Alphaserver/station reliability - VMS

This is a discussion on Alphaserver/station reliability - VMS ; This is a little bit of an open request but for the installation and application I'm trying to justify it would really help me if I could get some 'real' uptime data from the wider industry on how reliable AlphaServers ...

+ Reply to Thread
Page 1 of 2 1 2 LastLast
Results 1 to 20 of 26

Thread: Alphaserver/station reliability

  1. Alphaserver/station reliability

    This is a little bit of an open request but for the installation and
    application I'm trying to justify it would really help me if I could
    get some 'real' uptime data from the wider industry on how reliable
    AlphaServers are. I know they have outstanding performance and are
    used by half of wall street etc. but it would be great to know from
    the real guys who have to keep these things running the real facts
    behind the claims.
    The two models I'm interested in are Alphaserver DS25 (a number of,
    both single and dual processor) and an Alphastation DS15. All will be
    running Tru64 Unix 5.1b.
    Simple feedback such as "Typical application, date installed, MTBF/
    uptime/failure rate etc., whether frequently power-cycled or running
    24/7" would be great. Direct contact details to confirm this would
    also be appreciated but not essential.
    The reason for this request is to support the use of the Alphas in a
    mission-critical role running a legacy app as part of the business
    justification.
    Posting to the forum preferred but mail to hp@pearson.in also accepted
    if confidentiality required.

  2. Re: Alphaserver/station reliability

    david.pearson@british-energy.com wrote:
    > This is a little bit of an open request but for the installation and
    > application I'm trying to justify it would really help me if I could
    > get some 'real' uptime data from the wider industry on how reliable
    > AlphaServers are. I know they have outstanding performance and are
    > used by half of wall street etc. but it would be great to know from
    > the real guys who have to keep these things running the real facts
    > behind the claims.
    > The two models I'm interested in are Alphaserver DS25 (a number of,
    > both single and dual processor) and an Alphastation DS15. All will be
    > running Tru64 Unix 5.1b.
    > Simple feedback such as "Typical application, date installed, MTBF/
    > uptime/failure rate etc., whether frequently power-cycled or running
    > 24/7" would be great. Direct contact details to confirm this would
    > also be appreciated but not essential.
    > The reason for this request is to support the use of the Alphas in a
    > mission-critical role running a legacy app as part of the business
    > justification.
    > Posting to the forum preferred but mail to hp@pearson.in also accepted
    > if confidentiality required.


    My current uptime on my Alphastation 200 is 97 days. It's protected by
    a UPS but the run time after the power goes off is eight to ten minutes.
    Power, once off, takes two-three hours to restore. That's two to three
    hours to get two men and a truck here and two to three minutes to get a
    a man to the top of the pole in the "cherry picker" and replace the
    blown fuse.

    With reliable power and air conditioning, uptimes of four hundred days
    or more are quite possible. Your downtimes then are for reboots to
    install O/S upgrades and/or patches and to change SYSGEN parameters.
    Also allow for a shutdown every three or four years for major
    maintenance on the electrical systems.

    You CAN set things up to eliminate even more of the occasional down time
    but it can get REALLY expensive. You would need dual Uninterruptable
    Power Systems, dual (at least) air conditioners, dual emergency
    generators for when the power goes off for several hours, etc. Few
    companies need the uptime badly enough to pay the bills!


  3. Re: Alphaserver/station reliability

    In article <9e34ca46-b248-4099-b4ce-30d5e5959c09@m3g2000hsc.googlegroups.com>, david.pearson@british-energy.com writes:
    >The two models I'm interested in are Alphaserver DS25 (a number of,
    >both single and dual processor) and an Alphastation DS15. All will be
    >running Tru64 Unix 5.1b.


    This is somewhat the wrong group. I could tell you something about the
    stability of our DS25 under OpenVMS.

