everything into one layer.

              Comment


              • #17
                Originally posted by flower View Post
                as a btrfs user you get: tiered storage
                I've never heard of that. There are open issues for it though. Am I missing something?

                Comment


                • #18
                  Originally posted by Mark Rose View Post

                  I've never heard of that. There are open issues for it though. Am I missing something?
                  tiered storage just means you have disks with different performance characteristics in the same array. eg m2, sata ssd and hdds.

                  for every subvol you can define where that data should be stored.

                  with zfs i currently have two pools:
                  two m2 mirrored for vm's
                  eight hdd raid6 for data (and two mirrored sata ssd's in a special vdev for metadata) .

                  with bcachefs i can merge those two pools into one and just tell bcachefs where to store data.

                  Comment


                  • #19
                    Originally posted by flower View Post

                    tiered storage just means you have disks with different performance characteristics in the same array. eg m2, sata ssd and hdds.

                    for every subvol you can define where that data should be stored.

                    with zfs i currently have two pools:
                    two m2 mirrored for vm's
                    eight hdd raid6 for data (and two mirrored sata ssd's in a special vdev for metadata) .

                    with bcachefs i can merge those two pools into one and just tell bcachefs where to store data.
                    Sure, that's for ZFS and bcachefs, but I've not heard of btrfs having tiered storage. I've actually been thinking it would be a fun project to get my feet wet with kernel development (no commitment yet).

                    Comment


                    • #20
                      Originally posted by EphemeralEft View Post

                      I’m using BTRFS on top of dm-crypt on top of bcache on top of dm-raid. I guess the only feature I’m missing is per-subvol/file settings, and combining everything into one layer.
                      For curiosity, could you describe better you stack ? Because btrfs over dm-raid/md-raid loses the capability to rebuild bad data from the good copy...

                      Comment

                      Working...
                      X