1. This forum section is a read-only archive which contains old newsgroup posts. If you wish to post a query, please do so in one of our main forum sections (here). This way you will get a faster, better response from the members on Motherboard Point.

Sun may have Solaris for Itanium already ready

Discussion in 'Sun Hardware' started by Yousuf Khan, Oct 27, 2003.

  1. Yousuf Khan

    Yousuf Khan Guest

    1. Advertising

  2. Yousuf Khan

    Rich Teer Guest

    Rich Teer, Oct 27, 2003
    #2
    1. Advertising

  3. Yousuf Khan

    Yousuf Khan Guest

    "Rich Teer" <> wrote in message
    news:pine.SOL.4.58.0310271419460.467@zaphod...
    > On Mon, 27 Oct 2003, Yousuf Khan wrote:
    >
    > > http://www.xbitlabs.com/news/other/display/20031027123749.html

    >
    > That's old news; Solaris was one of the first OSes to be ported
    > to Itanic, a couple of years ago.


    It was one of the first OSes to be cancelled for Itanic a couple of years
    ago! Sun and Intel had a "spat", and Sun cancelled Solaris, complaining
    Intel wasn't giving them enough help. It looks like Sun continued to develop
    it without Intel's help.

    Yousuf Khan
     
    Yousuf Khan, Oct 27, 2003
    #3
  4. "Yousuf Khan" <> writes in comp.unix.solaris:
    |http://www.xbitlabs.com/news/other/display/20031027123749.html

    Wow. That's a very clueless report. Solaris for Itanium was incredibly
    public years ago, along with the reasons it was cancelled. A google
    search could give you better information than that.

    --
    ________________________________________________________________________
    Alan Coopersmith
    http://www.CSUA.Berkeley.EDU/~alanc/ aka:
    Working for, but definitely not speaking for, Sun Microsystems, Inc.
     
    Alan Coopersmith, Oct 27, 2003
    #4
  5. Yousuf Khan

    Yousuf Khan Guest

    "Alan Coopersmith" <> wrote in message
    news:bnk8ld$2dr6$...
    > "Yousuf Khan" <> writes in

    comp.unix.solaris:
    > |http://www.xbitlabs.com/news/other/display/20031027123749.html
    >
    > Wow. That's a very clueless report. Solaris for Itanium was incredibly
    > public years ago, along with the reasons it was cancelled. A google
    > search could give you better information than that.


    I think what they are trying to say is that they completed the port two
    years ago, despite cancelling it.

    Yousuf Khan
     
    Yousuf Khan, Oct 27, 2003
    #5
  6. Yousuf Khan

    Rob Stow Guest

    Rich Teer wrote:
    > On Mon, 27 Oct 2003, Yousuf Khan wrote:
    >
    >
    >>http://www.xbitlabs.com/news/other/display/20031027123749.html

    >
    >
    > That's old news; Solaris was one of the first OSes to be ported
    > to Itanic, a couple of years ago.
    >


    I'm sure that in hindsight Sun wishes it had never spent a dime
    on that project. Even if Sun and Intel weren't fighting over
    patent/licensing issues, Sun surely would have wished to avoid
    spending so much money developing for a processor that so far
    seems destined to never acquire any significant market share.

    Itanic is a dead horse that Intel and HP continue to flog -
    everyone else seems to be putting their whips away.
     
    Rob Stow, Oct 28, 2003
    #6
  7. Yousuf Khan

    Nate Edel Guest

    In comp.arch Rob Stow <> wrote:
    > Itanic is a dead horse that Intel and HP continue to flog -
    > everyone else seems to be putting their whips away.


    Nah, it's reasonably successful as "the latest version of PA-RISC," and
    seems to be successful as far as it goes in that role.

    Of course, it still hasn't replaced HPaq's Alpha machines, but I suspect
    that if they can get VMS/Itanic stable, it will successfully replace the
    Alphas (at least in that role... is anyone still buying them as Unix
    machines, or have they already killed them for that?)

    --
    Nate Edel http://www.nkedel.com/

    "But Marge! I've never felt so accepted in my life. These people looked deep
    into my soul and assigned me a number based on the order in which I joined."
     