    Regards,
    Christoph Gartmann

    --
    Max-Planck-Institut fuer Phone : +49-761-5108-464 Fax: -452
    Immunbiologie
    Postfach 1169 Internet: gartmann@immunbio dot mpg dot de
    D-79011 Freiburg, Germany
    http://www.immunbio.mpg.de/home/menue.html

  4. Re: Alphaserver/station reliability

    In article <47ECE96F.7010504@comcast.net>, "Richard B. Gilbert" writes:
    >david.pearson@british-energy.com wrote:
    >> This is a little bit of an open request but for the installation and
    >> application I'm trying to justify it would really help me if I could
    >> get some 'real' uptime data from the wider industry on how reliable
    >> AlphaServers are. I know they have outstanding performance and are
    >> used by half of wall street etc. but it would be great to know from
    >> the real guys who have to keep these things running the real facts
    >> behind the claims.
    >> The two models I'm interested in are Alphaserver DS25 (a number of,
    >> both single and dual processor) and an Alphastation DS15. All will be
    >> running Tru64 Unix 5.1b.
    >> Simple feedback such as "Typical application, date installed, MTBF/
    >> uptime/failure rate etc., whether frequently power-cycled or running
    >> 24/7" would be great. Direct contact details to confirm this would
    >> also be appreciated but not essential.
    >> The reason for this request is to support the use of the Alphas in a
    >> mission-critical role running a legacy app as part of the business
    >> justification.
    >> Posting to the forum preferred but mail to hp@pearson.in also accepted
    >> if confidentiality required.

    >
    >My current uptime on my Alphastation 200 is 97 days. It's protected by
    >a UPS but the run time after the power goes off is eight to ten minutes.
    >Power, once off, takes two-three hours to restore. That's two to three
    >hours to get two men and a truck here and two to three minutes to get a
    >a man to the top of the pole in the "cherry picker" and replace the
    >blown fuse.
    >
    >With reliable power and air conditioning, uptimes of four hundred days
    >or more are quite possible. Your downtimes then are for reboots to
    >install O/S upgrades and/or patches and to change SYSGEN parameters.
    >Also allow for a shutdown every three or four years for major
    >maintenance on the electrical systems.


    Did you notice that he asked for a DS25 under Unix? I assume yours is
    running under OpenVMS ;-)

    Regards,
    Christoph Gartmann

    --
    Max-Planck-Institut fuer Phone : +49-761-5108-464 Fax: -452
    Immunbiologie
    Postfach 1169 Internet: gartmann@immunbio dot mpg dot de
    D-79011 Freiburg, Germany
    http://www.immunbio.mpg.de/home/menue.html

  5. Re: Alphaserver/station reliability

    Christoph Gartmann wrote:
    > In article <47ECE96F.7010504@comcast.net>, "Richard B. Gilbert" writes:
    >
    >>david.pearson@british-energy.com wrote:
    >>
    >>>This is a little bit of an open request but for the installation and
    >>>application I'm trying to justify it would really help me if I could
    >>>get some 'real' uptime data from the wider industry on how reliable
    >>>AlphaServers are. I know they have outstanding performance and are
    >>>used by half of wall street etc. but it would be great to know from
    >>>the real guys who have to keep these things running the real facts
    >>>behind the claims.
    >>>The two models I'm interested in are Alphaserver DS25 (a number of,
    >>>both single and dual processor) and an Alphastation DS15. All will be
    >>>running Tru64 Unix 5.1b.
    >>>Simple feedback such as "Typical application, date installed, MTBF/
    >>>uptime/failure rate etc., whether frequently power-cycled or running
    >>>24/7" would be great. Direct contact details to confirm this would
    >>>also be appreciated but not essential.
    >>>The reason for this request is to support the use of the Alphas in a
    >>>mission-critical role running a legacy app as part of the business
    >>>justification.
    >>>Posting to the forum preferred but mail to hp@pearson.in also accepted
    >>>if confidentiality required.

    >>
    >>My current uptime on my Alphastation 200 is 97 days. It's protected by
    >>a UPS but the run time after the power goes off is eight to ten minutes.
    >>Power, once off, takes two-three hours to restore. That's two to three
    >>hours to get two men and a truck here and two to three minutes to get a
    >>a man to the top of the pole in the "cherry picker" and replace the
    >>blown fuse.
    >>
    >>With reliable power and air conditioning, uptimes of four hundred days
    >>or more are quite possible. Your downtimes then are for reboots to
    >>install O/S upgrades and/or patches and to change SYSGEN parameters.
    >>Also allow for a shutdown every three or four years for major
    >>maintenance on the electrical systems.

    >
    >
    > Did you notice that he asked for a DS25 under Unix? I assume yours is
    > running under OpenVMS ;-)
    >
    > Regards,
    > Christoph Gartmann
    >



    No, I didn't notice that and mine do run VMS.

