1 INFO-VAX	Wed, 06 Apr 2005	Volume 2005 : Issue 191       Contents: ! The Greatest News Ever ! Re: Broken vax4000 Re: Broken vax4000 Re: Broken vax4000 Re: Broken vax4000 Re: Broken vax40000 Re: Different behaviour of HP C on Alpha and VAX5 Re: Frustration!  EB64+ too old to run OpenVMS 7.3-1? 4 Re: High File Watermarking question for Hoff HoffmanP Re: I love a sunburnt Country. . . (or VMS opportunities in Perth Western AustraP Interex Announces Record Number of Papers Submitted for HP World 2005 Conference Make Easy $$$$!!6 Maximum Mailbox limitations with OVMS732 and TCPIP54E4$ Re: poor disk I/O performace on ds25$ Re: poor disk I/O performace on ds25$ Re: poor disk I/O performace on ds25$ Re: poor disk I/O performace on ds25$ Re: poor disk I/O performace on ds25/ Re: Procedure for installing PERL modules (LWP) / Re: Procedure for installing PERL modules (LWP) / Re: Procedure for installing PERL modules (LWP)  Re: Remote printing woes.  Re: shadow minicopy  Re: shadow minicopy 
 Site down? Re: Time change questions  Re: Time change questions  Re: Time change questions  Re: Time change questions  Re: Time change questions  Re: VMS Command language.  Re: VMS Command language.  VMS:  nothing else can compare? # Re: VMS:  nothing else can compare? # Re: VMS:  nothing else can compare? # Re: VMS:  nothing else can compare? # Re: VMS:  nothing else can compare? B Which release notes say sts$manager:utc$configure_tdf  is obsoleteF Re: Which release notes say sts$manager:utc$configure_tdf  is obsoleteF Re: Which release notes say sts$manager:utc$configure_tdf  is obsoleteF Re: Which release notes say sts$manager:utc$configure_tdf  is obsolete  F ----------------------------------------------------------------------   Date: 5 Apr 2005 16:05:23 -0700  From: Ron2005_010@yahoo.com # Subject: ! The Greatest News Ever ! B Message-ID: <1112742323.328842.56430@o13g2000cwo.googlegroups.com>     ; http://www.print-it.blogspot.com << The Greatest News Ever!    ------------------------------  # Date: Tue, 05 Apr 2005 18:37:34 GMT # From: hoff@hp.nospam (Hoff Hoffman)  Subject: Re: Broken vax4000 0 Message-ID: <OVA4e.3014$US.668@news.cpqcorp.net>  M In article <4252D6D2.68F70223@mist.com>, GreyCloud <cumulus@mist.com> writes: ; :My vax just died...  I notice that the router box has only 8 :one light lit instead of two.  The Full/Col light isn't= :lit.  The o/s doesn't boot up.  So I installed a spare drive   :with a spare o/s on it.  No go.  E   VAXstation 4000 VLC?  VAXstation 4000 model 60?  One of the various E   VAXstation 4000 model 90-series boxes?  Or is this a VAX 4000 model D   10x series box?  Model 200?  300?  400?  500?  505A? ... Well, youE   get the idea.  :-)  There are a *lot* of similarly named VAX boxes, 3   and details can help with the hardware debugging.   D   Anything on the console, as most of these have console self-tests.B   If the console isn't showing anything at all (and the console isA   known to have worked in the past), then what I *think* you are  >   reporting tends to be a power, power supply, or a low-level A   motherboard failure.  Do the fans spin up?   Any indicator LEDs C   on whichever VAX this is?  These can be on the CPU module, on the A   power supply, and elsewhere, depending on the VAX model and the    particular configuration.   A   Are there symptoms or behaviours that led you to swap the disk? B   (For example, the sounds or smells of a head crash -- I haven't B   witnessed a really notable head crash in years, as compared withB   the sturm and drang of disk crashes of old.  But I digress.  :-)  A   If you are neither sentimentally nor hardware-attached to this  C   particular VAX box :-), replacing it can be an option, of course.     N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq N  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com    ------------------------------  % Date: Tue, 05 Apr 2005 15:02:54 -0600 " From: GreyCloud <cumulus@mist.com> Subject: Re: Broken vax4000 ( Message-ID: <4252FCFE.723B104A@mist.com>   Hoff Hoffman wrote:  > O > In article <4252D6D2.68F70223@mist.com>, GreyCloud <cumulus@mist.com> writes: = > :My vax just died...  I notice that the router box has only : > :one light lit instead of two.  The Full/Col light isn't? > :lit.  The o/s doesn't boot up.  So I installed a spare drive " > :with a spare o/s on it.  No go. >  >   VAXstation 4000 VLC?     Yes, this model.  / > VAXstation 4000 model 60?  One of the various G >   VAXstation 4000 model 90-series boxes?  Or is this a VAX 4000 model F >   10x series box?  Model 200?  300?  400?  500?  505A? ... Well, youG >   get the idea.  :-)  There are a *lot* of similarly named VAX boxes, 5 >   and details can help with the hardware debugging.  >    Using Putty to access the box.  F >   Anything on the console, as most of these have console self-tests.D >   If the console isn't showing anything at all (and the console isB >   known to have worked in the past), then what I *think* you are? >   reporting tends to be a power, power supply, or a low-level / >   motherboard failure.  Do the fans spin up?    : Yes, the fans spin up.  The power supply seems to be fine,8 and the hard drive makes an attempt at booting but stops< quite quickly.  I have a MAU external converter and the leds+ light up.  Just no communications going on.    >  Any indicator LEDs E >   on whichever VAX this is?  These can be on the CPU module, on the C >   power supply, and elsewhere, depending on the VAX model and the  >   particular configuration.  > C >   Are there symptoms or behaviours that led you to swap the disk?   4 Yes, the lack of any booting or hard drive activity.  C >   (For example, the sounds or smells of a head crash -- I haven't D >   witnessed a really notable head crash in years, as compared withD >   the sturm and drang of disk crashes of old.  But I digress.  :-) >   - The hard drive is a new one, a 2Gb barracuda.    