    Nate Edel, Oct 28, 2003
    #7
  8. =?ISO-8859-1?Q?Peter_Perls=F8?=, Oct 28, 2003
    #8
  9. In article <jBhnb.96157$>,
    Yousuf Khan wrote:
    >"Rich Teer" <> wrote in message
    >news:pine.SOL.4.58.0310271419460.467@zaphod...
    >> On Mon, 27 Oct 2003, Yousuf Khan wrote:


    >> > http://www.xbitlabs.com/news/other/display/20031027123749.html


    >> That's old news; Solaris was one of the first OSes to be ported
    >> to Itanic, a couple of years ago.


    >It was one of the first OSes to be cancelled for Itanic a couple of years
    >ago! Sun and Intel had a "spat", and Sun cancelled Solaris, complaining
    >Intel wasn't giving them enough help. It looks like Sun continued to develop
    >it without Intel's help.


    Sun have probably "continued to develop it" in the sense that most of
    the current Solaris development work can easily be applied to the Itanium
    code tree. The amount of platform dependant code in Solaris is quite small
    so you could probably argue that they also "continue to develop" the Power
    PC version of Solaris they had some years ago.

    --
    Phillip Fayers School of Physics & Astronomy, Cardiff University (*).
    [ (*) - the official trading name of the University of Wales, Cardiff. ]
    http://www.astro.cf.ac.uk/pub/Phillip.Fayers/
    Tel: +44 (0)29 2087 5282 Attribute these comments to me, not UWC.
     
    Phillip Fayers, Oct 28, 2003
    #9
  10. Yousuf Khan

    Greg Menke Guest

    Peter Perlsø <> writes:

    > Yousuf Khan wrote:
    > > http://www.xbitlabs.com/news/other/display/20031027123749.html
    > >

    >
    > Big deal. Intanium sales are nil, and people are realizing that they
    > can get better functionality for far less the price of a Sun, from a
    > Linux box build from off the shelf hardware.


    Theres more to functionality than plopping down minimum
    dollars/gigabyte of fileserver. No doubt Linux makes for a cheaper
    and sometimes faster fileserver than Sun does- but price is not the
    only metric.

    Gregm
     
    Greg Menke, Oct 28, 2003
    #10
  11. Yousuf Khan

    Gavin Maltby Guest

    Peter Perlsø wrote:
    > Yousuf Khan wrote:
    >
    >> http://www.xbitlabs.com/news/other/display/20031027123749.html
    >>
    >>

    >
    > Big deal. Intanium sales are nil, and people are realizing that they can
    > get better functionality for far less the price of a Sun, from a Linux
    > box build from off the shelf hardware.


    This is such an ignorant statement, heard (sadly) all too often from both
    the clueless and many who should know better :-( While nobody denies
    that such a solution can be better or equivalent or at least
    more cost-efficient in some cases, such glib dismissals only show
    complete ignorance of all the hardware and software involved
    (ie all of Sun SPARC systems amd PCs, and of Solaris and Linux).

    Gavin
     
    Gavin Maltby, Oct 28, 2003
    #11
  12. Yousuf Khan

    Ade Guest

    Greg Menke wrote:
    > Peter Perlsø <> writes:
    >
    >
    >>Yousuf Khan wrote:
    >>
    >>>http://www.xbitlabs.com/news/other/display/20031027123749.html
    >>>

    >>
    >>Big deal. Intanium sales are nil, and people are realizing that they
    >>can get better functionality for far less the price of a Sun, from a
    >>Linux box build from off the shelf hardware.

    >
    >
    > Theres more to functionality than plopping down minimum
    > dollars/gigabyte of fileserver. No doubt Linux makes for a cheaper
    > and sometimes faster fileserver than Sun does- but price is not the
    > only metric.
    >
    > Gregm
    >



    Here here that man! About time people realised that.
     