    Unix, however, is fairly reliable these days. I'm running Sun SPARC
    systems with Solaris 8, 9, and 10 and these boxes seem to be quite
    reliable. Not quite up to VMS standards but close enough for government
    work ;-)


  6. Re: Alphaserver/station reliability

    Richard B. Gilbert wrote:
    > Christoph Gartmann wrote:
    >> In article <47ECE96F.7010504@comcast.net>, "Richard B. Gilbert"
    >> writes:
    >>
    >>> david.pearson@british-energy.com wrote:
    >>>
    >>>> This is a little bit of an open request but for the installation and
    >>>> application I'm trying to justify it would really help me if I could
    >>>> get some 'real' uptime data from the wider industry on how reliable
    >>>> AlphaServers are. I know they have outstanding performance and are
    >>>> used by half of wall street etc. but it would be great to know from
    >>>> the real guys who have to keep these things running the real facts
    >>>> behind the claims.
    >>>> The two models I'm interested in are Alphaserver DS25 (a number of,
    >>>> both single and dual processor) and an Alphastation DS15. All will be
    >>>> running Tru64 Unix 5.1b.
    >>>> Simple feedback such as "Typical application, date installed, MTBF/
    >>>> uptime/failure rate etc., whether frequently power-cycled or running
    >>>> 24/7" would be great. Direct contact details to confirm this would
    >>>> also be appreciated but not essential.
    >>>> The reason for this request is to support the use of the Alphas in a
    >>>> mission-critical role running a legacy app as part of the business
    >>>> justification.
    >>>> Posting to the forum preferred but mail to hp@pearson.in also accepted
    >>>> if confidentiality required.
    >>>
    >>> My current uptime on my Alphastation 200 is 97 days. It's protected
    >>> by a UPS but the run time after the power goes off is eight to ten
    >>> minutes.
    >>> Power, once off, takes two-three hours to restore. That's two to
    >>> three hours to get two men and a truck here and two to three minutes
    >>> to get a a man to the top of the pole in the "cherry picker" and
    >>> replace the blown fuse.
    >>>
    >>> With reliable power and air conditioning, uptimes of four hundred
    >>> days or more are quite possible. Your downtimes then are for reboots
    >>> to install O/S upgrades and/or patches and to change SYSGEN
    >>> parameters. Also allow for a shutdown every three or four years for
    >>> major maintenance on the electrical systems.

    >>
    >>
    >> Did you notice that he asked for a DS25 under Unix? I assume yours is
    >> running under OpenVMS ;-)
    >>
    >> Regards,
    >> Christoph Gartmann
    >>

    >
    >
    > No, I didn't notice that and mine do run VMS.
    >
    > Unix, however, is fairly reliable these days. I'm running Sun SPARC
    > systems with Solaris 8, 9, and 10 and these boxes seem to be quite
    > reliable. Not quite up to VMS standards but close enough for government
    > work ;-)
    >


    And, besides, the O.P was specificaly asking about the *hardware*
    reliability, so unix vs. VMS might not be that important in this
    case...

    Jan-Erik.

  7. Re: Alphaserver/station reliability


    We have one DS15 in the field running OpenVMS V7.3-2. It was
    installed August 2004, replacing a DS10 that had similar reliability
    stats except for two service replacements of a TLZ10 DAT tape drive.

    There have been no hardware service events with the DS15. System has
    been running 24x7 except for four downs; two were power failures that
    outlasted the UPS, one for ECO update installations, and the most
    recent for rack installation.

    System was the primary application server at the site until late 2007
    and averaged 70 concurrent interactive users until then. It is
    currently used for access to archival information, about 4-8
    interactive users. It was also and remains the site email and web
    server and the central mail and time server for the customer's domain.

    Current uptime: 518 days. Last reboot cause: power interruption due
    to installation of the system in a rack.


  8. Re: Alphaserver/station reliability

    david.pearson@british-energy.com wrote:
    > The two models I'm interested in are Alphaserver DS25 (a number of,
    > both single and dual processor) and an Alphastation DS15.


    These are presumably going to be used systems since HP doesn't sell them
    anymore. So you need to consider not just how well these have run for
    their former owners for the last 5 years, but how well they will run for
    you for the next 5 years. For hardware purposes really the only things
    you need to be concerned about as computers age are disks, fans and
    electrolytic capacitors.