B >   If you are neither sentimentally nor hardware-attached to thisE >   particular VAX box :-), replacing it can be an option, of course.  >   ; I've considered this, but I have to look around.  I may pay ; Island computing a visit.  It would be nice to get a CD-ROM ; drive with one so I can acquire a Hobbyist license.  I'm an 6 old retired VAX programmer and sort of like to keep in2 touch.  I don't find anything of any interest from: Microsoft, so I made a decision to stay with VMS and their< version of Fortran.  I have Fortran on this old VLC now. :-(  P >  ---------------------------- #include <rtfaq.h> -----------------------------M >     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq P >  --------------------------- pure personal opinion ---------------------------G >         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com    Thnx for the faq.    ------------------------------  # Date: Tue, 05 Apr 2005 22:43:39 GMT # From: hoff@hp.nospam (Hoff Hoffman)  Subject: Re: Broken vax4000 2 Message-ID: <vwE4e.3049$qd1.1389@news.cpqcorp.net>  M In article <4252FCFE.723B104A@mist.com>, GreyCloud <cumulus@mist.com> writes:  :Hoff Hoffman wrote: :>  P :> In article <4252D6D2.68F70223@mist.com>, GreyCloud <cumulus@mist.com> writes: :> :My vax just died...  :>   VAXstation 4000 VLC?  :Yes, this model.    :Using Putty to access the box.   I   How you connect into the console serial line is rather less interesting I   here than what is displayed during the console self-test.  If anything.   ; :Yes, the fans spin up.  The power supply seems to be fine, 9 :and the hard drive makes an attempt at booting but stops  :quite quickly.     D   I would be interested in the output from the console self-test, ifH   anything.  (Not knowing what you are hearing, I'm not certain if what G   you are reporting here is the disk spinning up and loading its heads, G   or if you are hearing the start of the system bootstrap and some sort G   of a bootstrap failure.  This is why the console output is a key clue E   here, as that -- if you can get as far as the console self-tests -- D   will indicate if there is a hardware problem, or a problem booting   OpenVMS VAX.)   , I have a MAU external converter and the leds, :light up.  Just no communications going on.  C   Putty can be used over the network or over a serial line, and the @   reference to MAU implies a network connection and not a serialB   connection.  You need to use the console serial line here.  (TheC   pinouts and such are listed in the FAQ -- I'm obviously guessing  B   here that you do not presently utilize the console serial line.)  C   Select the S3 switch upwards, connect to the MMJ on the back (and A   the pinouts and adapters are in the FAQ), set the terminal, the C   terminal emulator or what-ever to 9600 8N1.  Power-cycle the box.   B   You should see the console output.  This output is likely key toB   the troubleshooting of your reported bootstrap failure.  (If youA   get no console output on power-up, the box is _really_ sick, or C   the adapter, wiring, terminal or terminal emulator, or the serial +   connection is somehow incorrect or sick.)   . :The hard drive is a new one, a 2Gb barracuda.  E   Have you ever bootstrapped this OpenVMS VAX configuration from this C   disk?  (Or did this question just turn into a "fun with SCSI disk    installations" question? :-)    N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq N  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com    ------------------------------  % Date: Tue, 05 Apr 2005 19:10:32 -0700 # From: "Tom Linden" <tom@kednos.com>  Subject: Re: Broken vax4000 ( Message-ID: <opsosgbuiizgicya@hyrrokkin>  F On Tue, 05 Apr 2005 22:43:39 GMT, Hoff Hoffman <hoff@hp.nospam> wrote:  I > In article <4252FCFE.723B104A@mist.com>, GreyCloud <cumulus@mist.com>   	 > writes:  > :Hoff Hoffman wrote: > :>L > :> In article <4252D6D2.68F70223@mist.com>, GreyCloud <cumulus@mist.com>  	 > writes:  > :> :My vax just died...  > :>   VAXstation 4000 VLC?  > :Yes, this model.  > ! > :Using Putty to access the box.  > K >   How you connect into the console serial line is rather less interesting K >   here than what is displayed during the console self-test.  If anything.   I PuTTY is a telnet or SSH interface, so it is relevant.  It appears he has G some sort of BillyBox running PuTTY.  What he needs to do is to connect I hyperterminal 9600-8-N-1 to the serial port.  Of course if CONSOLE is not G set serial then there is a problam and he will need to connect directly  to the graphics port > = > :Yes, the fans spin up.  The power supply seems to be fine, ; > :and the hard drive makes an attempt at booting but stops  > :quite quickly.  > F >   I would be interested in the output from the console self-test, ifI >   anything.  (Not knowing what you are hearing, I'm not certain if what I >   you are reporting here is the disk spinning up and loading its heads, I >   or if you are hearing the start of the system bootstrap and some sort I >   of a bootstrap failure.  This is why the console output is a key clue G >   here, as that -- if you can get as far as the console self-tests -- F >   will indicate if there is a hardware problem, or a problem booting >   OpenVMS VAX.)  > . > I have a MAU external converter and the leds. > :light up.  Just no communications going on. > E >   Putty can be used over the network or over a serial line, and the B >   reference to MAU implies a network connection and not a serialD >   connection.  You need to use the console serial line here.  (TheD >   pinouts and such are listed in the FAQ -- I'm obviously guessingD >   here that you do not presently utilize the console serial line.) > E >   Select the S3 switch upwards, connect to the MMJ on the back (and C >   the pinouts and adapters are in the FAQ), set the terminal, the E >   terminal emulator or what-ever to 9600 8N1.  Power-cycle the box.  > D >   You should see the console output.  This output is likely key toD >   the troubleshooting of your reported bootstrap failure.  (If youC >   get no console output on power-up, the box is _really_ sick, or E >   the adapter, wiring, terminal or terminal emulator, or the serial - >   connection is somehow incorrect or sick.)  > 0 > :The hard drive is a new one, a 2Gb barracuda. > G >   Have you ever bootstrapped this OpenVMS VAX configuration from this E >   disk?  (Or did this question just turn into a "fun with SCSI disk   >   installations" question? :-) >  > 4 >  ---------------------------- #include <rtfaq.h>   > ----------------------------- 5 >     For additional, please see the OpenVMS FAQ --    > www.hp.com/go/openvms/faq 6 >  --------------------------- pure personal opinion   > --------------------------- G >         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com  >    ------------------------------  % Date: Tue, 05 Apr 2005 21:15:57 -0600  From: Kofi Anan <crook@UN.com> Subject: Re: Broken vax4000 0 Message-ID: <urSdndDoRc9nyc7fRVn-1g@bresnan.com>   Hoff Hoffman wrote: O > In article <4252FCFE.723B104A@mist.com>, GreyCloud <cumulus@mist.com> writes:  > :Hoff Hoffman wrote: > :>  R > :> In article <4252D6D2.68F70223@mist.com>, GreyCloud <cumulus@mist.com> writes: > :> :My vax just died...  > :>   VAXstation 4000 VLC?  > :Yes, this model.  > ! > :Using Putty to access the box.  > K >   How you connect into the console serial line is rather less interesting K >   here than what is displayed during the console self-test.  If anything.  > = > :Yes, the fans spin up.  The power supply seems to be fine, ; > :and the hard drive makes an attempt at booting but stops  > :quite quickly.    > F >   I would be interested in the output from the console self-test, ifJ >   anything.  (Not knowing what you are hearing, I'm not certain if what I >   you are reporting here is the disk spinning up and loading its heads, I >   or if you are hearing the start of the system bootstrap and some sort I >   of a bootstrap failure.  This is why the console output is a key clue G >   here, as that -- if you can get as far as the console self-tests -- F >   will indicate if there is a hardware problem, or a problem booting >   OpenVMS VAX.)  > . > I have a MAU external converter and the leds. > :light up.  Just no communications going on. > E >   Putty can be used over the network or over a serial line, and the B >   reference to MAU implies a network connection and not a serialD >   connection.  You need to use the console serial line here.  (TheE >   pinouts and such are listed in the FAQ -- I'm obviously guessing  D >   here that you do not presently utilize the console serial line.) > E >   Select the S3 switch upwards, connect to the MMJ on the back (and C >   the pinouts and adapters are in the FAQ), set the terminal, the E >   terminal emulator or what-ever to 9600 8N1.  Power-cycle the box.  > D >   You should see the console output.  This output is likely key toD >   the troubleshooting of your reported bootstrap failure.  (If youC >   get no console output on power-up, the box is _really_ sick, or E >   the adapter, wiring, terminal or terminal emulator, or the serial - >   connection is somehow incorrect or sick.)   D I'll try these suggestions and get back to you if I do see anything.   > 0 > :The hard drive is a new one, a 2Gb barracuda. > G >   Have you ever bootstrapped this OpenVMS VAX configuration from this 
 >   disk?   E Yes, and it was working great too.  A question remains as to storage   temperatures as well. F This may have killed the vlc... the temp was -27F low for the winter.  I didn't use the vlc% for at least a year, as I have moved.   2 Anyway, the spare drive also used to boot the vlc.  <   (Or did this question just turn into a "fun with SCSI disk  >   installations" question? :-) >  > P >  ---------------------------- #include <rtfaq.h> -----------------------------M >     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq P >  --------------------------- pure personal opinion ---------------------------G >         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com  >    ------------------------------  % Date: Tue, 05 Apr 2005 15:29:19 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> 9 Subject: Re: Different behaviour of HP C on Alpha and VAX B Message-ID: <1112729433.63b41e7a656b95dd3c39204cd37e87b6@teranews>  A by the wya, you should always compile with cc/UNSIGNED when using F characters. Otherwise the C compiler will complain when you try to useG the run time functions to manipulate strings which you have declared as  unsigned char.  F when you compile with /UNSIGNED, you can declare all your strings withE "char" and not worry about compiler generating code that purposefully G chops off the 8th bit of characters or considers that  is smaller than " e because it thinks  is negative.   ------------------------------  % Date: Tue, 05 Apr 2005 15:02:41 -0700 - From: Jake Hamby <jhamby IS AT pobox DOT com> > Subject: Re: Frustration!  EB64+ too old to run OpenVMS 7.3-1?8 Message-ID: <5v1651pg6874f4c5126javfhghp23d4k9p@4ax.com>  F On Tue, 05 Apr 2005 16:23:27 GMT, hoff@hp.nospam (Hoff Hoffman) wrote:  D >  Confirm the hardware configuration, and check for newer firmware.B >  (AFAIK, all supported Alpha systems have (re)loadable firmware.C >  Given the age of the system, I'd expect to see the firmware only D >  on older distributions -- I'd not expect to find recent updates.)  E This is the *one* Alpha motherboard without a flash ROM.  I found the C manual in PDF and it specifically says (as have all the other sites F that mention the EB64+) that it has a UVPROM that must be reprogrammedB externally.  Also, the ROM image is only available as part of someC firmware EBSDK that was once available for $75 and I'm not going to * pay that kind of money for this old thing.  ? >:Or could it be the ISA video card that's causing the problem?  > B >  Yes, or any other unsupported hardware.  AFAIK, there were few B >  (one?) supported ISA video card, and AFAIK it was not supported> >  on this platform.  