[SLE] Large Block Device Support [WAS: Largefile support on SUSE]

From: Greg Freemyer (freemyer-ml_at_NorcrossGroup.com)
Date: 10/30/03

  • Next message: Fred Miller: "Re: [SLE] 9.0 and nVidia"
    To: suse-linux-e@suse.com
    Date: 29 Oct 2003 19:18:55 -0500
    
    

    If anyone cares besides me, I found these limits with x86:

    Vanilla 2.4.20 kernel
        2 TB
    With LBD kernel patch --
        Effectively None. ie. Multiple PBs or bigger
    Vanilla 2.6.0 kernel with LBD_CONFIG off
        2 TB
    With LBD_CONFIG ON --
        Effectively None. ie. Multiple PBs or bigger
    XFS --
        Effectively None. ie. Multiple PBs or bigger
    DM --
        2 TB even in 2.6 (not sure about this)
    LVM --
        1 PB, but you requires large extents
    NFS --
        No restriction
    Samba --
        No restriction.

    Also, the actual hardware driver (SCSI, etc.) can limit to 2 TB, or even
    1 TB.

    Greg

    On Fri, 2003-10-17 at 17:07, Greg Freemyer wrote:

    > SuSE 9.0 (2.4 kernel):
    > XFS -- 2 TB (I assume)
    > XFS w/LBD patch -- greater than 2 TB (I assume)
    > XFS w/LBD patch on top of MD -- ??
    > XFS w/LBD patch on top of LVM -- 1 PB
    > XFS w/LBD patch on top of DM (w/DM patch) -- ??
    > NFS exported XFS -- ??
    > Samba exported XFS -- ??
    >
    > SuSE 9.0 (2.6 test kernel (includes LBD patch)):
    > XFS -- ??
    > XFS on top of MD -- ??
    > XFS on top of DM (assuming DM kernel patches installed) -- ??
    > NFS exported XFS -- ??
    > Samba exported XFS -- ??
    >
    > Thanks
    > Greg
    > --
    > Greg Freemyer
    >
    >
    > On Fri, 2003-10-17 at 10:20, Tony Vickers wrote:
    > > For Large file systems use XFS from SGI. I have files that are 65 Gig in
    > > size.
    > >
    > > --Tony
    > >
    > > On Fri, 2003-10-17 at 02:41, Nick Katsamas wrote:
    > > > Hi,
    > > >
    > > > I have Suse 7.3 with kernel 2.4.19 compiled by my own, and
    > > > i want to create a 5GB file.
    > > >
    > > > I use the command:
    > > >
    > > > dd if=/dev/zero of=largefile bs=1024k count=5120
    > > >
    > > > and this is the answer that i get:
    > > >
    > > > dd: writing `largefile': File too large
    > > > 2049+0 records in
    > > > 2048+0 records out
    > > >
    > > > ls -h
    > > >
    > > > 2.0G largefile
    > > >
    > > > I try as root and as normal user.
    > > >
    > > > How can i create this file under Linux?
    > > >
    > > > --
    > > > Best regards
    > > > ****************************
    > > > Nick Katsamas
    > > > e-Mail: nkat@forthnet.gr
    > > > ****************************
    > > --
    > >
    > >
    > > --Tony
    > >
    > > Tony Vickers 866-585-1771 Office
    > > Chief Technology Officer 619-934-5276 Fax
    > > I.D.L. Technology Group, Inc. 619-672-1027 Cell
    > > http://www.idltechnology.com
    > >
    > > "Solutions and Services that Fit."
    > >
    > > This message was sent using SuSE Linux Openexchange. See how Openexchange
    > > can change the way you think about doing business. http://www.idltechnology.com
    >

    -- 
    Check the headers for your unsubscription address
    For additional commands send e-mail to suse-linux-e-help@suse.com
    Also check the archives at http://lists.suse.com
    Please read the FAQs: suse-linux-e-faq@suse.com
    

  • Next message: Fred Miller: "Re: [SLE] 9.0 and nVidia"

    Relevant Pages

    • Re: [patch] voluntary-preempt-2.6.9-rc1-bk4-R0
      ... >> patch for the upstream or BK tree during that process, as small patches ... > fixes should apply just as easily to -mm as vanilla. ... > will certainly be using the vanilla kernel. ... > AGNULA people are planning to start distributing test VP-kernel packages ...
      (Linux-Kernel)
    • Re: [RFC][PATCH 2/7] UBC: core (structures, API)
      ... Basic structures: ... It provides some kind of soft and hard limits. ... * Kernel internal part. ... I presume UB_RESOURCES value is going to change as different resources ...
      (Linux-Kernel)
    • Re: [RFC][PATCH 2/7] UBC: core (structures, API)
      ... It provides some kind of soft and hard limits. ... * Kernel internal part. ... I presume UB_RESOURCES value is going to change as different resources ... ...just that user land will need to be some how informed about that. ...
      (Linux-Kernel)
    • Re: PNP: increase the maximum number of resources
      ... And the kernel silently ignored a BIOS exceeding these limits ... And the kernel complains with KERN_ERR when a BIOS has more resources ... Thomas worked on a patch to make resource allocation dynamic and do away ...
      (Linux-Kernel)
    • Re: Finding kernel memory starting address on HPUX 11.0
      ... code that compiles under more than one compiler ... data limits initially are the ... There aren't really enforced sizes on shared memory usage (or anonymous ... Again - as with all kernel tunables, ...
      (comp.sys.hp.hpux)