Announcement

Collapse
No announcement yet.

The Performance Of Ubuntu Software Running On Windows 10 With The New Linux Subsystem

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

  • fredm
    replied
    Originally posted by Michael View Post

    Phoronix Premium readers already know, but new benchmarks of it coming on Thursday.
    Hi,

    To reply to myself

    This Creative upgrade of Windows, bash on windows moved from ubuntu 14.04 to the ubuntu 16.04, but there is no improvement on the file system benchmark


    The benchmark filesystem, need to be done in the external folder than bash, the result seems better.
    https://superuser.com/questions/1114...windows-drives -> /mnt/d

    And also I tried to create a workspace in the $HOME but as if you try to create a file with windows explorer or other windows apps, in the "bash" you can't access to them.
    https://github.com/Microsoft/BashOnWindows/issues/45
    => confirmed that when using mklink /J to create a windows link to the Lxss/rootfs path, files could be updated but newly created files weren't visible in WSL.


    to share file edition/creation between windows and bash , it advices to use external folder.

    Leave a comment:


  • Michael
    replied
    Originally posted by FredericM View Post
    Hi,

    Would it be possible to have an update of this benchmark ?

    About the file system low result's, is there a method to improve it in bash on windows ?

    Regards,
    Frédéric
    Phoronix Premium readers already know, but new benchmarks of it coming on Thursday.

    Leave a comment:


  • FredericM
    replied
    Hi,

    Would it be possible to have an update of this benchmark ?

    About the file system low result's, is there a method to improve it in bash on windows ?

    Regards,
    Frédéric

    Leave a comment:


  • nasyt
    replied
    Originally posted by torsionbar28 View Post
    For example, every major OS out there has a software package manager that tracks all the installed files, with checksums, tracks dependencies, versioning, etc. Except Windows.
    Thats it. No software package manager and proud of it. These package manager are often despised by those, who come from the windows world.

    Originally posted by torsionbar28 View Post
    Every major OS out there has an LVM for storage management. Except Windows. etc. etc
    Recently, windows got one. It is called "Storage Spaces".

    https://technet.microsoft.com/en-us/...=ws.11%29.aspx

    Leave a comment:


  • sireangelus
    replied
    so... where is the so much famed linux superiority in resources overhead? Where did all that talk about so many windows service running into the background that will slow down your onw performance?... i'm disappointed. I do really believe in linux, but sometimes the difference is just not there.

    Leave a comment:


  • trek
    replied
    so in the end starting the wine project was a good move, so good that microsoft copied it

    probably the linux ecosystem is grown to a point that it is a good source of applications even for windows

    Leave a comment:


  • speculatrix
    replied
    I hope Microsoft decide to provide some sort of native Wayland compatibility layer rather than X11. And if Apple did something like that too a then it would mean Linux user space could be a better universal platform than Windows or OSX.

    Leave a comment:


  • p91paul
    replied
    Originally posted by duby229 View Post

    I don't understand, filesystem formats are really just a layout of bits on a medium. The filesystem itself doesn't have any performance, it's just a layout.
    Actually, a badly designed filesystem (i.e. the bits layout) can "have" bad performance if it makes difficult or impossible to write an efficient implementation.

    Originally posted by duby229 View Post
    The driver implementation which reads and writes that layout is where performance is measured. Does it matter which driver is used or what OS the driver runs on?
    Yes it does: as you say, the implementation is where performance is measured. If you test ntfs-3g on linux you have no idea of the performance of the windows driver because they are different implementations; you have no way of testing the actual ntfs windows driver on linux. And the OS probably matters because the driver might use kernel code, and the OS userspace calls the driver; so even the same driver could perform very differently in different OSes (because the OS userspace is using it right/wrong, or because kernel code the driver is calling is faster/slower).

    Since we are talking about measuring the performance of the ubuntu on windows layer, a good test would be to compare ubuntu on windows with plain windows. Ubuntu on Windows (I believe) uses the windows ntfs driver; so if windows is faster than ubuntu, probably that means that the conversion layer is not so efficient (but it could still be that linux applications optimize system calls to be fast on an ext4 filesystem, though that's less likely).

    Leave a comment:


  • duby229
    replied
    Originally posted by Kushan View Post

    No, I am saying Comparing the different code bases is simply not possible because one codebase only runs on Linux and one codebase only runs on Windows. That's not a fair test in the slightest, we still wouldn't know if it was NTFS that was slow or just Windows.
    I don't understand, filesystem formats are really just a layout of bits on a medium. The filesystem itself doesn't have any performance, it's just a layout. The driver implementation which reads and writes that layout is where performance is measured. Does it matter which driver is used or what OS the driver runs on?

    Leave a comment:


  • Kushan
    replied
    Originally posted by duby229 View Post

    Isn't comparing the different codebases the whole point? In your opinion then comparing anything with different code doesn't make sense? Amdgpu vs Catalyst? MS Office vs LibreOffice?

    Determining the capability and performance of different implementations at the settings you use is kinda the whole point.
    No, I am saying Comparing the different code bases is simply not possible because one codebase only runs on Linux and one codebase only runs on Windows. That's not a fair test in the slightest, we still wouldn't know if it was NTFS that was slow or just Windows.

    Leave a comment:

Working...
X