Previewing AIX 6 and open beta program - Aix

This is a discussion on Previewing AIX 6 and open beta program - Aix ; http://www-03.ibm.com/servers/aix/6/preview.html...

+ Reply to Thread
Results 1 to 16 of 16

Thread: Previewing AIX 6 and open beta program

  1. Previewing AIX 6 and open beta program


  2. Re: Previewing AIX 6 and open beta program

    On 2007-05-21, Dieter Mosbach wrote:
    > http://www-03.ibm.com/servers/aix/6/preview.html


    "- Dynamic tracing
    AIX 6 provides a new dynamic tracing capability that can simplify
    debugging complex system or application code. This dynamic tracing
    facility will be introduced via a new tracing command, probevue, that
    allows a developer or system administrator to dynamically insert trace
    breakpoints in existing code without having to recompile the code."

    Sounds something like Solaris 10s DTrace. Does anyone know if
    it's going to be as powerful as DTrace, or is it just a "damn,
    we should have this one too"-quickshot?

    "Participants can install the open beta version of AIX 6 on IBM systems
    that run on the POWER4?, PPC970, POWER5? or POWER6? processors including
    the IBM System p?, IBM System p5?, IBM eServer? p5, IBM eServer pSeries®
    server product lines, as well as IBM BladeCenter® JS2x blades and
    IntelliStation® POWER? workstations."

    Damn! Someone got an affordable Power4 system to sell ;-)

    What i'd *really* wish for is a replacement of the SRC at
    least as good and flexible as the SMF in Solaris 10. Don't
    get me wrong, i still like AIX very much, but am a bit spioled
    by the shiny new features in the Solaris development ;-)

    Regards,

    Frank

  3. Re: Previewing AIX 6 and open beta program

    Frank Fegert wrote:
    > On 2007-05-21, Dieter Mosbach wrote:
    >> http://www-03.ibm.com/servers/aix/6/preview.html

    >
    > "Participants can install the open beta version of AIX 6 on IBM systems
    > that run on the POWER4?, PPC970, POWER5? or POWER6? processors including
    > the IBM System p?, IBM System p5?, IBM eServer? p5, IBM eServer pSeries®
    > server product lines, as well as IBM BladeCenter® JS2x blades and
    > IntelliStation® POWER? workstations."
    >
    > Damn! Someone got an affordable Power4 system to sell ;-)


    I'm going to try the beta on my POWER3-II 7044-270. I can't imagine a
    technical reason they dropped POWER3 support (64bit), so there is a
    chance it could work.

    Kevin

  4. Re: Previewing AIX 6 and open beta program

    Kevin Bowling wrote:

    > Frank Fegert wrote:
    >
    >>On 2007-05-21, Dieter Mosbach wrote:
    >>
    >>>http://www-03.ibm.com/servers/aix/6/preview.html

    >>
    >> "Participants can install the open beta version of AIX 6 on IBM systems
    >> that run on the POWER4?, PPC970, POWER5? or POWER6? processors including
    >> the IBM System p?, IBM System p5?, IBM eServer? p5, IBM eServer pSeries®
    >> server product lines, as well as IBM BladeCenter® JS2x blades and
    >> IntelliStation® POWER? workstations."
    >>
    >>Damn! Someone got an affordable Power4 system to sell ;-)

    >
    >
    > I'm going to try the beta on my POWER3-II 7044-270. I can't imagine a
    > technical reason they dropped POWER3 support (64bit), so there is a
    > chance it could work.
    >
    > Kevin


    Just speculation on my part, but:

    One possible technical reason to drop older processors
    is so that the new AIX 6 kernel and runtime can be compiled
    with optimization for the later processors and therefore
    will hopefully run faster.

    Of course there are several business reasons to drop
    the older processors, one being to reduce the cost of
    testing the s/w before releasing it.

