Page 1 of 2 12 LastLast
Results 1 to 10 of 11

Thread: The First Benchmarks Of Red Hat Enterprise Linux 6.0

  1. #1
    Join Date
    Jan 2007
    Posts
    14,369

    Default The First Benchmarks Of Red Hat Enterprise Linux 6.0

    Phoronix: The First Benchmarks Of Red Hat Enterprise Linux 6.0

    The first beta release of Red Hat Enterprise Linux 6.0 was made available yesterday morning. RHEL 6.0 is set to offer many virtualization enhancements, power management improvements, new security features, many package updates, and even some reported performance enhancements. With Red Hat mentioning this major upgrade to their enterprise operating system carrying "performance enhancements", these claims have now been tested using the Phoronix Test Suite within our labs. There are some improvements for Red Hat Enterprise Linux 6.0 to note, but also some losses.

    http://www.phoronix.com/vr.php?view=14826

  2. #2
    Join Date
    Mar 2010
    Posts
    60

    Default

    does anyone have a good link how to implement this nobarrier option referenced in the article.


    I'm guessing is through fstab, but i've not found a good guide yet.

    Cheers

  3. #3
    Join Date
    Dec 2007
    Location
    Edinburgh, Scotland
    Posts
    574

    Default

    Please could you update the title, you've benchmarked the beta not the final product like it suggests

  4. #4
    Join Date
    Apr 2010
    Posts
    1

    Default

    Thanks for the review. But when the review articles always claim something like "the performance loss of the new distro is due to the EXT4 vs EXT3 file system", could you add the same test on the same distro with EXT3 then? There might be an article to compare different file systems, but putting the same test/distro with different FS would be much clear and better than words.
    My $0.02.

  5. #5
    Join Date
    Nov 2008
    Location
    Madison, WI, USA
    Posts
    862

    Default

    Quote Originally Posted by megakilo View Post
    Thanks for the review. But when the review articles always claim something like "the performance loss of the new distro is due to the EXT4 vs EXT3 file system", could you add the same test on the same distro with EXT3 then? There might be an article to compare different file systems, but putting the same test/distro with different FS would be much clear and better than words.
    My $0.02.
    I know this doesn't respond to your request for addition of multiple FS in the cross-distribution reviews, but it should give you an idea about the relative performance of current filesystems:

    http://www.phoronix.com/scan.php?pag..._2634_fs&num=1

  6. #6

    Default

    Why is it that whenever Phoronix sees a massive performance regression, they immediately speculate on what is causing it rather than redoing the benchmarks with their speculated cause fixed so people can see whether or not Phoronix's speculation is the case?

    The regressions in Apache Bench merit adding benchmarks for the RHEL 6.0 and Fedora where the file systems are ext4 with nobarrier and ext3 with/without nobarrier so we can see to what extent things are being affected by changes in the file systems and to what extent things are helped/harmed by attempting to undo them. Despite that, Phoronix never does this.

  7. #7

    Default

    Quote Originally Posted by xir_ View Post
    does anyone have a good link how to implement this nobarrier option referenced in the article.


    I'm guessing is through fstab, but i've not found a good guide yet.

    Cheers
    http://www.kernel.org/doc/Documentat...stems/ext4.txt

    I used few options from here and I put them into fstab and they're working.

    ext4 noatime,errors=remount-ro 0 1

  8. #8

    Default

    And from here:

    - When comparing performance with other filesystems, remember that
    ext3/4 by default offers higher data integrity guarantees than most. So
    when comparing with a metadata-only journalling filesystem, use `mount -o
    data=writeback'. And you might as well use `mount -o nobh' too along
    with it. Making the journal larger than the mke2fs default often helps
    http://lwn.net/Articles/203915/
    performance with metadata-intensive workloads.

  9. #9
    Join Date
    Sep 2008
    Posts
    130

    Default

    Quote Originally Posted by xir_ View Post
    does anyone have a good link how to implement this nobarrier option referenced in the article.


    I'm guessing is through fstab, but i've not found a good guide yet.

    Cheers
    man mount

    10 char limit

  10. #10
    Join Date
    Mar 2010
    Posts
    60

    Default

    cheers for the replies guys. I really do appreciate the help.

    Unfortunately enabling this option didn't improve a serious performance problem i'm having with disk access slowing my system to the point of uselessness.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •