RE: Huge Partition

D_at_7@k|N&
Date: 04/21/04

  • Next message: -=Brian Truter=-: "RE: XDMCP broadcast"
    To: "'For users of Fedora Core releases'" <fedora-list@redhat.com>
    Date: Wed, 21 Apr 2004 07:36:14 -0700
    
    

     
    -----BEGIN PGP SIGNED MESSAGE-----
    Hash: SHA1

    Would the filesystem type make a difference here? Or are the problems happening at a lower level than the filesystem? I don't know ext3's limitations are, but it seems like this might be better handled with a different filesystem. Just a thought.

    - -
    - -=D@7@k|N&=-
     
     

    - -----Original Message-----
    From: fedora-list-bounces@redhat.com [mailto:fedora-list-bounces@redhat.com] On Behalf Of Rick Mohr
    Sent: Wednesday, April 21, 2004 7:00 AM
    To: For users of Fedora Core releases
    Subject: RE: Huge Partition

    OK, I haven't dealt with this on Fedora, but I did run across a similar problem when I installed RHAS3. I had a 3Ware card with about 1.5TB of disk space, most of which I wanted on a single partition. The anaconda installer couldn't handle it, but I believe I read something about how there used to be a limit hardcoded into it that caused the problem. (Don't know if this was fixed or not.)

    I also tried creating the partition after installation, and that also failed. Actually, I used parted and it appeared to work. But when I formatted it as ext3 it would always hang part way through. There were also a bunch of I/O error messages in the syslogs.

    I ended up finding a reference to the problem on 3Ware's web site (although I can't seem to track down the URL now). It mentioned that there was a problem with using 'int' instead of 'unsigned long' in the sd.c file in the scsi driver. The suggested fix was this:

    static int sd_init_command(Scsi_Cmnd * SCpnt)
    {
            /* int dev, block, this_count; */
            int dev, this_count;
            unsigned long block;

    I am not sure if that is the proper solution, but it did eliminate the I/O errors from my logs. However, I have not come close to filling the partition up, so I cannot verify that all the space is actually usable.

    Oh, and my /proc/partitions file still shows a negative number of blocks for this partition even after I made the change.

    Hope that helps.

    - --Rick

    On Wed, 21 Apr 2004, Bill Gradwohl wrote:

    > The testing I started yesterday afternoon completed, and I'm sorry to
    > say the result is a failure.
    >
    > Although parted created a 2TB partition as verified by fdisk, only 1TB
    > is useable. I wrote 1001 files of just slightly under 1G each when the
    > script failed for out of disk space. The script simply used the cp
    > command to copy a single master file to FILE.XXXX where XXXX is an
    > increasing number.
    >
    > I tried googling for disk space limits, but got too many bogus hits to
    > be of value.
    >
    > Does anyone know what the "official" disk space limits are? Does
    > anyone know what I need to do to get a single partition of greater
    > than 1TB? I have a test platform with 2TB of physical disk if someone
    > can point me in the right direction to recompile whatever.
    >
    > Martin:
    >
    > If you have a 2TB partition, have you tested it for availability
    > beyond 1TB?
    >
    > --
    > Bill Gradwohl
    > YCC
    > (817) 224-9400 x211
    > www.ycc.com
    > SPAMstomper Protected E-mail
    > www.stomperware.com
    >
    >
    >
    > --
    > fedora-list mailing list
    > fedora-list@redhat.com
    > To unsubscribe: http://www.redhat.com/mailman/listinfo/fedora-list
    >
    >

    - --
    fedora-list mailing list
    fedora-list@redhat.com
    To unsubscribe: http://www.redhat.com/mailman/listinfo/fedora-list

    - ----------------------------------------
    My Inbox is protected by SPAMfighter
    4308 spam mails have been blocked so far.
    Download free www.spamfighter.com today!

    -----BEGIN PGP SIGNATURE-----
    Version: PGP 8.0

    iQA/AwUBQIaG3qbnQHwVWKJjEQJcPgCgml/HK6YoR9oG3WD37gW844hp2KgAoPKd
    qmg7622IkYNg4mkastY9eSE3
    =2D/P
    -----END PGP SIGNATURE-----

    -- 
    fedora-list mailing list
    fedora-list@redhat.com
    To unsubscribe: http://www.redhat.com/mailman/listinfo/fedora-list
    

  • Next message: -=Brian Truter=-: "RE: XDMCP broadcast"

    Relevant Pages

    • RE: Large discrepancy in reported disk usage on USR partition
      ... tunefs: maximum blocks per file in a cylinder group: 2048 ... Is it possible on UFS2 for disk space to be allocated but hidden somehow? ... Large discrepancy in reported disk usage on USR partition ... Filesystem 1024-blocks Used Avail Capacity Mounted on ...
      (freebsd-questions)
    • RE: lvm on RH8 _ SOLVED_
      ... Apparently, the lvm module wasn't getting initialised at boot, so I ... Using fdisk, change the partition usage to 0x8e (free space, logical ... Next the filesystem needs to be created in the logical volumes. ...
      (RedHat)
    • Re: checking total and free space
      ... The df command will tell you how much space in the filesystem ... It tells you nothing about unallocated disk space. ... >> and the size of each partition to get the free space. ... Partition Tag Flags Sector Count Sector Mount Directory ...
      (comp.unix.solaris)
    • Re: Hard Drive Question
      ... > hardlined partition logic is usually overkill and needlessly complicated. ... * To impose restrictions on the size of certain directory subtrees (i.e. /tmp ... amount of filesystem checked at any one time, ... * To provide alternate (or recovery) directory subtrees, ...
      (alt.os.linux)
    • read/only root filesystem strategy
      ... the installed system as possible be running under a read/only mount. ... or a widely shared network filesystem. ... Modifying the kernel to perform the setup. ... Partition hda1 would house the kernel and a small rescue ...
      (comp.os.linux.development.system)