    Paul Landay

  5. Re: Previewing AIX 6 and open beta program

    Paul Landay wrote:
    > Kevin Bowling wrote:
    >>> On 2007-05-21, Dieter Mosbach wrote:
    >>> Damn! Someone got an affordable Power4 system to sell ;-)

    >>
    >>
    >> I'm going to try the beta on my POWER3-II 7044-270. I can't imagine a
    >> technical reason they dropped POWER3 support (64bit), so there is a
    >> chance it could work.
    >>
    >> Kevin

    >
    > Just speculation on my part, but:
    >
    > One possible technical reason to drop older processors
    > is so that the new AIX 6 kernel and runtime can be compiled
    > with optimization for the later processors and therefore
    > will hopefully run faster.
    >
    > Of course there are several business reasons to drop
    > the older processors, one being to reduce the cost of
    > testing the s/w before releasing it.
    >
    > Paul Landay



    I can't see this as much of a business decision, and I hope someone at
    IBM hears me. The fact remains that POWER4 and up systems are
    significantly more expensive. This means that admins like myself will
    have a much harder time getting a hold of AIX6 capable gear for
    practice, and it will ripple down line. Less AIX admins means higher
    cost, means less AIX users, means companies less willing to deploy, etc
    even more so than now.

    Even if it does not make business sense to open source AIX or do an
    amd64 port now[1], there is no doubt that open standards have increased
    the worth and use of AIX. I think these "open practices" are excellent,
    and IBM should expand the AIX community and get it involved as much as
    possible, similar to the power.org effort.

    [1] ..Which is too bad. If Sun sun can do it, IBM can too. Fact is
    System P are the best midrange to highend systems on the market and sell
    because of this fact, not because of AIX. AIX exists in a state of
    limbo in my eyes. No doubt it is a powerful and stable OS, but it would
    be nice to see it expand and thrive.


  6. Re: Previewing AIX 6 and open beta program

    On May 23, 3:31 pm, Kevin Bowling
    wrote:
    > Paul Landay wrote:
    > > Kevin Bowling wrote:
    > >>> On 2007-05-21, Dieter Mosbach wrote:
    > >>> Damn! Someone got an affordable Power4 system to sell ;-)

    >
    > >> I'm going to try the beta on my POWER3-II 7044-270. I can't imagine a
    > >> technical reason they dropped POWER3 support (64bit), so there is a
    > >> chance it could work.

    >
    > >> Kevin

    >
    > > Just speculation on my part, but:

    >
    > > One possible technical reason to drop older processors
    > > is so that the new AIX 6 kernel and runtime can be compiled
    > > with optimization for the later processors and therefore
    > > will hopefully run faster.

    >
    > > Of course there are several business reasons to drop
    > > the older processors, one being to reduce the cost of
    > > testing the s/w before releasing it.

    >
    > > Paul Landay

    >
    >
    > I can't see this as much of a business decision, and I hope someone at
    > IBM hears me. The fact remains that POWER4 and up systems are
    > significantly more expensive. This means that admins like myself will
    > have a much harder time getting a hold of AIX6 capable gear for
    > practice, and it will ripple down line. Less AIX admins means higher
    > cost, means less AIX users, means companies less willing to deploy, etc
    > even more so than now.
    >
    > Even if it does not make business sense to open source AIX or do an
    > amd64 port now[1], there is no doubt that open standards have increased
    > the worth and use of AIX. I think these "open practices" are excellent,
    > and IBM should expand the AIX community and get it involved as much as
    > possible, similar to the power.org effort.
    >
    > [1] ..Which is too bad. If Sun sun can do it, IBM can too. Fact is
    > System P are the best midrange to highend systems on the market and sell
    > because of this fact, not because of AIX. AIX exists in a state of
    > limbo in my eyes. No doubt it is a powerful and stable OS, but it would
    > be nice to see it expand and thrive.
    >


    Agreed completely... I'm swimming in old 7025-F80s and p620-6F1s. I'll
    probably try it on one of them, but I doubt that the RS64-III will be
    beefy enough, not to mention how well crafted the firmwares are to
    only allow certain things to boot.


  7. Re: Previewing AIX 6 and open beta program

    Kevin Bowling wrote:
    > significantly more expensive. This means that admins like myself will
    > have a much harder time getting a hold of AIX6 capable gear for
    > practice, and it will ripple down line. Less AIX admins means higher
    > cost, means less AIX users, means companies less willing to deploy, etc
    > even more so than now.


    agree on professional and personal levels (as i have a 43p @ home ..
    at the same time the virt. features of the power4/power5 systems you do
    have also grant you the 'option' to whip a practice system out of thin
    air. i know it still consumes resources depending on the
    platform/vintage of stuff.. but we _can_use all this virtualization
    stuff to _our_ advantage as admins, too..