    Of these, the fans are by far the easiest to deal with since they come
    in standard sizes. You can usually find replacements from Delta, Sunon,
    or whoever made the fans in the first place. If ultra-reliability is a
    concern then you might consider putting in new fans before putting the
    units into service. Especially the CPU cooling fan, if there is one.
    You may be able to find the MTBF numbers for the fans, which will help
    you decide which ones should be preemptively replaced. In any case,
    before putting the machine into service (and with the power unplugged!)
    give each fan a spin with a finger or a blast of compressed air and
    verify that it moves smoothly and quietly.

    If the disks are more than a few years old you might prefer to install
    new ones before putting the machines into service. Again, this assumes
    reliability is a paramount concern. If a disk failure is not a show
    stopper then you can wait for failures before putting in new disks.

    The capacitors are the most difficult issue. I'm speaking generally
    here, since I don't have access to this specific hardware, but
    electrolytic caps are usually found in the power supply and scattered
    around the motherboard. Some machines have no electrolytic capacitors on
    the motherboard, and that may be true here as these were not run of the
    mill PCs. Electrolytic caps are difficult to test in situ, and the small
    ones on the motherboard are difficult to replace. About the best you
    can do is, before putting a used machine into service, to inspect all
    the capacitors carefully looking for bulging, leaking, or signs of past
    problems such as a stain next to a capacitor. Most failed capacitors
    are not kind enough to give any visual indications that they have left
    the building. My impression is that often when older machines just stop
    working, and it isn't the power supply, the cause is often a bad
    capacitor. Repair at that point is a costly option, but so may be
    finding replacements for these machines.

    Regards,

    David Mathog

  9. Re: Alphaserver/station reliability

    On Mar 28, 5:04 am, david.pear...@british-energy.com wrote:
    > This is a little bit of an open request but for the installation and
    > application I'm trying to justify it would really help me if I could
    > get some 'real' uptime data from the wider industry on how reliable
    > AlphaServers are. I know they have outstanding performance and are
    > used by half of wall street etc. but it would be great to know from
    > the real guys who have to keep these things running the real facts
    > behind the claims.
    > The two models I'm interested in are Alphaserver DS25 (a number of,
    > both single and dual processor) and an Alphastation DS15. All will be
    > running Tru64 Unix 5.1b.
    > Simple feedback such as "Typical application, date installed, MTBF/
    > uptime/failure rate etc., whether frequently power-cycled or running
    > 24/7" would be great. Direct contact details to confirm this would
    > also be appreciated but not essential.
    > The reason for this request is to support the use of the Alphas in a
    > mission-critical role running a legacy app as part of the business
    > justification.
    > Posting to the forum preferred but mail to h...@pearson.in also accepted
    > if confidentiality required.


    David,

    At this instant, I do not have any of the DS-class boxes in-house.
    However, in addition to the comments that have already been made, I
    will make the observation that atmospheric dust can be a major issue.

    Here in New York City, I got a demonstration of how much ambient
    particulate there actually is. A few years ago, a DEC axp 3000 (early
    generation AlphaStation) failed. When I opened the cover, the volume
    of dust was impressive. A quick cleaning with an air compressor and a
    vacuum (along the lines of "computer peridontics" -- removing the
    accumulated dust collection), and the system was back in service.

    Of course, it is not recommended to let things get to that point.

    - Bob Gezelter, http://www.rlgsc.com

  10. Re: Alphaserver/station reliability

    Bob Gezelter wrote:
    > On Mar 28, 5:04 am, david.pear...@british-energy.com wrote:
    >
    >>This is a little bit of an open request but for the installation and
    >>application I'm trying to justify it would really help me if I could
    >>get some 'real' uptime data from the wider industry on how reliable
    >>AlphaServers are. I know they have outstanding performance and are
    >>used by half of wall street etc. but it would be great to know from
    >>the real guys who have to keep these things running the real facts
    >>behind the claims.
    >>The two models I'm interested in are Alphaserver DS25 (a number of,
    >>both single and dual processor) and an Alphastation DS15. All will be
    >>running Tru64 Unix 5.1b.
    >>Simple feedback such as "Typical application, date installed, MTBF/
    >>uptime/failure rate etc., whether frequently power-cycled or running
    >>24/7" would be great. Direct contact details to confirm this would
    >>also be appreciated but not essential.
    >>The reason for this request is to support the use of the Alphas in a
    >>mission-critical role running a legacy app as part of the business
    >>justification.
    >>Posting to the forum preferred but mail to h...@pearson.in also accepted
    >>if confidentiality required.

