Announcement

Collapse
No announcement yet.

Btrfs vs. F2FS Multi-SSD Performance On Linux 4.11

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

  • Btrfs vs. F2FS Multi-SSD Performance On Linux 4.11

    Phoronix: Btrfs vs. F2FS Multi-SSD Performance On Linux 4.11

    Last week I posted benchmarks showing off F2FS performance with its multi-drive feature that isn't formal RAID but can still yield better I/O performance. For additional context, here are some results on that same system and with the Linux 4.11 kernel when using Btrfs with its native RAID capabilities.

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    Btrfs - Broken trash fs? All I see after all these years of development are performance that is erratic at best. Granted, there are other metrics to a fs besides performance. And some would rightfully argue that almost all other qualities beside speed are more important... but come on... Btrfs is painfully slow at times.

    Comment


    • #3
      Originally posted by milkylainen View Post
      Btrfs - Broken trash fs? All I see after all these years of development are performance that is erratic at best. Granted, there are other metrics to a fs besides performance. And some would rightfully argue that almost all other qualities beside speed are more important... but come on... Btrfs is painfully slow at times.
      Why people is still stupid after all these years? Compare it with ext4 with data=journal speed (that's similar to what CoW is) and add some penality because it is running checksumming on top.

      Of course it is slower than f2fs (where it can't boost its performance by RAID-ing anyway), only dumb people could think it could eventually outperform filesystems that lack its features.

      Comment


      • #4
        Originally posted by milkylainen View Post
        Btrfs - Broken trash fs? All I see after all these years of development are performance that is erratic at best. Granted, there are other metrics to a fs besides performance. And some would rightfully argue that almost all other qualities beside speed are more important... but come on... Btrfs is painfully slow at times.
        While I agree that btrfs can feel slow at times, if you look at the benches, there's only two tests where BTRFS has noticeably worse performance: The SQLite bench and the compile bench. Everything else put btrfs on par or better than f2fs.

        Comment


        • #5
          Originally posted by Niarbeht View Post
          While I agree that btrfs can feel slow at times, if you look at the benches, there's only two tests where BTRFS has noticeably worse performance: The SQLite bench and the compile bench. Everything else put btrfs on par or better than f2fs.
          That's only because btrfs is RAID-ing while F2FS is not (immature implementation I suppose).

          When F2FS's "RAID0" (i.e. spreading writes on all drives) is tuned it's going to bruhahahahahahah btrfs or anything else on speed alone.

          Comment


          • #6
            Not a single comment mentioning ZFS yet?
            Well... There you go.

            Oh.. @Michael why powersave intel_pstate?
            Last edited by Zucca; 01 May 2017, 05:53 PM. Reason: Added last line.

            Comment


            • #7
              Originally posted by milkylainen View Post
              Btrfs - Broken trash fs? All I see after all these years of development are performance that is erratic at best. Granted, there are other metrics to a fs besides performance. And some would rightfully argue that almost all other qualities beside speed are more important... but come on... Btrfs is painfully slow at times.
              Btrfs has to be functional, not fast. Development isn't focusing on performance. Show me transparent compression, snapshots, copy on write or deduplication in F2FS or another "fast" file system. Of course it has to be fast with those features. Nothing? How sad.

              Comment


              • #8
                What filesystem would you guys recommend when storing large git projects on SSD? I'm thinking of storing projects such as linux kernel, android (aosp), freedesktop projects. F2FS or BTRFS with LZO compression?

                Many times I have end up in situations where it takes forever to delete thousands of small files.

                Comment


                • #9
                  Originally posted by starshipeleven View Post
                  Why people is still stupid after all these years? Compare it with ext4 with data=journal speed (that's similar to what CoW is) and add some penality because it is running checksumming on top.

                  Of course it is slower than f2fs (where it can't boost its performance by RAID-ing anyway), only dumb people could think it could eventually outperform filesystems that lack its features.
                  Erratic, you mentally retarded yokel. I said both features and other qualities are more important. But as you apparently thrive on personal insults.

                  Comment


                  • #10
                    Originally posted by dragon321 View Post

                    Btrfs has to be functional, not fast. Development isn't focusing on performance. Show me transparent compression, snapshots, copy on write or deduplication in F2FS or another "fast" file system. Of course it has to be fast with those features. Nothing? How sad.
                    I know. And I said erratic.

                    Comment

                    Working...
                    X