    -r (unless your company doesn't buy micropartitioning with that shiny
    new p570)
    (but that's a rant for a.s.r..

  8. Re: Previewing AIX 6 and open beta program


    >>I can't see this as much of a business decision, and I hope someone at
    >>IBM hears me. The fact remains that POWER4 and up systems are
    >>significantly more expensive. This means that admins like myself will
    >>have a much harder time getting a hold of AIX6 capable gear for
    >>practice, and it will ripple down line. Less AIX admins means higher
    >>cost, means less AIX users, means companies less willing to deploy, etc
    >>even more so than now.


    IMHO, If a company is willing to spend $$$$ on p5/p6 kit then *should*
    to have the sense to factor in training or techical handover from the
    pre/post sales bods. And lets face it, picking up a redbook or a
    whitepaper to learn new stuff is not rocket science and is generally
    considered part of an admins role nowadays. I dont think that it will
    ripple down to less AIX admins at the end of the day .. as a matter of
    fact, with the p5 sales being at the top of the Enterprise UNIX
    Hardware sales last year, we are going to end up with more admins as
    companies are buying IBM kit over HP and SUN - hoick phut

    My 2 cents anyhoo ..

    Rgds
    Mark Taylor




  9. Re: Previewing AIX 6 and open beta program

    Mark Taylor wrote:
    >
    >
    > >>I can't see this as much of a business decision, and I hope someone at
    > >>IBM hears me. The fact remains that POWER4 and up systems are
    > >>significantly more expensive. This means that admins like myself will
    > >>have a much harder time getting a hold of AIX6 capable gear for
    > >>practice, and it will ripple down line. Less AIX admins means higher
    > >>cost, means less AIX users, means companies less willing to deploy, etc
    > >>even more so than now.

    >
    > IMHO, If a company is willing to spend $$$$ on p5/p6 kit then *should*
    > to have the sense to factor in training or techical handover from the
    > pre/post sales bods. And lets face it, picking up a redbook or a
    > whitepaper to learn new stuff is not rocket science and is generally
    > considered part of an admins role nowadays. I dont think that it will
    > ripple down to less AIX admins at the end of the day .. as a matter of
    > fact, with the p5 sales being at the top of the Enterprise UNIX
    > Hardware sales last year, we are going to end up with more admins as
    > companies are buying IBM kit over HP and SUN - hoick phut


    Well, for what it's worth, the healthcare and financial sectors will soon be
    abandoning OpenVMS in droves due to a lack of suitable follow-on hardware to the
    Enterprise-scale Alpha kit that just went end-of-sale. (I64 SuperDomes have
    serious have serious memory latency issues in large RAM, multi-CPU
    configurations.)

    The number of AIX sites/Admins should be the least of anyone's worries.

    Now, security, on the other hand, ... well, let's face it: was AIX ever banned
    from hacker conventions for being too secure? (OpenVMS was!)

    --
    David J Dachtera
    dba DJE Systems
    http://www.djesys.com/

    Unofficial OpenVMS Marketing Home Page
    http://www.djesys.com/vms/market/

    Unofficial Affordable OpenVMS Home Page:
    http://www.djesys.com/vms/soho/

    Unofficial OpenVMS-IA32 Home Page:
    http://www.djesys.com/vms/ia32/

    Unofficial OpenVMS Hobbyist Support Page:
    http://www.djesys.com/vms/support/

  10. Re: Previewing AIX 6 and open beta program

    David J Dachtera wrote:

    [snip]

    >
    > Well, for what it's worth, the healthcare and financial sectors will soon be
    > abandoning OpenVMS in droves due to a lack of suitable follow-on hardware to the
    > Enterprise-scale Alpha kit that just went end-of-sale. (I64 SuperDomes have
    > serious have serious memory latency issues in large RAM, multi-CPU
    > configurations.)
    >


    Its also a problem the p5/p5+ 570s and 595s have. And according to the
    numbers from IBM, its worse than the Superdomes.

    --
    Dave Harrold,
    Lead Software Systems Engineer
    Aurora Health Care
    3031 W. Montana Street
    Milwaukee, WI 53215

    Phone: (414) 647-6204
    FAX: (414) 647-4999
    Email: David.Harrold@aurora.org
    http://www.AuroraHealthCare.org

  11. Re: Previewing AIX 6 and open beta program

    On 29 Mai, 19:43, Dave Harrold wrote:
    ....
    > Its also a problem the p5/p5+ 570s and 595s have. And according to the
    > numbers from IBM, its worse than the Superdomes.


    Hi Dave,
    could supply more information or links regarding the memory issue on
    p5/p5+ machines ?

    tia
    Hajo


  12. Re: Previewing AIX 6 and open beta program

    Hajo Ehlers wrote:
    > On 29 Mai, 19:43, Dave Harrold wrote:
    > ...
    >> Its also a problem the p5/p5+ 570s and 595s have. And according to the
    >> numbers from IBM, its worse than the Superdomes.

    >
    > Hi Dave,
    > could supply more information or links regarding the memory issue on
    > p5/p5+ machines ?
    >


    Well, it a problem if it effects your work load. The issue is memory
    latency of the system. According to a presentation I found, it lists a
    p5 595 as having a latency in the 210ns range.

    Name of the presentation is:

    POWER5 Processor and System Evolution
    ScicomP 11
    Charles Grassl
    IBM
    May, 2005

    > tia
    > Hajo
    >


    The only way to know if this effects you is to test your workload on the
    system in question.


    --
    Dave Harrold,
    Lead Software Systems Engineer
    Aurora Health Care
    3031 W. Montana Street
    Milwaukee, WI 53215

    Phone: (414) 647-6204
    FAX: (414) 647-4999
    Email: David.Harrold@aurora.org
    http://www.AuroraHealthCare.org

  13. Re: Previewing AIX 6 and open beta program

    Dave Harrold wrote:
    >
    > Hajo Ehlers wrote:
    > > On 29 Mai, 19:43, Dave Harrold wrote:
    > > ...
    > >> Its also a problem the p5/p5+ 570s and 595s have. And according to the
    > >> numbers from IBM, its worse than the Superdomes.

    > >
    > > Hi Dave,
    > > could supply more information or links regarding the memory issue on
    > > p5/p5+ machines ?
    > >

    >
    > Well, it a problem if it effects your work load. The issue is memory
    > latency of the system. According to a presentation I found, it lists a
    > p5 595 as having a latency in the 210ns range.
    >
    > Name of the presentation is:
    >
    > POWER5 Processor and System Evolution
    > ScicomP 11
    > Charles Grassl
    > IBM
    > May, 2005
    >
    > The only way to know if this effects you is to test your workload on the
    > system in question.


    I entered "POWER5 Processor and System Evolution" into Yahoo! and found this
    link:

    http://www.cisl.ucar.edu/computers/blueice/

    ....on which page I found this link near the bottom, which appears to be a .PDF
    of the presentation Dave mentioned:

    http://www.spscicomp.org/ScicomP11/P...ial-power5.pdf

    It's an almost 9MB .PDF, so be advised.

    I'll see if I can re-find a link that was posted to the comp.os.vms newsgroup
    about memory access comparisons of Alpha GS1280, Integrity I64 and SuperDome
    I64 showing that the SuperDomes are almost 50% slower at memory access than the
    GS1280s, similarly configured.

    Maybe I can find someone who has done a similar comparison of Alpha, I64 and P5
    in Enterprise-Class configurations just to get it all on one graph.

    --
    David J Dachtera
    dba DJE Systems
    http://www.djesys.com/

    Unofficial OpenVMS Marketing Home Page
    http://www.djesys.com/vms/market/

    Unofficial Affordable OpenVMS Home Page:
    http://www.djesys.com/vms/soho/

    Unofficial OpenVMS-IA32 Home Page:
    http://www.djesys.com/vms/ia32/

    Unofficial OpenVMS Hobbyist Support Page:
    http://www.djesys.com/vms/support/

  14. Re: Previewing AIX 6 and open beta program

    David J Dachtera wrote:
    > I'll see if I can re-find a link that was posted to the comp.os.vms
    > newsgroup about memory access comparisons of Alpha GS1280, Integrity
    > I64 and SuperDome I64 showing that the SuperDomes are almost 50%
    > slower at memory access than the GS1280s, similarly configured.