    Ade, Oct 28, 2003
    #12
  13. Yousuf Khan

    Gavin Maltby Guest

    Peter Perlsø wrote:
    > Yousuf Khan wrote:
    >
    >> http://www.xbitlabs.com/news/other/display/20031027123749.html
    >>
    >>

    >
    > Big deal. Intanium sales are nil, and people are realizing that they can
    > get better functionality for far less the price of a Sun, from a Linux
    > box build from off the shelf hardware.
    >


    The important word in that sentence is "can", that it is not "will".
    Yes your Linux on off-the-shelf hardware _can_ offer better
    functionality and or lower price etc. It's a sad fact, but
    many an armchair hardware, operating system and application software
    "expert" makes these decisions with the same depth of thought and
    understanding displayed in the above statement. One day, perhaps,
    the industry will mature and stabilize enough that this will be
    sufficient (there's a Sun N1 pitch in there somewhere!). But in
    the interim any respectable business decision to choose a particular
    solution should be made after weighing many factors, not just
    apparent functionality and price. Obviously the weights differ
    depending on the nature of the business, and in some cases
    some factors can be ignored. But to work off the assumption
    that "an OS is an OS", "hardware is hardware" etc is just
    plain stupid.

    Of course, much of the above becomes irrelevant for home systems
    for personal use. Budgets there typically make price the main
    consideration. You likely don't have a billion dollar database
    (the data integrity of which you'd like assured as much as the
    industry allows) running on it. Linux/PC is cool, open source, and
    free so an obvious strong candidate for home. But when you weigh
    the combination as a business proposal you need to consider all
    aspects of it, and there are many aspects there where it is not
    top (and not even second).

    Gavin

    (Speaking for myself)
     
    Gavin Maltby, Oct 28, 2003
    #13
  14. Gavin Maltby wrote:
    > Peter Perlsø wrote:
    >
    >> Yousuf Khan wrote:
    >>
    >>> http://www.xbitlabs.com/news/other/display/20031027123749.html
    >>>
    >>>

    >>
    >> Big deal. Intanium sales are nil, and people are realizing that they
    >> can get better functionality for far less the price of a Sun, from a
    >> Linux box build from off the shelf hardware.
    >>

    >
    > The important word in that sentence is "can", that it is not "will".
    > Yes your Linux on off-the-shelf hardware _can_ offer better
    > functionality and or lower price etc. It's a sad fact, but
    > many an armchair hardware, operating system and application software
    > "expert" makes these decisions with the same depth of thought and
    > understanding displayed in the above statement.
    >


    So, let's hear it - give me a few good reasons to shell out, say 1,5 mil
    on a Sun solution, compared to the tenth of than on, say, Red Hat
    Enterprise on a custom built Xeon system.
     
    =?ISO-8859-1?Q?Peter_Perls=F8?=, Oct 28, 2003
    #14
  15. Yousuf Khan

    Rich Teer Guest

    On Wed, 29 Oct 2003, Peter Perlsø wrote:

    > So, let's hear it - give me a few good reasons to shell out, say 1,5 mil
    > on a Sun solution, compared to the tenth of than on, say, Red Hat
    > Enterprise on a custom built Xeon system.


    Oh come on, this one's easy: Xeon systems don't scale to
    anywhere near the number of processors that a big (or even
    medium) Sun system can. Some problems DO require horizontally
    scaled solutions; not everything can be done on little bitty
    2-way x86 servers, you know. And if you think otherwise,
    you are more than a little naive.

    And exactly how will that toy Xeon system keep, say, a
    100 GB database completely in RAM?

    I submit that Red Hat Enterprise is about as Enterprise
    ready as the Dull x86 hardware it runs on...

    --
    Rich Teer, SCNA, SCSA

    President,
    Rite Online Inc.

    Voice: +1 (250) 979-1638
    URL: http://www.rite-online.net
     
    Rich Teer, Oct 28, 2003
    #15
  16. Peter Perlsø wrote:

    > Gavin Maltby wrote:
    >> Peter Perlsø wrote:
    >>
    >>> Yousuf Khan wrote:
    >>>
    >>>> http://www.xbitlabs.com/news/other/display/20031027123749.html
    >>>>
    >>>>
    >>>
    >>> Big deal. Intanium sales are nil, and people are realizing that they
    >>> can get better functionality for far less the price of a Sun, from a
    >>> Linux box build from off the shelf hardware.
    >>>

    >>
    >> The important word in that sentence is "can", that it is not "will".
    >> Yes your Linux on off-the-shelf hardware _can_ offer better
    >> functionality and or lower price etc. It's a sad fact, but
    >> many an armchair hardware, operating system and application software
    >> "expert" makes these decisions with the same depth of thought and
    >> understanding displayed in the above statement.
    >>

    >
    > So, let's hear it - give me a few good reasons to shell out, say 1,5 mil
    > on a Sun solution, compared to the tenth of than on, say, Red Hat
    > Enterprise on a custom built Xeon system.


