Re: SGI Altix and Apple Xserve RAID? - SGI

This is a discussion on Re: SGI Altix and Apple Xserve RAID? - SGI ; The following header lines retained to effect attribution: >Date: Mon, 14 Jun 2004 18:39:08 +0000 (GMT) >From: "Tony 'Nicoya' Mantler" >Subject: Re: SGI Altix and Apple Xserve RAID? >To: info-iris-admin@ARL.ARMY.MIL >X-Spam-Level: >In article , > "Randolph J. Herber" wrote: >: ...

+ Reply to Thread
Results 1 to 2 of 2

Thread: Re: SGI Altix and Apple Xserve RAID?

  1. Re: SGI Altix and Apple Xserve RAID?

    The following header lines retained to effect attribution:
    >Date: Mon, 14 Jun 2004 18:39:08 +0000 (GMT)
    >From: "Tony 'Nicoya' Mantler"
    >Subject: Re: SGI Altix and Apple Xserve RAID?
    >To: info-iris-admin@ARL.ARMY.MIL
    >X-Spam-Level:


    >In article ,
    > "Randolph J. Herber" wrote:


    >: >http://www.t10.org/ftp/t10/drafts/sbc2/sbc2r14.pdf


    >: >See section 5, the READ/WRITE/ETC (16) and (32) commands allow 8 bytes for
    >: >LBA.


    >: >Also, section 4.4 states:


    >: >"Logical block addresses are no larger than 8 bytes. Some commands support
    >: >only
    >: >4 byte LOGICAL BLOCK ADDRESS fields (e.g., READ CAPACITY (10), READ (10),
    >: >and
    >: >WRITE (10))."


    >: I would have thought that the 4 byte LOGICAL BLOCK ADDRESS fields
    >: for reading and writing would have been the problem.


    >: 2^9 * 2^32 = 2^41 = 2TiB = 2199023255552B


    >Read more closely. It is only the READ/WRITE/ETC (6), (10) and (12) commands
    >that are limited.


    I did read closely. That is how I noticed that limiting
    factor was short READ and WRITE commands. What you omitted
    from your posting was the signifiance of the parenthensized
    numbers. Those are the lengths of the commands. SCSI-3
    permits a length of 16 bytes for a command descriptor
    block as well.

    http://www.t10.org/ftp/t10/drafts/spc3/spc3r12.pdf

    Page 22 shows the Long LBA 16-byte commands with an
    8-byte LBA. Page 256 shows further data.

    By reading the document closely, it is possible to extract the
    information about parenthensized numbers. For equipment that
    supports the SCSI-3 standard and these large size command
    descriptor blocks (the standard leaves the impression that the
    devices do not have to support them), then:

    2^9 * 2^(8*8) = 9TiB = 9444732965739290427392B
    (as you stated earlier).

    >The (6) commands are limited to 21 bit LBAs, while the (10) and (12) commands
    >are limited to 4-byte/32 bit LBAs. As I noted above, the (16) and (32) commands
    >allow a full 8-byte/64-bit LBA.


    >Cheers - Tony 'Nicoya' Mantler


    >Tony 'Nicoya' Mantler -- Master of Code-fu -- nicoya@ubb.ca
    >-- http://nicoya.feline.pp.se/ -- http://www.ubb.ca/ --


    Randolph J. Herber, herber@fnal.gov, +1 630 840 2966, CD/CDFTF PK-149F,
    Mail Stop 318, Fermilab, Kirk & Pine Rds., PO Box 500, Batavia, IL 60510-0500,
    USA. (Speaking for myself and not for US, US DOE, FNAL nor URA.) (Product,
    trade, or service marks herein belong to their respective owners.)

  2. Re: SGI Altix and Apple Xserve RAID?

    In article ,
    "Randolph J. Herber" wrote:

    : I did read closely. That is how I noticed that limiting
    : factor was short READ and WRITE commands. What you omitted
    : from your posting was the signifiance of the parenthensized
    : numbers. Those are the lengths of the commands. SCSI-3
    : permits a length of 16 bytes for a command descriptor
    : block as well.

    Sorry, I should have been more clear on that. I assumed that anyone with basic
    knowledge of modern SCSI would remember that there are multiple command lengths,
    or would be able to pick it up from the context.


    : 2^9 * 2^(8*8) = 9TiB = 9444732965739290427392B

    8ZiB, 8192EiB, 8388608PiB, 8589934592TiB, 8796093022208GiB, 9007199254740992MiB,
    9223372036854775808KiB, 9444732965739290427392B.

    If you insist on using the 'i' style prefixes (which haven't been widely adopted
    outside the IEC and certain circles of pedants), please remember to actually
    divide by 1024, and to use the proper magnitude.


    Cheers - Tony 'Nicoya' Mantler

    --
    Tony 'Nicoya' Mantler -- Master of Code-fu -- nicoya@ubb.ca
    -- http://nicoya.feline.pp.se/ -- http://www.ubb.ca/ --

+ Reply to Thread