You will have fun with this box as it is= >  old enough that it lacks support for the more current and  B >  more widely-available graphics controllers -- a PowerStorm 3D30D >  or 4D20 will *probably* work here, but some of the newer graphicsE >  controllers require EV6-class processors and non-swizzled bus I/O.   A Yeah, I am 99% certain now that the video card is the problem.  I C don't know who found this bizarre card to put in the machine, but I F was finally able to determine from the FCC ID that it was made by "MGV# Manufacturing" of Madison, Alabama.   H http://www.fcc.gov/fcc-bin/ead?Search=%5EI7ZVN&offset=1&db=Certification  F It also appears to be some sort of floppy or MFM hard drive controllerD because it has five DIP switches and a strange connector opposite toF the ISA end that looks like an ISA connector but is only 10 pins wide.F I found a site selling Diamond Trio64 PCI cards for $4.95 so I'll haveF to wait for that to arrive and hopefully it will solve the problem.  IC did see the EB64+ listed as supported in the OpenVMS 7.3-2 SPD, and 4 there are no other cards in it (the NIC is onboard).  F >:I'm going to try to bid on a cheap DEC Personal Workstation 600AU onD >:eBay this week, of which there are several for sale at the moment. > B >  I would personally aim for an AlphaStation series, including anA >  XP900, XP1000 or equivalent box.  I would not acquire an Alpha B >  system with less than an EV56 processor, and I would prefer and5 >  would look for an EV6-class processor, or better.     Thanks for the info.   -Jake Hamby    ------------------------------   Date: 5 Apr 2005 23:02:53 -0500 - From: Kilgallen@SpamCop.net (Larry Kilgallen) = Subject: Re: High File Watermarking question for Hoff Hoffman 3 Message-ID: <bZNpq01qOQDe@eisner.encompasserve.org>   Z In article <BE787C79.B0C0%roktsci@comcast.net>, Jeff Cameron <roktsci@comcast.net> writes: > Dear Hoff, > $ > Settle an argument, if you please. > D > Is the following true or not, about High file Watermarking in VMS?  > Please feel free to elaborate. >  > Assertion:E > On a disk with High File Watermarking enabled, blocks are zeroed on  > allocation/extent.   >  > Supplemental: M > In the case of sequential files, the bytes between the logical and physical E > EOF are also zeroed when the logical file mark location is written.   C Note that this is a _different_ assertion than the one in the other @ thread, which suggested that blocks on disk would be zeroed whenC a file is opened, even though the EOF was at the start of the file.    ------------------------------  % Date: Tue, 05 Apr 2005 15:32:24 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> Y Subject: Re: I love a sunburnt Country. . . (or VMS opportunities in Perth Western Austra B Message-ID: <1112729559.1348c861ae6455ed6c2a633487b5f9dd@teranews>   Richard Maher wrote: >  > Hi,  > J > We've finally bitten the bullet and decided to pack the kids up and head > back to Perth.  G Make sure you leave sufficient sized holes in the boxes for the kids to G breathe ! And give them a blanket because it may get mighty cold in the  cargo hold of the aircraft     :-) :-) :-) :-) :-)    ------------------------------  % Date: Wed, 06 Apr 2005 07:43:43 +0200 - From: Didier Morandi <prenom.nom@freesurf.fr> Y Subject: Interex Announces Record Number of Papers Submitted for HP World 2005 Conference 4 Message-ID: <42537711$0$26512$636a15ce@news.free.fr>   http://minilien.com/?kBbQDXq79E   F Interex Announces Record Number of Papers Submitted for HP World 2005 E Conference; HP Signs up for Largest Single Sponsorship in Event's 10   Year History  
 HP World 2005   @ SUNNYVALE, Calif.--(BUSINESS WIRE)--April 5, 2005--Interex, The F International Association of Hewlett-Packard Computing Professionals, B announced today that they received more than 500 submissions from I respected IT technologists for the opportunity to speak at HP World 2005  2 in San Francisco, August 14-18, at Moscone Center.  F As a result Interex is able to offer HP World 2005 attendees a robust G conference program that features 250+ technical sessions, 20+ hands-on  D workshops, a full day of seminars, a new Storage Networking Program E featuring SNIA sessions and certification seminar, Microsoft Tech-Ed  C sessions, in-depth Masters Courses, keynote presentations from Ann  A Livermore, Executive VP of HP's Technology Solutions Group; Kris  E Hagerman, Executive VP of the Storage and Server Management Group at  H VERITAS Software; Chuck Rozwat, Executive VP of Server Technologies for > Oracle; and Patrick J. Martin, Chairman, President and CEO of H StorageTek; plenary sessions from Ira Winkler, Bruce Perens and John C. H Dvorak, on-site Thomson Prometric certification testing, and a Meet the C Experts Forum that will provide attendees with access to dozens of  K leading industry technologists. For more information go to www.hpworld.com.   A "The response from the IT community to our call for speakers was  A remarkable this year," said Nadine Flaherty, HP World Conference  F Director. "As a result we are able to offer training opportunities at 8 this year's HP World that can't be found anywhere else."  I Interex also announced that HP has purchased a Premier Sponsorship at HP  F World 2005, the largest financial commitment for a single sponsorship , from HP in the 10 year history of the event.  G "We are pleased that HP has chosen to make a significant investment in  B HP World 2005," said Ronald W. Evans, Executive Director & CEO of D Interex. "The support of all of our sponsors has enabled Interex to E offer IT professionals a wide variety of valuable technical training   opportunities this year."   E HP World 2005 Conference & Expo will be held August 14-18 at Moscone  E Center in San Francisco, CA. The Expo will be held August 16-18. For  F information about HP World 2005 please visit the event's official web 9 site at www.hpworld.com. Online registration is now open.   
