Announcement

Collapse
No announcement yet.

Benchmarking Debian's GNU/kFreeBSD

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

  • #21
    Originally posted by david.given View Post
    Some of those filesystem benchmarks look very dubious to me --- UFS shouldn't be that much slower than ext3. Can you confirm that you did have SoftUpdates turned on? Without that, you'll cripple the FreeBSD file system performance.
    Yeah, I agree. What's the partition layout that Phoronix uses? If you do just a single slice of / then by default softupdates are disabled.

    Comment


    • #22
      Phoronix always test the default settings of each OS/distro. Guess what would happen if this wa not the case. Others would complain that you should enable this and not that and others the oppossite. So the default settings is the best option since that's what the users have when they install the system. Most of them afterall stay with the default settings for ever.

      Comment


      • #23
        Originally posted by david.given View Post
        Without that, you'll cripple the FreeBSD file system performance.
        The same probably can be said about Ext3 if it wasn't using writeback mode etc.

        Comment


        • #24
          FreeBSD's UFS is well known for it's speed with soft updates enabled. It should have been enabled. Using FreeBSD's UFS without soft updates is like using ext4 with extents turned off. I would even be so bold as to say that Debian's installer is broken if the default is to have soft updates disabled. BTW, I have been running a Debian GNU/kFreeBSD server since Lenny was still in testing.

          Originally posted by Apopas View Post
          Most of them afterall stay with the default settings for ever.
          True if the user uses Windows or Mac OS X, or some other non-FOSS OS. FOSS OS's are designed to be tweaked and hacked. In fact, that's why most people use FOSS OS's. On a server, you will rarely ever go with the default install options, since most servers have a specific role, and you would want to tweak your server for that role. I do see your point for the sake of a comparison, but I sure hope you didn't intend that as a general statement.

          Comment


          • #25
            Originally posted by tux9656 View Post
            True if the user uses Windows or Mac OS X, or some other non-FOSS OS. FOSS OS's are designed to be tweaked and hacked. In fact, that's why most people use FOSS OS's. On a server, you will rarely ever go with the default install options, since most servers have a specific role, and you would want to tweak your server for that role. I do see your point for the sake of a comparison, but I sure hope you didn't intend that as a general statement.
            I agree with you and that's one of the greatest benefits of FOSS. But since the tweaking options cover a large amount of different settings, from filesystems to kernels and desktop environments, it's impossible to satisfy everyone.
            So the default options, maybe is not the best options around, but is something objective and at least it satisfies the default user.
            At least there is phoronix global around, to test our tweaked systems.

            Comment


            • #26
              We can go further and say Linux kernel was using non optimal config, file system, file system mount options etc. Linux kernel used here was also quite old.

              Comment

              Working...
              X