Announcement

Collapse
No announcement yet.

Btrfs RAID: Linux 3.10 To Linux 3.18 Benchmarks

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

  • Btrfs RAID: Linux 3.10 To Linux 3.18 Benchmarks

    Phoronix: Btrfs RAID: Linux 3.10 To Linux 3.18 Benchmarks

    As a follow-up to this week's Btrfs RAID HDD testing on Ubuntu 14.10, I ran some benchmarks of Btrfs in RAID0 while benchmarking every major kernel release from Linux 3.10 to Linux 3.18-rc1.

    http://www.phoronix.com/vr.php?view=21078

  • #2
    Btrfs needs more corporate support

    For all the brouhaha over Btrfs being the future of the Linux filesystem (including from Theodore Tso of EXT4 fame as far back as 2009 https://lists.fedoraproject.org/pipe...ch/009583.html ), it seems that corporate support and improvement patches are still going to EXT4 instead of Btrfs. If Btrfs is going to get any better, it needs big companies like IBM, Redhat, SUSE and Oracle to pay their developers to work on Btrfs.

    Comment


    • #3
      Originally posted by phoronix View Post
      Phoronix: Btrfs RAID: Linux 3.10 To Linux 3.18 Benchmarks

      As a follow-up to this week's Btrfs RAID HDD testing on Ubuntu 14.10, I ran some benchmarks of Btrfs in RAID0 while benchmarking every major kernel release from Linux 3.10 to Linux 3.18-rc1.

      http://www.phoronix.com/vr.php?view=21078
      Now it would be good to a benchmark using postresql and/or mysql/sqlite and/or apt.

      My perception is that doing an upgrade of a system with btrfs takes like double the time compared to and ext system. I also have the impression that a system with a database also takes a large hit, but that that has improved significantly over time.

      Comment


      • #4
        Ummm. Facebook uses and supports Btrfs. Is that big enough for you? Or do you need more examples?

        Comment


        • #5
          Originally posted by ua=42 View Post
          Ummm. Facebook uses and supports Btrfs. Is that big enough for you? Or do you need more examples?
          It's great that Facebook uses and contributes to Btrfs. But overall, the number of patches being submitted to the Ext 4 mailing list is significantly higher than on the Btrfs mailing list, indicating a higher corporate support for Ext4 than for Btrfs. The ratio of needy amateurs (complaining that things are broken) to professionals (submitting patches to improve things) is also much higher on the Btrfs mailing list.

          Comment


          • #6
            What's the current state of RAID 5/6 in BTRFS? It is already suitable for replacing ZFS in small NAS servers?

            Comment


            • #7
              Originally posted by stan View Post
              It's great that Facebook uses and contributes to Btrfs. But overall, the number of patches being submitted to the Ext 4 mailing list is significantly higher than on the Btrfs mailing list, indicating a higher corporate support for Ext4 than for Btrfs. The ratio of needy amateurs (complaining that things are broken) to professionals (submitting patches to improve things) is also much higher on the Btrfs mailing list.
              ext4 is good, simple and stable file system, not everyone needs the futures with have Btrfs. Btrfs will never replace ext4 for simple desktop users, for Android etc. Well f2fs have a higher chance to replace ext4 in this place.

              Comment


              • #8
                Originally posted by newwen View Post
                What's the current state of RAID 5/6 in BTRFS? It is already suitable for replacing ZFS in small NAS servers?
                It's still at a proof of concept level. It can't repair itself automatically or detect and drop a malfunctioning drive automatically. Stay away from it.

                Comment


                • #9
                  Raid 0

                  What's the point of testing Raid 0? There is no drive level redundancy and this doesn't test the efficiency of parity calculations. I'd much rather see a RAID5 performance comparison between BTRFS and MDRAID.

                  Comment


                  • #10
                    Originally posted by sheldonl View Post
                    What's the point of testing Raid 0? There is no drive level redundancy and this doesn't test the efficiency of parity calculations. I'd much rather see a RAID5 performance comparison between BTRFS and MDRAID.
                    I agree a RAID 0 test is of little use to most people. At best it maybe gives some insight into how the more useful RAID levels would perform. I would love to see how BTRFS RAID 10 has performed across versions.

                    Comment

                    Working...
                    X