 About Interex   F Interex is the world's largest independent organization of IT leaders A incorporating HP and HP-compatible technology in their computing  C strategy. For more than 30 years, Interex has produced the world's  H leading HP-centric conferences, publications, online services and other < educational tools supporting its members' multi-platform IT E infrastructures, including UNIX systems, Windows server, Windows XP,  F OpenView, Linux, NonStop, OpenVMS, MPE/iX, HP-UX, Pocket PC, .NET and D other environments. For more information on Interex -- where shared > knowledge equals shared power -- please visit www.interex.org. 	  Contacts 			  Interex  Simon Tonner, 408-743-4616  = (super, that Google Alerts (http://www.google.com/alerts) :-)    D. --  ,        Didier MORANDI - Expert informaticien0   13 chemin du Gu, 1213 Petit-Lancy (GE) Suisse0 Tl. : +33(0)6 7983 6418 ~ www.didiermorandi.com   ------------------------------   Date: 5 Apr 2005 11:48:53 -0700  From: glenn9466@netscape.net Subject: Make Easy $$$$!! C Message-ID: <1112726933.940789.286130@l41g2000cwc.googlegroups.com>   > Follow the directions below and in two weeks you'll have up to> $20000.00 in your PayPal account. There is a very high rate ofC participation in the program because of its low investment and high ) rate of return. Just $5.00 to one person!    THAT'S ALL !!!  B If you are a skeptic and don't think the program will work, I urge8 you to give it a try anyway! It REALLY WORKS! Why do you% think so many people are promoting it   G You'll immediately think "scam", and I don't blame you. It shouldn't be G this easy to make money. But THIS IS NOT A SCAM. It's known as a "cycle G system". Person A recieves $5 from x amount of people, depending on how C well they circulate the message. Person B recieves $5 from the next E allotment of participants, and so on, and so on. While $20,000 is not @ common $300-$500 IS. For a $5 investment, that's not too shabby.  E LOOK AT IT THIS WAY If the Program is a total failure for you and you 4 never get even $1.00 in return, your total loss willE be the $5.00! If you are not yet a paypal member, there is no risk at 7 all!!! If the Program is only moderately successful for D you, your PayPal account will have several hundred dollars deposited1 into it within the next few days! If you actively C participate in the Program, you could have up to $20,000.00 in your   PayPal account within two weeks!  ' Now let me tell you the simple details.    Getting Started!!   E If you're not already a user of PayPal, the very first thing you need 7 to do is go to PayPal and sign up. It takes two minutes B and Pay Pal will deposit $5.00 in your account just for becoming a9 member. That makes this program's total cost $0!!! Follow % this link to open your PayPal account    http://www.paypal.com   D Now log into your PayPal account, and send the PayPal account of the8 person listed in Position 1 $5.00 PayPal will ask you to7 select type. (Select service and put $5.00 donation for A subject.) When person in Position 1 receives notification of your > payment, you can simply copy this page and change the names in> position #1 & #2 & #3 as instructed. Remember, only the person? in Position 1 on the list gets your $5.00 donation. Send them a > donation then remove #1 PayPal account from the list. Move theE other two accounts up & add your Paypal account to #3 position. After , you have retyped the names in the new order,  C IMMEDIATELY send the revised message to as many people as possible. 7 PROMOTE! PROMOTE! The more you promote the Program, the ? more you will receive in donations!! That's all there is to it. B You are reading this message in usenet, and usenet is the best wayC to spread the word about the program. Post this message to AT LEAST B 200 groups in usenet(there are over 24, 000), after you send the 5B dollars to the person at #1. This will guarentee you a profit fromE this program. The more groups you post it to, the more money you will ? make!!!.You can use a program like postXpert to post to all the 0 newsgroups at once. You can find this program at httpwww.download.com. C Use Netscape or Internet Explorer and try searching for various new  groups; (on- line forums, message boards, chat sites, discussions.) > Visit message boards and post this article as a new message by highlighting? the text of this letter and selecting paste from the edit menu. B Fill in the subject, this will be the header that everyone sees as they scroll3 through the list of postings in a particular group, + click the post message button. You're done.   B When your name reaches Position 1 (usually in less than a week) it9 will be your turn to receive the cash. $5.00 will be sent F to your PayPal account by people just like you who are willing to send6 $5.00 dontation and receive up to $20,000 in less than? two weeks. Because there are only (3) names on the list you can - anticipate 80% of your cash within two weeks.   @ Anytime you find yourself short on cash just take out your $5.00< donation program and send it to 50 prospects. Imagine if youB sent it to 100 or even more. Most people spend more than $5 on the5 lottery every week with no real hope of ever winning.   F IMPORTANT!!! IN ORDER FOR THIS PROGRAM TO WORK, YOU MUST BE HONEST. DOC NOT ADD YOUR EMAIL IMMEDIATELY TO THE #1 POSITION, AND MAKE SURE TO C SEND YOUR 5 DOLLARS TO THE PERSON AT #1. IF EVERYONE WHO TRIED THIS C DIDNT SEND THE MONEY, THEN NOBODY WOULD MAKE A DIME. 5 DOLLARS IS A C VERY SMALL INVESTMENT, ESPECIALLY WHEN YOUR ABOUT TO MAKE MANY MANY  MANYA TIMES THAT AMOUNT. IF WE ARE ALL HONEST, THEN WE ALL MAKE LOTS OF  MONEY!!   C REMEMBER, you add your email to the #3 spot, and then move everyone  elseE up 1(deleting the person who was formally in the #1 spot, and who you E should have sent your money to). DO NOT add your email to #1 when you B start this program. If everyone did this, then NO ONE would make a cent.     THIS PROGRAM WORKS - JUST TRY IT      ( POSITION 1 PAYPAL ACCOUNT clatte@shaw.ca  4 POSITION 2 PAYPAL ACCOUNT joannejenkins1@bigpond.com  / POSITION 3 PAYPAL ACCOUNT glenn1966@verizon.net   * Integrity and honesty make this plan work.C Participants who actively promote this program will average between  $8000  and 2 $12000 and receive the donations within two weeks.  