Announcement

Collapse
No announcement yet.

An Exciting Btrfs Update With Encoded I/O, Fsync Performance Improvements

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

  • #11
    Originally posted by intelfx View Post

    How exactly is ZFS "next-gen" compared to btrfs, or was it just generic fanboy commentary?
    With BTRFS, to do everything ZFS does you have to use BTRFS, LUKS, and LVM. Granted that OpenZFS also needs to be running on top of a LUKS container for full security. OpenZFS with encryption still shows dataset names. To hide that from an a Live environment, hacker who stole your HDD, etc you have to put OpenZFS on top of LUKS. Since that applies to pretty much everything I don't consider that to be a fault for ZFS.

    On top of the full integration, it's cross platform with FreeBSD, Solaris, and Linux with MacOS and Windows ports in the works. One has to dip into the free Windows file systems to get that feature.

    For me, those are what makes ZFS next-gen. Everything necessary is integrated within and it has the most potential to just work everywhere.

    My biggest problem is I tend to swap "zpool" and "zfs" when running commands...the biggest offenders being "zpool status $POOL" and "zfs get all". Those are the correct commands, but sometimes I'm high and run "zfs status $POOL" or "zpool get all". I wish their tools were a bit smarter and would redirect commands like those while notifying me I input the incorrect command.

    Seriously, that's my biggest issue. I swap the zpool and zfs commands around.

    As far as being next-gen, just read my story above. You'd be hard pressed to do all that disk juggling with any other file system or setup and still have your data nice and safe 6 years later on completely different disks and systems. What I didn't mention in that story is that I switched PCs 4 times during that and that same ZFS volume went with me the entire time (4 if you include buying a new motherboard and straight swapping all the rest of the hardware, 3 if not).

    Comment


    • #12
      Originally posted by skeevy420 View Post
      With BTRFS, to do everything ZFS does you have to use BTRFS, LUKS, and LVM. Granted that OpenZFS also needs to be running on top of a LUKS container for full security. OpenZFS with encryption still shows dataset names. To hide that from an a Live environment, hacker who stole your HDD, etc you have to put OpenZFS on top of LUKS. Since that applies to pretty much everything I don't consider that to be a fault for ZFS.

      On top of the full integration, it's cross platform with FreeBSD, Solaris, and Linux with MacOS and Windows ports in the works. One has to dip into the free Windows file systems to get that feature.

      For me, those are what makes ZFS next-gen. Everything necessary is integrated within and it has the most potential to just work everywhere.

      My biggest problem is I tend to swap "zpool" and "zfs" when running commands...the biggest offenders being "zpool status $POOL" and "zfs get all". Those are the correct commands, but sometimes I'm high and run "zfs status $POOL" or "zpool get all". I wish their tools were a bit smarter and redirect commands like those while notifying me I input the incorrect command.

      Seriously, that's my biggest issue. I swap the zpool and zfs commands around.
      Cool story. Except you don't need LVM with btrfs (why would you even??), and you say it yourself that you still need LUKS with ZFS. And that's completely irrelevant to what makes a technology "next-gen" or not.

      Originally posted by skeevy420 View Post
      As far as being next-gen, just read my story above. You'd be hard pressed to do all that disk juggling with any other file system or setup and still have your data nice and safe 6 years later on completely different disks and systems. What I didn't mention in that story is that I switched PCs 4 times during that and that same ZFS volume went with me the entire time (4 if you include buying a new motherboard and straight swapping all the rest of the hardware, 3 if not).
      I get it that you want to ascribe every last bit of your experience to perceived ZFS' greatness, but how is this different from literally any other non-brain-dead filesystem out there?
      Last edited by intelfx; 22 March 2022, 10:42 AM.

      Comment


      • #13
        Originally posted by kdemello1980 View Post

        OpenZFS isn't in the kernel tree, nor is it license-compatible with Linux.
        The same can be said of the Nvidia kernel module. That doesn't mean that Nvidia GPU's suddenly suck or are unusable.

