Announcement

Collapse
No announcement yet.

Benchmarking Debian's GNU/kFreeBSD

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

  • Benchmarking Debian's GNU/kFreeBSD

    Phoronix: Benchmarking Debian's GNU/kFreeBSD

    There has been an effort underway within the Debian development community to pull the FreeBSD kernel within this distribution to provide an alternative to using the Linux kernel. In essence with this Debian GNU/kFreeBSD project you have the standard Debian package set providing a GNU user-land with a GNU C library, but the FreeBSD kernel is running underneath. The Debian project has also been working on Debian GNU/Hurd to effectively do the same thing but with the GNU Mach microkernel. But unlike Debian GNU/Hurd, with the release of Debian 6.0 "Squeeze", Debian GNU/kFreeBSD will reach a release status. With the Debian Squeeze release being just two months away we have decided to provide the first public set of benchmarks that compare the Debian GNU/kFreeBSD performance to that of Debian GNU/Linux. We have tested both the 32-bit and 64-bit builds of Debian with the Linux and FreeBSD kernels.

    http://www.phoronix.com/vr.php?view=14511

  • #2
    Another analyse-yourself-the-results benchmark.
    It's sad Phoronix doesn't make the most of PTS. Describing figures is far less interesting than interpreting them : I'm not blind and the vast majority of your reader either I guess.

    Comment


    • #3
      64 bit vs 32 bit

      Personally, the thing I found most interesting with these results was that in most benchmarks there was a bigger difference between 32 bit vs 64 bit than between Linux vs kFreeBSD...

      Comment


      • #4
        Disclaimer: This is not a sarcastic question.

        Why is Debian porting the FreeBSD kernel to Debian?

        The reason I ask this is because:
        - It doesn't look like the FreeBSD kernel is better so there has to be another reason.
        - Last time I checked Hurd was supposed to be the next generation über-kernel, so investing money/time in FreeBSD must have a good reason.

        EDIT:
        Nevermind, I found the answer:
        "Debian's main motivation for the inclusion of the FreeBSD kernel into the official release process is the opportunity to offer to its users a broader choice of kernels and also include a kernel that provides features such as jails, the OpenBSD Packet Filter and support for NDIS drivers in the mainline kernel with full support."
        Last edited by MaestroMaus; 01-18-2010, 07:21 AM.

        Comment


        • #5
          Originally posted by MaestroMaus View Post
          "Debian's main motivation for the inclusion of the FreeBSD kernel into the official release process is the opportunity to offer to its users a broader choice of kernels and also include a kernel that provides features such as jails, the OpenBSD Packet Filter and support for NDIS drivers in the mainline kernel with full support."
          This way, both kernels' developers have a target to compare (and perhaps "compete") with, using the same user space software.

          Thanks for this test!

          Comment


          • #6
            About the only outstanding problem we have at this point is NVIDIA's binary FreeBSD driver not working with Debian GNU/kFreeBSD.
            So what about the state of the foss-drivers (radeon, intel, nouveou) in the *BSDs? Most recent development seems to be targeted at moving features from user-space (portable) to the (linux-)kernel (less portable), like kernel-based-mode-setting and memory-management. How far behind are their *BSD equivalents? This might even be worth a separate article, as that'd make a pretty interesting read imho (contrary to boring graphs with little to no explanation/interpretation).

            Comment


            • #7
              IIRC the KMS stuff will be a huge gain for OpenBSD, since they can now do 3D without allowing a userspace process to scribble all over VRAM unchecked.

              Comment


              • #8
                Honestly, this "analyse-it-yourself" stuff gets dull. We're not blind, and for the most part probably not stupid. Pretty much every other major site that does benchmarking sits and actually analyses and explains the results far beyond "Oh look it's 5.8x faster than xxx". If you're not in a position to be able to provide such analysis for whatever reason, just chuck up the graphs and stuff and do your usual summary/conclusion at the end.

                Some of the Debian GNU/kFreeBSD losses were in the area of disk performance where the common UFS file-system was outdone by EXT3, which is currently the default choice for Debian GNU/Linux. The results may have shaken out differently had the ZFS file-system been used, but unfortunately that is not a viable option at this point with the Debian GNU/kFreeBSD project, and if EXT4 was the default for Debian GNU/Linux it would also be interesting.
                Still likely to be slower though. ZFS deliberately chooses to sacrifice some speed for the benefit of overall file system stability.