E This is not a chain letter. You are simply making a donation of $5.00 5 to another person. The Program does not violate title / 18 section 1302 of the Postal and lottery code.   8 Remember -TIME is of the essence. YOU can choose to liveB Paycheck-to-Paycheck or live FREE from FINANCIAL BONDAGE. Become a9 part of the donation program and help people help people.   ) This program is about helping each other!   * Success is a journey - Not a destination!    Start Your Journey TODAY!!!!   ------------------------------   Date: 5 Apr 2005 13:19:11 -0700 " From: dave.baxter@bannerhealth.com? Subject: Maximum Mailbox limitations with OVMS732 and TCPIP54E4 C Message-ID: <1112732351.281800.268140@l41g2000cwc.googlegroups.com>   @ With OpenVMS 7.3-2 and TCPIP V5.4 ECO4, the maximum number of BGD devices is increased to 32767.     I understand that this is done byE setting the sysconfig parameter "ovms_unit_maximum = 32767".   Do any A of the other parameters have to be changed?   In particular, I am  thinking about  ) "ovms_unit_limit" (currently set to 9998) ( "ovms_unit_fast_credel"  (currently = 0)   thanks   Dave.    ------------------------------   Date: 5 Apr 2005 13:29:29 -0700 $ From: "AEF" <spamsink2001@yahoo.com>- Subject: Re: poor disk I/O performace on ds25 C Message-ID: <1112732969.921227.246600@l41g2000cwc.googlegroups.com>    Larry Kilgallen wrote:E > In article <1112645711.055624.211920@o13g2000cwo.googlegroups.com>, & "AEF" <spamsink2001@yahoo.com> writes: > >  > > Larry Kilgallen wrote: > > > >> For a test of of overhead from Highwater Marking, try the	 following 6 > >> both with Highwater Marking enabled and disabled. > >> > >> 	Create a new file + > >> 	Write 100000 blocks of non-zero bytes  > >> 	Close the file( > >> 	Report the resources and time used > >>F > >> A different test would be to allocate the file to a large size on
 > > creation.: > >h > >a9 > > OK. It's near instantaneous either way using COPY NL:e A.A/ALLOC=100000.e >rF > Yes, because that does not write 100000 blocks of non-zero (or zero)E > bytes.  It creates a file with a large allocation but leaves EOF at  > the _start_ of the file.     Correct.    G > > However, setting the eof mark from the beginning to the end is neartA > > instantaneous without highwater but took 27 seconds with! OK, D > > non-shared sequential files have the "security pattern" of zeros" > > written when the eof is moved. >d& > As described earlier in this thread.    D Correct. So part of my post was off. Sorry. I can't be top notch all
 the time! :-)a   ------------------------------   Date: 5 Apr 2005 16:38:46 -0500n- From: Kilgallen@SpamCop.net (Larry Kilgallen) - Subject: Re: poor disk I/O performace on ds25o3 Message-ID: <cCdQbC$7d5Zy@eisner.encompasserve.org>   Z In article <BE7769E6.AFF5%roktsci@comcast.net>, Jeff Cameron <roktsci@comcast.net> writes:M > On 4/4/05 7:38 AM, in article i4Q0tsJXozr8@eisner.encompasserve.org, "Larry + > Kilgallen" <Kilgallen@SpamCop.net> wrote:  > ? >> In article <BE76088B.AE0F%roktsci@comcast.net>, Jeff Camerone  >> <roktsci@comcast.net> writes: >>  N >>> I can appreciate the meaning behind the nomenclature, but that is not whatO >>> is happening. I just got finished writing a test program in FORTRAN where IgO >>> used SYS$ASSIGN to open a file of 2048 blocks and used SYS$QIOW to read theiN >>> blocks, and when done on a disk with HFWM active all blocks were zero, but5 >>> were not on the same disk when HFWM was disabled.r >>  8 >> That is exactly the description of Highwater Marking.? >> But it does not tell you whether those blocks were zeroed on & >> disk before you tried to read them.. >> And that does not mimic any useful program.K > Are you saying that the 100 or so times that the program is run with HFWMgJ > enabled always yields all blocks as zeros vs. 100 or so runs of the sameK > program on the same disk with HFWM disabled always yields non zero blocksj > was just a coincidence?e   No.D  K > I was not attempting to "mimic" a useful program, nor measure overhead ofmN > HFWM. It does, however exhibit what HFWM does what the documentation says it> > does, that being blocks are zeroed out on allocation/extent.  E No, it does not and they are not.  You have only shown that the bytes & were zero when you tried to read them.  F >> For a test of of overhead from Highwater Marking, try the following4 >> both with Highwater Marking enabled and disabled. >> X >> Create a new file( >> Write 100000 blocks of non-zero bytes >> Close the file % >> Report the resources and time used  >> mN >> A different test would be to allocate the file to a large size on creation. > K > I performed this test too, which also validated my assertion. The time itcJ > takes to open a large pre-allocated file takes significantly longer when- > HFWM is enabled vs. when it is not enabled.   B It should take no longer to open the file so long as EOF is at the beginning of the file.   ------------------------------  % Date: Tue, 05 Apr 2005 17:25:52 -0700c( From: Jeff Cameron <roktsci@comcast.net>- Subject: Re: poor disk I/O performace on ds25e/ Message-ID: <BE787AA0.B0BE%roktsci@comcast.net>   K On 4/5/05 2:38 PM, in article cCdQbC$7d5Zy@eisner.encompasserve.org, "Larryh) Kilgallen" <Kilgallen@SpamCop.net> wrote:e  > > In article <BE7769E6.AFF5%roktsci@comcast.net>, Jeff Cameron > <roktsci@comcast.net> writes:dN >> On 4/4/05 7:38 AM, in article i4Q0tsJXozr8@eisner.encompasserve.org, "Larry, >> Kilgallen" <Kilgallen@SpamCop.net> wrote: >> x@ >>> In article <BE76088B.AE0F%roktsci@comcast.net>, Jeff Cameron! >>> <roktsci@comcast.net> writes:> >>> O >>>> I can appreciate the meaning behind the nomenclature, but that is not whatpP >>>> is happening. I just got finished writing a test program in FORTRAN where IP >>>> used SYS$ASSIGN to open a file of 2048 blocks and used SYS$QIOW to read theO >>>> blocks, and when done on a disk with HFWM active all blocks were zero, but06 >>>> were not on the same disk when HFWM was disabled. >>> 9 >>> That is exactly the description of Highwater Marking.f@ >>> But it does not tell you whether those blocks were zeroed on' >>> disk before you tried to read them.a/ >>> And that does not mimic any useful program.oL >> Are you saying that the 100 or so times that the program is run with HFWMK >> enabled always yields all blocks as zeros vs. 100 or so runs of the sameiL >> program on the same disk with HFWM disabled always yields non zero blocks >> was just a coincidence? >  > No.  > L >> I was not attempting to "mimic" a useful program, nor measure overhead ofO >> HFWM. It does, however exhibit what HFWM does what the documentation says it:? >> does, that being blocks are zeroed out on allocation/extent.a > G > No, it does not and they are not.  You have only shown that the byteso( > were zero when you tried to read them. > G >>> For a test of of overhead from Highwater Marking, try the following 5 >>> both with Highwater Marking enabled and disabled.  >>>  >>> Create a new file*) >>> Write 100000 blocks of non-zero bytes, >>> Close the file& >>> Report the resources and time used >>> O >>> A different test would be to allocate the file to a large size on creation.  >> aL >> I performed this test too, which also validated my assertion. The time itK >> takes to open a large pre-allocated file takes significantly longer when . >> HFWM is enabled vs. when it is not enabled. > D > It should take no longer to open the file so long as EOF is at the > beginning of the file.H I don't know why you are fighting this. The documentation and HP support) disagree with you. Believe what you wish!    ------------------------------   Date: 5 Apr 2005 20:01:02 -0500 + From: young_r@encompasserve.org (Rob Young)c- Subject: Re: poor disk I/O performace on ds25b3 Message-ID: <hzfCNpxU$ZpK@eisner.encompasserve.org>r  a In article <42514F5C.2D1C8015@oracle.com>, Norman Lastovica <norman.lastovica@oracle.com> writes:r6 > COPY is just about the worst possible way to measure8 > performance in that it is basically a synchrous copy. 6 > (I'm told that there are plans afoot for VMS V8.3 to  > make copy really rip however). >   9 	Hmmmm... Hope the fastness is introduced via a COPY/FASTg5 	switch.  There could be unintended side-effects witht8 	long standing DCL routines suddenly causing performance
 	problems.   				Rob.   ------------------------------   Date: 5 Apr 2005 23:00:35 -0500d- From: Kilgallen@SpamCop.net (Larry Kilgallen)F- Subject: Re: poor disk I/O performace on ds25(3 Message-ID: <fV8wknhPRi2n@eisner.encompasserve.org>>  Z In article <BE787AA0.B0BE%roktsci@comcast.net>, Jeff Cameron <roktsci@comcast.net> writes:M > On 4/5/05 2:38 PM, in article cCdQbC$7d5Zy@eisner.encompasserve.org, "Larry-+ > Kilgallen" <Kilgallen@SpamCop.net> wrote:H    E >> It should take no longer to open the file so long as EOF is at the- >> beginning of the file. J > I don't know why you are fighting this. The documentation and HP support > disagree with you.  B Then please post a specific test that demonstrates your assertion.   ------------------------------   Date: 5 Apr 2005 21:32:25 +0100s6 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)8 Subject: Re: Procedure for installing PERL modules (LWP)* Message-ID: <425303e9@NEWS.LANGSTOEGER.AT>  } In article <craigberry-876A05.19385804042005@news.isp.giganews.com>, "Craig A. Berry" <craigberry@mac.com.spamfooler> writes: H >Note that Perl 5.6.1 is pretty long in the tooth now; 5.8.6 is current.  # Don't tell us, tell HPQ instead ;-)p  E And also tell them, MOZILLA is replaced by FIREFOX/THUNDERBIRD now...m   -- o Peter "EPLAN" LANGSTOEGERi% Network and OpenVMS system specialistt E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------   Date: 5 Apr 2005 18:07:07 -0700a From: mcbill20@yahoo.com8 Subject: Re: Procedure for installing PERL modules (LWP)C Message-ID: <1112749627.475702.165500@z14g2000cwz.googlegroups.com>a    Peter 'EPLAN' LANGSTOEGER wrote:F > In article <craigberry-876A05.19385804042005@news.isp.giganews.com>,8 "Craig A. Berry" <craigberry@mac.com.spamfooler> writes:A > >Note that Perl 5.6.1 is pretty long in the tooth now; 5.8.6 isu current. >c% > Don't tell us, tell HPQ instead ;-)  >rG > And also tell them, MOZILLA is replaced by FIREFOX/THUNDERBIRD now...a >t > -- > Peter "EPLAN" LANGSTOEGER ' > Network and OpenVMS system specialists > E-mail  peter@langstoeger.at@ > A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realisto  E Yeah, Peter took the words right out of my mouth. I installed the onewD from HP to make it a little easier and to hopefully work better with2 the Apache (or CSWS) that I also downloaded there.  F As you can see from some of my other recent posts, I have been playingA around with some open source software for VMS (Apache, PHP, PERL,hE Python, GDchart, GDlib, XML, etc.). However, since I don't have a tonoF of spare time, I find it frustrating to try to get much of it working.? I am a developer among other things, so I am used to debugging.cC However, I don't enjoy having to spend so much time debugging other F code that I didn't write just to get it to install. So often it seems,A when attempting to install this stuff on VMS, I find myself going  through:  ' 1. Find a link for the package for VMS.  2. Download it.- 3. Un-tar, unzip, etc.@ 4. Remove the trailing carriage-returns from .c, .h, .mms files.C 5. Try running MMK. When it fails, it could be sloppy code from the-E UNIX world that the DEC C compiler won't accept, or problems with the . name of the directory the install is in, or...G 6. Try to get help. Often there are version issues and with much of the5E software it's hard to find the pervious versions that will supposedlyo work with VMS.  ? And of course, with much of the software, there are quite a few:E prerequisites that also require the above steps. For example, I still7G haven't gotten GDchart or GDlib to work correctly. After all the buildsoF showed successful statuses, when I tried the test software, the imagesG were always invalid. Now, I'm certainly not saying that I didn't make aVF mistake (or more) as I see in this forum that others have gotten theseE packages working. But, when my time is limited, I find that it's manyhD time easier to just do it on LINUX (and I am a hard-core VMS person, not a LINUX person).  F I sure wish HP would spend a little more time keeping widely used open> source software up to date. I definitely enjoy programming andC debugging, but not spending hours and hours doing this every time IfD want to add any new software to VMS. For example, I'd much rather beD learning the intricacies of Apache, PERL, etc. than trying to figure  out why the build on VMS failed.   Just my .o2.  " Thanks again for your help though.   Bill   ------------------------------  % Date: Tue, 05 Apr 2005 22:32:25 -0500o6 From: "Craig A. Berry" <craigberry@mac.com.spamfooler>8 Subject: Re: Procedure for installing PERL modules (LWP)D Message-ID: <craigberry-D3A75C.22322505042005@news.isp.giganews.com>  C In article <1112749627.475702.165500@z14g2000cwz.googlegroups.com>,o  mcbill20@yahoo.com wrote:  " > Peter 'EPLAN' LANGSTOEGER wrote:H > > In article <craigberry-876A05.19385804042005@news.isp.giganews.com>,: > "Craig A. Berry" <craigberry@mac.com.spamfooler> writes:C > > >Note that Perl 5.6.1 is pretty long in the tooth now; 5.8.6 isU