        Heck, there's a bunch of modules in the AUR and software in repositories called non-free for that very reason -- not license-compatible with Linux.

        Frankly, license-compatible always turns into a philosophical debate because that's all it really is. Let's not go there today.

        Comment


        • #14
          Originally posted by skeevy420 View Post

          With BTRFS, to do everything ZFS does you have to use BTRFS, LUKS, and LVM.
          would you mind sharing an example of something you can do with ZFS that is not feasible with btrfs without using LUKS and LVM?


          Comment


          • #15
            Originally posted by intelfx View Post

            Cool story. Except you don't need LVM with btrfs (why would you even??), and you say it yourself that you still need LUKS with ZFS. And that's completely irrelevant to what makes a technology "next-gen" or not.
            I've read the BTRFS documentation. They go into detail as to why you would even. I suggest you read it. TLDR: Just because they have similar features doesn't mean they have a perfect feature overlap or that features that they share work the same.

            I get it that you want to ascribe every last bit of your experience to perceived ZFS' greatness, but how is this different from literally any other non-brain-dead filesystem out there?
            How about you share something cool you've done with BTRFS over the course of a half-dozen years instead of shitting on what I've done? Someone asked what I've used it for and I shared my experience on how great and flexible ZFS can be in regards to disks, partitions, and upgrading mirrors and raids. I could go into how I think ZFS is also useful for creating case-insensitive volumes for Wine because Wine gets a slight performance increase if it gets to act like Windows and not care. You can only do that on Ext4 with custom formatting options. That's why the Steam Deck uses Ext4 for home and BTRFS for A/B root partitioning.

            Whoops, a ZFS fan just shared a neat use of BTRFS -- Steam Deck A/B root partitioning

            Comment


            • #16
              Originally posted by cynic View Post

              would you mind sharing an example of something you can do with ZFS that is not feasible with btrfs without using LUKS and LVM?

              Would you?

              There's so much feature overlap I feel like that's a loaded question.

              Comment


              • #17
                Originally posted by skeevy420 View Post
                BTRFS I've always used for root and have been hit by GRUB and compression issues in the past. Sometimes it's from using esoteric setups and other times it's simply due to using a rolling release distribution and finding out a bug exists the hard way. Whatever the case may be, it has happened enough in the past 10 years that I've gone back to using Ext4 for my root volumes. Ext4 just works and I've never been hit by an issue so bad that a reinstall of the OS seemed like the easier fix.
                Do you use EXT4 for your root and ZFS for other partitions? Can you explain a bit more and share how your hard disk structure is? Thanks!

                Comment


                • #18
                  Originally posted by skeevy420 View Post

                  Would you?

                  There's so much feature overlap I feel like that's a loaded question.
                  ok, got it: you just wrote something that you cannot support with facts.

                  Comment


                  • #19
                    Originally posted by skeevy420 View Post

                    I've read the BTRFS documentation. They go into detail as to why you would even. I suggest you read it. TLDR: Just because they have similar features doesn't mean they have a perfect feature overlap or that features that they share work the same.
                    Aaand... what? All of those points apply to ZFS equally.

                    Originally posted by skeevy420 View Post
                    How about you share something cool you've done with BTRFS over the course of a half-dozen years instead of shitting on what I've done? Someone asked what I've used it for and I shared my experience on how great and flexible ZFS can be in regards to disks, partitions, and upgrading mirrors and raids.
                    I'm an equal opportunity skeptic. I find and point out logical flaws in any kind of statements, regardless of who was the author or what sort of emotional attachment do they have to their opinion.

                    I don't care how "cool" do you think your use-case is. I'm simply stating that there is nothing special in it. Everything you said, and more, can be achieved with btrfs (even more, because unlike ZFS, btrfs permits transparent conversion of data between storage profiles).

                    Comment


                    • #20
                      What's the status of BTRFS when it comes to:
                      - RAID 5 & 6 bug - Write Caching on separate disks
                      - Read Caching on separate disks

                      ?

                      Comment

                      Working...
                      X