                Comment


                • #9
                  Originally posted by Garp View Post
                  ZFS deliberately chooses to sacrifice some speed for the benefit of overall file system stability.
                  ZFS deliberately chose to sacrifice some freedom by using the CDDL license (http://en.wikipedia.org/wiki/CDDL):

                  In the words of Danese Cooper, who is no longer with Sun, one of the reasons for basing the CDDL on the Mozilla license was that the Mozilla license is GPL-incompatible. Cooper stated, at the 6th annual Debian conference, that the engineers who had written the Solaris kernel requested that the license of OpenSolaris be GPL-incompatible. "Mozilla was selected partially because it is GPL incompatible. That was part of the design when they released OpenSolaris.
                  This is also why Debian probably never will include ZFS "natively" anyhow, but rather upgrade their default file system to ext4, which btw was good enough for Google, as noted by the trolling Stephen J. Vaughan-Nichols, in "The best Linux file system of all?" http://blogs.computerworld.com/15413..._system_of_all and who quoted Google's Michael Rubin, a senior staff engineer:

                  Google is currently in the middle of upgrading from ext2 to a more up to date file system. We ended up choosing ext4.
                  So, despite all the hype of ZFS, maybe Google had their reasons.

                  Comment


                  • #10
                    Originally posted by sabriah View Post
                    ZFS deliberately chose to sacrifice some freedom by using the CDDL license (http://en.wikipedia.org/wiki/CDDL):

                    This is also why Debian probably never will include ZFS "natively" anyhow, but rather upgrade their default file system to ext4, which btw was good enough for Google, as noted by the trolling Stephen J. Vaughan-Nichols, in "The best Linux file system of all?" http://blogs.computerworld.com/15413..._system_of_all and who quoted Google's Michael Rubin, a senior staff engineer:

                    So, despite all the hype of ZFS, maybe Google had their reasons.
                    Certainly don't deny any of that. My point was merely that ZFS architects made deliberate choices that result in a performance hit on the basis that the performance hit was worth the benefits in data safety.

                    As to whether Debian will or will not influde ZFS natively, they sure as heck will never do it under their "Linux" flavour due to licensing differences.

                    Google's choice of ext4 is rather interesting, particularly given the serious performance regressions in the more recent kernels. I wonder if they're not just lining themselves up for a world of hurt. They've done benchmarking for their choice, but very importantly point out that it's the best file system for their particular needs and intended use. Doesn't make it the best file system every need. For that matter I still tend to prefer XFS over ext3 due to it's perfomance advantages in certain areas.

                    Like with the Windows/FreeBSD/Linux etc. arguments galore, no one size fits all. All have advantages and disadvantages. People would be best served by looking at all of them and evaluating them for their specific needs.

                    Comment


                    • #11
                      Originally posted by sabriah View Post
                      ZFS deliberately chose to sacrifice some freedom by using the CDDL license (http://en.wikipedia.org/wiki/CDDL):
                      Still, the CDDL license is Free Software (according to FSF) and Open Source (according to OSI) and is less restrictive than GPLv2, let alone GPLv3 - for example, it doesn't prohibit one from linking with code under any other license, which is what GPL does.


                      Originally posted by sabriah View Post
                      This is also why Debian probably never will include ZFS "natively" anyhow, but rather upgrade their default file system to ext4, which btw was good enough for Google, as noted by the trolling Stephen J. Vaughan-Nichols, in "The best Linux file system of all?" http://blogs.computerworld.com/15413..._system_of_all and who quoted Google's Michael Rubin, a senior staff engineer:



                      So, despite all the hype of ZFS, maybe Google had their reasons.
                      Google has very different requirements compared to the usual server uses - for example, their data is replicated between machines, so reliability of a single machine is less important - and reliability (snapshots, checksums, built-in RAID etc) is a very important part of ZFS. Also, I don't think Google needs all the flexibility ZFS provides. They just want to replace 1980's technology (ext2 is basically a simplified version of the original UFS, released in 1984, without most of the features UFS got during the following 20 years) with something more 1990's (extents, journalling, defragmentation - VxFS had all this over a decade ago).

                      Comment


                      • #12
                        Hello,

                        In the article, you wrote a couple of time that Debian Squeeze would be released in March. This is most probably wrong, as March 2010 is only the beginning of the Testing Freeze. The release usually comes a few months after the freeze.

                        http://lists.debian.org/debian-devel.../msg00002.html

                        Comment


                        • #13
                          Originally posted by Garp View Post
                          Google's choice of ext4 is rather interesting, particularly given the serious performance regressions in the more recent kernels.
                          This is just a matter of the default configurations of Ubuntu.

                          Comment


                          • #14
                            Overload?

                            Is this the reason why Debian-stable is a year behind everyone else?

                            **shot**

                            Not only do they support two dozen arcitectures, but now they'll have three kernels?

                            (for the ill-humored, I'm using hyperbole)

                            Comment


                            • #15
                              Originally posted by MaestroMaus View Post
                              - Last time I checked Hurd was supposed to be the next generation über-kernel, so investing money/time in FreeBSD must have a good reason.
                              Hurd... in development since 1990.
                              20 years and there is YET to be a stable release.
                              BSD has been around and stable since.... about the beginning of time.

                              If they want a kernel option besides Linux, BSD is ready and waiting. Debian's been going around in circles with Hurd since they tried to pick it up back in 1998.


                              Seriously, I'll believe in Hurd when it actually becomes more than vaporware.

                              Comment

                              Working...
                              X