    FWIW, there have been three separate CEC's for the systems HP calls
    "SuperDomes." The first was for the HP 9000 Superdome, with PA-RISC
    CPUs in it. That is, PA-RISC 8700 and earlier.

    Then there was the first of the CECs for an Integrity SuperDome, which
    was called "sx1000" externally. That is also used, IIRC for PA-8800
    and PA-8900 SuperDomes.

    The current CEC is called "sx2000" which is used for the Integrity
    SuperDome's with the Itanium2 9000's, aka Montecito, processors.

    Each has had differences in memory access latency and aggregate
    bandwidth, so it would be best to be specific about which CEC in any
    discussion.

    Similarly, there have been two separate CECs for the smaller Integrity
    systems. The first was zx1 (rx[12]6[02]0, rx4640, BL60p and some
    others) the current is zx2 (rx2660, rx3600, rx6600, BL860)

    sincerely,

    rick jones
    --
    Wisdom Teeth are impacted, people are affected by the effects of events.
    these opinions are mine, all mine; HP might not want them anyway...
    feel free to post, OR email to rick.jones2 in hp.com but NOT BOTH...

  15. Re: Previewing AIX 6 and open beta program

    Rick Jones wrote:
    >
    > David J Dachtera wrote:
    > > I'll see if I can re-find a link that was posted to the comp.os.vms
    > > newsgroup about memory access comparisons of Alpha GS1280, Integrity
    > > I64 and SuperDome I64 showing that the SuperDomes are almost 50%
    > > slower at memory access than the GS1280s, similarly configured.

    >
    > FWIW, there have been three separate CEC's


    %ELTRAN-E-NOCONTEXT, undefined term not supported by context
    -ELTRAN-E-UNDEF, undefined term
    \CEC\

    > for the systems HP calls
    > "SuperDomes." The first was for the HP 9000 Superdome, with PA-RISC
    > CPUs in it. That is, PA-RISC 8700 and earlier.
    >
    > Then there was the first of the CECs for an Integrity SuperDome, which
    > was called "sx1000" externally. That is also used, IIRC for PA-8800
    > and PA-8900 SuperDomes.
    >
    > The current CEC is called "sx2000" which is used for the Integrity
    > SuperDome's with the Itanium2 9000's, aka Montecito, processors.
    >
    > Each has had differences in memory access latency and aggregate
    > bandwidth, so it would be best to be specific about which CEC in any
    > discussion.


    Whichever model(s) were used in the tests used to compile the data and graphs
    which have been shown at the past two HP Technology Forum Symposia and numerous
    other sites around the world by then-current and now past HP / OpenVMS
    Engineering employees.

    --
    David J Dachtera
    dba DJE Systems
    http://www.djesys.com/

    Unofficial OpenVMS Marketing Home Page
    http://www.djesys.com/vms/market/

    Unofficial Affordable OpenVMS Home Page:
    http://www.djesys.com/vms/soho/

    Unofficial OpenVMS-IA32 Home Page:
    http://www.djesys.com/vms/ia32/

    Unofficial OpenVMS Hobbyist Support Page:
    http://www.djesys.com/vms/support/

  16. Re: Previewing AIX 6 and open beta program

    David J Dachtera wrote:
    > Rick Jones wrote:
    > > FWIW, there have been three separate CEC's


    > %ELTRAN-E-NOCONTEXT, undefined term not supported by context
    > -ELTRAN-E-UNDEF, undefined term
    > \CEC\


    The translation I use is Core Electronics Complex but there may be
    others - basically all the "other" chips used to allow the CPUs to do
    useful work.

    rick jones
    --
    The glass is neither half-empty nor half-full. The glass has a leak.
    The real question is "Can it be patched?"
    these opinions are mine, all mine; HP might not want them anyway...
    feel free to post, OR email to rick.jones2 in hp.com but NOT BOTH...

+ Reply to Thread