    >
    >
    > David,
    >
    > At this instant, I do not have any of the DS-class boxes in-house.
    > However, in addition to the comments that have already been made, I
    > will make the observation that atmospheric dust can be a major issue.
    >
    > Here in New York City, I got a demonstration of how much ambient
    > particulate there actually is. A few years ago, a DEC axp 3000 (early
    > generation AlphaStation) failed. When I opened the cover, the volume
    > of dust was impressive. A quick cleaning with an air compressor and a
    > vacuum (along the lines of "computer peridontics" -- removing the
    > accumulated dust collection), and the system was back in service.
    >
    > Of course, it is not recommended to let things get to that point.
    >
    > - Bob Gezelter, http://www.rlgsc.com


    Dust bunnies are a common problem but they usually do not, or are not
    allowed to, interfere with system functionality.

    A computer does, however, act as a vacuum cleaner. Any dust it sucks up
    will likely remain inside the machine. A reasonably clean data center
    helps a lot. A thorough vacuum cleaning every couple of years is a good
    idea anyway. Just remove the cover, vacuum out the dust bunnies, and
    close it up again. With workstations used in an office environment,
    once a year is a good idea!

    I've never heard of a dust explosion inside a computer but there's a
    first time for everything!


  11. Re: Alphaserver/station reliability

    CHAIN: DS10L : Up since 11-OCT-2007 Current CPU temp: 45░
    BIKE: DS10L : Up since 11-OCT-2007 Current CPU temp: 40░
    VELO: VAX4000-600: Up since 11-OCT-2007.

    AKA: Last power failure was 11-OCT

    All these machines were used when I got them. They have been up without
    any problems for a couple years now. Not a data centre.

    They are on self maintenance.


    So if you have machines in a real data centre with conditioned power
    supply and UPS, they should run for a very long time.

    Make sure fans work properly.

    If configured in a cluster, you can make it such that you can bring one
    node down with littel disruption to service, perform hardwarde and or
    software maintenance and bring that node back up to re-integrate into
    the cluster. OOPSL: forgot, you're not running VMS.

  12. Re: Alphaserver/station reliability

    david.pearson@british-energy.com wrote:
    > This is a little bit of an open request but for the installation and
    > application I'm trying to justify it would really help me if I could
    > get some 'real' uptime data from the wider industry on how reliable
    > AlphaServers are. I know they have outstanding performance and are
    > used by half of wall street etc. but it would be great to know from
    > the real guys who have to keep these things running the real facts
    > behind the claims.
    > The two models I'm interested in are Alphaserver DS25 (a number of,
    > both single and dual processor) and an Alphastation DS15. All will be
    > running Tru64 Unix 5.1b.
    > Simple feedback such as "Typical application, date installed, MTBF/
    > uptime/failure rate etc., whether frequently power-cycled or running
    > 24/7" would be great. Direct contact details to confirm this would
    > also be appreciated but not essential.
    > The reason for this request is to support the use of the Alphas in a
    > mission-critical role running a legacy app as part of the business
    > justification.
    > Posting to the forum preferred but mail to hp@pearson.in also accepted
    > if confidentiality required.


    Both hardware and software are pretty bulletproof. Have run several vax
    and Alpha systems - one Alpha system / Tru64 almost none stop for nearly
    ten years without a single crash or hardware failure. It would probably
    have had indefinate uptimes were it not for the yearly shutdown to clean
    the dust out, fans, check connectors etc. This machine is semi retired
    now, but would probably do another ten years without a problem, since
    there are no parts running outside their design ratings.

    The added attention to detail is one of the things you pay the extra for
    with high end kit - all the extended environmental testing, quality
    parts, real engineering etc. May look the same as the cheap pc, but the
    difference is all under the skin and the fact that the system never
    breaks down has quite an impact on TCO :-)...