    Very much depends on your application.

    If you are already predominantly Sun and it is going to take you a
    substantial amount of development time (say 1500 man days) to port to Linux
    then Sun is a good choice.

    If you like unexplained hardware faults that the vendor forces you to NDA
    for then Sun is a good choice.

    If you want limited scalability, or at least scalability by box count rather
    than system size then Intel is a good call.

    There are relatively few environments that actually need to cram a 100GB
    database into RAM, but where they exist then Sparc is a better bet than
    x86.

    If you want extreme scalability then by all means go Sparc, but from FSC,
    not Sun. The FSC boxes are currently up to 30% larger than the equivalent
    Sun and their Sparc implementation tends to show out as around 10-15%
    faster. All this for around 1/3-1/2 the investment across a 3 year period.

    Most big purchases of Sun kit tend to go that way for one of a number of
    reasons:

    1) The enterprise is already predominantly Sun and it makes sense to
    centralise on a single vendor.

    2) The enterprise doesn't know what it's doing and slavishly follow the "NT
    Bad, UNIX good" mantra and buy Sun because "that's what everybody buys"

    3) There is kickback to be had

    4) Sun do something silly with their pricing to get the deal (seen it
    happen)

    5) Sun is technically the best solution.

    Of all of the above 1 and 2 are the most common, 5 is by far and away the
    least common.

    The big thing Sun have going for them is Solaris, and Solaris on Sparc at
    that. They have ISV buyin.

    At the point where *any* other commercial UNIX has that level of ISV support
    Sun suddenly start looking less appealing.

    Most purchasing decisions in enterprises who need the mega-high-end
    scalability espoused by Rich Teer and his ilk tend to go to inertia, as
    these enterprises have generally been round a while and at the time they
    started buying UNIX Sun was the best on the block and they haven't
    re-evaluated their policies since then.

    P.
     
    Paul S. Brown, Oct 29, 2003
    #16
  17. On Tue, 28 Oct 2003 23:28:40 +0000, Rich Teer wrote:

    > On Wed, 29 Oct 2003, Peter Perlsø wrote:
    >
    >> So, let's hear it - give me a few good reasons to shell out, say 1,5 mil
    >> on a Sun solution, compared to the tenth of than on, say, Red Hat
    >> Enterprise on a custom built Xeon system.

    >
    > Oh come on, this one's easy: Xeon systems don't scale to
    > anywhere near the number of processors that a big (or even
    > medium) Sun system can.


    True.

    > And exactly how will that toy Xeon system keep, say, a
    > 100 GB database completely in RAM?


    Could be a problem. But then again, I suspect there will
    be Opteron Servers that can do this very soon.

    > I submit that Red Hat Enterprise is about as Enterprise
    > ready as the Dull x86 hardware it runs on...


    Can you back this assertion up in any way?

    I submit that Sun will have a difficult time surviving another
    5 years. 64-bit x86 and Linux are going to destroy it.
    It's not that I want Sun to die, I just don't see any way out.
     
    Sheldon Simms, Oct 29, 2003
    #17
  18. Yousuf Khan

    Rich Teer Guest

    On Tue, 28 Oct 2003, Sheldon Simms wrote:

    > Could be a problem. But then again, I suspect there will
    > be Opteron Servers that can do this very soon.


    we'll see...

    > Can you back this assertion up in any way?


    Not with any hard data; but my (limited) Linux experience
    suggests that it is not ready for enterprise deployment.
    I've heard (from sources I trust) of Linux machines panicing
    under high load (perhaps due to running out of swap); that,
    in my book, is totally unacceptable for an enterprise environment.

    The Gates-spawned home computer user, who is used to crashes
    and reboots for no apparently good reason, might think otherwise,
    however...

    > I submit that Sun will have a difficult time surviving another
    > 5 years. 64-bit x86 and Linux are going to destroy it.
    > It's not that I want Sun to die, I just don't see any way out.


    If x86 (in the form of AMD's 64-bit stuff) is indeed the
    way forward, I can think an OS that is much more suited
    to this endeavor than Linux, and that is Solaris x86.

    --
    Rich Teer, SCNA, SCSA

    President,
    Rite Online Inc.

    Voice: +1 (250) 979-1638
    URL: http://www.rite-online.net
     
    Rich Teer, Oct 29, 2003
    #18
  19. On Wed, 29 Oct 2003 01:59:48 +0000, Rich Teer wrote:

    > On Tue, 28 Oct 2003, Sheldon Simms wrote:
    >
    >> Could be a problem. But then again, I suspect there will
    >> be Opteron Servers that can do this very soon.

    >
    > we'll see...
    >
    >> Can you back this assertion up in any way?

    >
    > Not with any hard data; but my (limited) Linux experience
    > suggests that it is not ready for enterprise deployment.
    > I've heard (from sources I trust) of Linux machines panicing
    > under high load (perhaps due to running out of swap); that,
    > in my book, is totally unacceptable for an enterprise environment.


    Obviously. But Linux is improving all the time. Also,
    I feel quite confident there are differences in behavior
    between (for example) Red Hat Advanced Server and the
    typical self-compiled kernel. The compilation environment,
    vendor patches, performance testing, etc. all have an
    influence.

    > The Gates-spawned home computer user, who is used to crashes
    > and reboots for no apparently good reason, might think otherwise,
    > however...
    >
    >> I submit that Sun will have a difficult time surviving another
    >> 5 years. 64-bit x86 and Linux are going to destroy it.
    >> It's not that I want Sun to die, I just don't see any way out.

    >
    > If x86 (in the form of AMD's 64-bit stuff) is indeed the
    > way forward, I can think an OS that is much more suited
    > to this endeavor than Linux, and that is Solaris x86.


    What does Sun do when they can't sell hardware with a decent
    profit margin anymore? When Sparc is history? If they try to
    get the money back with Solaris they won't be able to compete
    with Linux.
     
    Sheldon Simms, Oct 29, 2003
    #19
  20. Yousuf Khan

    CJT Guest

    Sheldon Simms wrote:

    > On Wed, 29 Oct 2003 01:59:48 +0000, Rich Teer wrote:
    >
    >
    >>On Tue, 28 Oct 2003, Sheldon Simms wrote:
    >>
    >>
    >>>Could be a problem. But then again, I suspect there will
    >>>be Opteron Servers that can do this very soon.

    >>
    >>we'll see...
    >>
    >>
    >>>Can you back this assertion up in any way?

    >>
    >>Not with any hard data; but my (limited) Linux experience
    >>suggests that it is not ready for enterprise deployment.
    >>I've heard (from sources I trust) of Linux machines panicing
    >>under high load (perhaps due to running out of swap); that,
    >>in my book, is totally unacceptable for an enterprise environment.

    >
    >
    > Obviously. But Linux is improving all the time. <snip>


    .... and so is Solaris.


    --
    After being targeted with gigabytes of trash by the "SWEN" worm, I have
    concluded we must conceal our e-mail address. Our true address is the
    mirror image of what you see before the "@" symbol. It's a shame such
    steps are necessary. ...Charlie
     
    CJT, Oct 29, 2003
    #20
    1. Advertising

Want to reply to this thread or ask your own question?

It takes just 2 minutes to sign up (and it's free!). Just click the sign up button to choose a username and then you can ask your own questions on the forum.
Similar Threads
  1. lyon_wonder
    Replies:
    1
    Views:
    266
    Jim-Ed Browne
    Apr 27, 2004
  2. Yousuf Khan
    Replies:
    107
    Views:
    1,916
  3. Matt Simis
    Replies:
    5
    Views:
    530
    Matt Simis
    Dec 20, 2003
  4. Yousuf Khan

    OpenVMS on Itanium almost ready

    Yousuf Khan, Dec 21, 2003, in forum: Intel
    Replies:
    1
    Views:
    334
    Bill Todd
    Dec 24, 2003
  5. BORGWARD

    D:\ is not ready, Device not ready

    BORGWARD, Sep 15, 2004, in forum: Hardware
    Replies:
    1
    Views:
    373
    Jim Phelps
    Sep 15, 2004
Loading...

Share This Page