    Chris

    --

    ----------------------
    Greenfield Designs Ltd
    Electronic and Embedded System Design
    Oxford, England
    (44) 1865 750 681

  13. Re: Alphaserver/station reliability

    Richard B. Gilbert wrote:
    > david.pearson@british-energy.com wrote:
    >> This is a little bit of an open request but for the installation and
    >> application I'm trying to justify it would really help me if I could
    >> get some 'real' uptime data from the wider industry on how reliable
    >> AlphaServers are. I know they have outstanding performance and are
    >> used by half of wall street etc. but it would be great to know from
    >> the real guys who have to keep these things running the real facts
    >> behind the claims.
    >> The two models I'm interested in are Alphaserver DS25 (a number of,
    >> both single and dual processor) and an Alphastation DS15. All will be
    >> running Tru64 Unix 5.1b.
    >> Simple feedback such as "Typical application, date installed, MTBF/
    >> uptime/failure rate etc., whether frequently power-cycled or running
    >> 24/7" would be great. Direct contact details to confirm this would
    >> also be appreciated but not essential.
    >> The reason for this request is to support the use of the Alphas in a
    >> mission-critical role running a legacy app as part of the business
    >> justification.
    >> Posting to the forum preferred but mail to hp@pearson.in also accepted
    >> if confidentiality required.

    >
    > My current uptime on my Alphastation 200 is 97 days. It's protected by
    > a UPS but the run time after the power goes off is eight to ten minutes.
    > Power, once off, takes two-three hours to restore. That's two to three
    > hours to get two men and a truck here and two to three minutes to get a
    > a man to the top of the pole in the "cherry picker" and replace the
    > blown fuse.
    >
    > With reliable power and air conditioning, uptimes of four hundred days
    > or more are quite possible. Your downtimes then are for reboots to
    > install O/S upgrades and/or patches and to change SYSGEN parameters.
    > Also allow for a shutdown every three or four years for major
    > maintenance on the electrical systems.
    >
    > You CAN set things up to eliminate even more of the occasional down time
    > but it can get REALLY expensive. You would need dual Uninterruptable
    > Power Systems, dual (at least) air conditioners, dual emergency
    > generators for when the power goes off for several hours, etc. Few
    > companies need the uptime badly enough to pay the bills!
    >

    hmm.. I have that at home.. the 40KW diesel backup and UPS's.. Once
    those are installed, they cost little to maintain really. - and an
    XP1000 with OpenVMS been up for 330 days now and I need to blow the dust
    bunnies out occasionally, but I don't shut it off for that, I just
    remove the case and blow them out with some air, and this is in a 40-80
    degree 30-60% RH environment. I bet you won't have any problems with
    "DECompaqHP Alpha" type of hardware.

    (note: the case is usually part of the "cooling system", so don't do
    this 'hot cleaning' carelessly. It can be gotten away with on certain
    machines, and not on others.)

    Patrick J

  14. Re: Alphaserver/station reliability

    david.pearson@british-energy.com wrote:
    > This is a little bit of an open request but for the installation and
    > application I'm trying to justify it would really help me if I could
    > get some 'real' uptime data from the wider industry on how reliable
    > AlphaServers are. I know they have outstanding performance and are
    > used by half of wall street etc. but it would be great to know from
    > the real guys who have to keep these things running the real facts
    > behind the claims.
    > The two models I'm interested in are Alphaserver DS25 (a number of,
    > both single and dual processor) and an Alphastation DS15. All will be
    > running Tru64 Unix 5.1b.
    > Simple feedback such as "Typical application, date installed, MTBF/
    > uptime/failure rate etc., whether frequently power-cycled or running
    > 24/7" would be great. Direct contact details to confirm this would
    > also be appreciated but not essential.
    > The reason for this request is to support the use of the Alphas in a
    > mission-critical role running a legacy app as part of the business
    > justification.
    > Posting to the forum preferred but mail to hp@pearson.in also accepted
    > if confidentiality required.


    My experience with the hardware is >365 days of uptime - depending on
    necessary patching for production issues. Hardware runs forever..

  15. Re: Alphaserver/station reliability

    Bob Gezelter writes:

    >Here in New York City, I got a demonstration of how much ambient
    >particulate there actually is. A few years ago, a DEC axp 3000 (early
    >generation AlphaStation) failed. When I opened the cover, the volume
    >of dust was impressive. A quick cleaning with an air compressor and a
    >vacuum (along the lines of "computer peridontics" -- removing the
    >accumulated dust collection), and the system was back in service.


    Was it near/downwind of the World Trade Center on 9/11?

  16. Re: Alphaserver/station reliability

    As an indication of reliability, yesterday, for Earth Hour, I decided to
    shutdown one alpha. I figured I would use that opportunity to reboot
    the alpha in a special satellite mode without a local page file,
    allowing me to do an image backup of the main system disk to that
    workstation's disk.

    It had been so long since I rebooted a machine that i had to take out
    the boom for the magic incantation of the boot command to get it into
    sysboot mode.

  17. Re: Alphaserver/station reliability

    JF Mezei wrote:
    > As an indication of reliability, yesterday, for Earth Hour, I decided to
    > shutdown one alpha. I figured I would use that opportunity to reboot
    > the alpha in a special satellite mode without a local page file,
    > allowing me to do an image backup of the main system disk to that
    > workstation's disk.
    >
    > It had been so long since I rebooted a machine that i had to take out
    > the boom for the magic incantation of the boot command to get it into
    > sysboot mode.


    In my opinion planned reboots every 4 months or so is a good thing
    to keep people and startup scripts uptodate.

    Consider the famous irish rail system with 17 years of uptime.

    I assume that it did come up fine, but imagine if it had not.

    Guy at the console: hey it does not start up properly - has
    anyone changed anything in the startup scripts ?

    Group of other guys: since when ?

    Guy at the console: within the last 17 years !

    :-)

    Arne

  18. Re: Alphaserver/station reliability

    Arne Vajh°j wrote:
    > JF Mezei wrote:
    >
    >> As an indication of reliability, yesterday, for Earth Hour, I decided to
    >> shutdown one alpha. I figured I would use that opportunity to reboot
    >> the alpha in a special satellite mode without a local page file,
    >> allowing me to do an image backup of the main system disk to that
    >> workstation's disk.
    >>
    >> It had been so long since I rebooted a machine that i had to take out
    >> the boom for the magic incantation of the boot command to get it into
    >> sysboot mode.

    >
    >
    > In my opinion planned reboots every 4 months or so is a good thing
    > to keep people and startup scripts uptodate.
    >
    > Consider the famous irish rail system with 17 years of uptime.
    >
    > I assume that it did come up fine, but imagine if it had not.
    >
    > Guy at the console: hey it does not start up properly - has
    > anyone changed anything in the startup scripts ?
    >
    > Group of other guys: since when ?
    >
    > Guy at the console: within the last 17 years !
    >
    > :-)
    >
    > Arne


    But if you never restarted it in 17 years, why did you need to modify
    the startup scripts?

    And if you are going to keep a cluster running continuosly for 17 years
    I think you probably should have a test cluster on which you test all
    changes before putting them in production!


  19. Re: Alphaserver/station reliability

    On Mar 30, 8:54 pm, Arne Vajh°j wrote:
    > Consider the famous irish rail system with 17 years of uptime.


    That's *cluster*, not *system*.

    > I assume that it did come up fine, but imagine if it had not.
    >
    > Guy at the console: hey it does not start up properly - has
    > anyone changed anything in the startup scripts ?
    >
    > Group of other guys: since when ?
    >
    > Guy at the console: within the last 17 years !


    My cluster has been up for nearly 10 years but each node has been
    rebooted regularly - none of the nodes (nor the storage) in the
    cluster today were there when the cluster was formed in 1998. Most of
    the hardware hadn't even been invented yet! There were more Vaxes
    than Alphas orignally, but there hasn't been a single Vax in the
    cluster is several years. The OS has seen major upgrades as well as
    many minor patch updates. The links between data centers went from
    Gigaswitches to SAN ISLs. One of the data centers has lost power a
    few times. In all that time, the cluster has continued to serve end
    user applications.

    Reliability, Availability, Serviceability. When it comes to Alphas
    running VMS, pick any 3.

    .../Ed

  20. Re: Alphaserver/station reliability

    ewilts@ewilts.org wrote:
    > On Mar 30, 8:54 pm, Arne Vajh°j wrote:
    >> Consider the famous irish rail system with 17 years of uptime.

    >
    > That's *cluster*, not *system*.


    Sure ?

    I always thought it was a single system.

    Arne

+ Reply to Thread
Page 1 of 2 1 2 LastLast