 > current. > >e' > > Don't tell us, tell HPQ instead ;-)m  D HPQ is not the only source of Perl for OpenVMS, but I have recently G seen signs that HP personnel are once again working on Perl, including o8 upgrading it to take advantage of newer CRTL features.    I > > And also tell them, MOZILLA is replaced by FIREFOX/THUNDERBIRD now...   = Well, not exactly, but that is the direction for the Mozilla eE Foundation, and yes, life was better when Colin Blake was spending a c, lot of his time keeping the port up-to-date.  G > Yeah, Peter took the words right out of my mouth. I installed the oneoF > from HP to make it a little easier and to hopefully work better with4 > the Apache (or CSWS) that I also downloaded there.  G This depends on the definition of "work better."  If you need mod_perl eH for performance reasons, then the rather ancient HP version is the only B port I'm aware of.  If you just need Perl (which will do anything E mod_perl will do but with more image activations), there are options wH that are likely to work better for what you are trying to do.  I've got C a VMS-friendly build-from-source kit for Perl 5.8.4 where you just nD submit a batch job to build and install, and some binary kits built H from that kit.  These include some non-core extensions like LWP and its G dependencies as well as a number of VMS-specific extensions. Some nice  : person made sure these kits got on the latest freeware CD:  3 http://h71000.www7.hp.com/freeware/freeware70/perl/   C Send mail to vmsperl-subscribe AT perl DOT org to join discussions  8 about porting and using Perl and Perl extensions on VMS.  H > I sure wish HP would spend a little more time keeping widely used open > source software up to date.R  D I wish they would devote more resources, but I doubt "a little more H time" is really sufficient, at least for big, fast-moving projects like G Mozilla, Apache, or Perl.  It's very time-consuming to maintain a port uH of a code base that has hundreds of people on many platforms submitting A changes on a daily basis.  OpenVMS Engineering tends to port the eF current stable version and only submit its changes to the maintainers F well after the fact if at all.  This is understandable from the point G of view of short-term resource allocation, but dreadful from the point eB of view of giving the port a longer future.  If you don't develop H against the development version and take steps to get your changes into H the authoritative repository on a frequent basis, you may find yourself H having to do the port all over again with the next version (as they had G to do with Apache 2.0).  The work on Mozilla was a rare and refreshing rG exception to this tendency, but I haven't seen signs that this work is i ongoing.   ------------------------------  % Date: Tue, 05 Apr 2005 19:13:59 -0500r2 From: David J Dachtera <djesys.nospam@comcast.net>" Subject: Re: Remote printing woes.+ Message-ID: <425329C7.9B95379E@comcast.net>a   David Gray wrote:v > 	 > Hi all,e >  >         * OpenVMS 7.3-2nE >         * HP TCP/IP Services for OpenVMS Alpha Version V5.4 - ECO 1a% >         * Ricoh 1060 printer (PCL6)o > F > Getting the following error message when trying to print to a remoteG > printer via LPD.  The host system is in the UK and the remote printers > is in New Jersey.  > C > %TCPIP-E-LPD_REQREJECT, print request rejected by !AS (queue !AS)oE > %SYSTEM-F-CONNECFAIL, connect to network object timed-out or failedM > E > Both firewalls have telnet and also ports 9100,23 & 515 enabled.  I F > cannot verify whether the printer can be seen on the LAN here in theH > UK as the firewall administrators in the US cant or wont enabled ping./ > That said I can tracert without any problems.e > D > The printer queue is defined as follows and I also tried using the7 > TCPIP symbiont, getting no errors but also no prints.t >  > $ show queue  mercury1/fulleC > Server queue MERCURY1, idle, on TOM::"mercury1:515", mounted formt
 > WORD11$FORMsA >   /BASE_PRIORITY=4 /DEFAULT=(FEED,FORM=WORD11$FORM) /LIBRARY=HP 4 > Lowercase /OWNER=[SYSTEM] /PROCESSOR=TCPIP$LPD_SMB/ >   /PROTECTION=(S:M,O:D,G:R,W:S) /RETAIN=ERRORt > /SEPARATE=(RESET=(HP_RESET)) > * > Also tried using port 9100 - same error. >  > Any thoughts?r  G Well, I'm a Multinet guy, myself, so I can't speak to your queue setup.   B Instead, I'll refer you to the on-line doc.'s for TCP/IP Services.   Go to the OpenVMS page at HP:u http://www.hp.com/go/openvms  E ..., look under OpenVMS Information, and take the link to the OpenVMS"D Documentation page. There, you'll find a link to the TCP/IP ServicesF documentation (currently, http://h71000.www7.hp.com/doc/tcpip54.html -" subject to change without notice).  E By the way: can you Telnet to the printer? ...on port 515? ...on port- 9100?-  H Should you happen to have a system running Multinet, try this command to6 see if it will show you what ports it is listening on:   $ MULT SHOW/CONN/SNMP=a.b.c.dm  3 (insert your printer's IP address, as appropriate).:   -- e David J Dachtera dba DJE SystemsA http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page:>" http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/n   Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Tue, 05 Apr 2005 14:16:32 -0700d, From: Ken Fairfield <my.full.name@intel.com> Subject: Re: shadow minicopy+ Message-ID: <d2uv7h$4vk$1@news01.intel.com>s  / Phillip Helbig---remove CLOTHES to reply wrote: ; > In article <d2sah6$so8$1@news01.intel.com>, Ken Fairfield2# > <my.full.name@intel.com> writes: I  
 [big snip]  I >>     Finally, its recommended VAX nodes in a mixed-architecture clustergG >>have SHADOW_MAX_COPY set to 0 because they cannot have Write Bitmaps, F >>and if they "accidently" pick up the copy, it will be done as a full >>copy, not a mini-copy. >  >  > Glad you mentioned that. > H > I have VAXes and ALPHAs in the cluster.  I have shadow sets with both I > members on the same VAX, on two different VAXes, and on the same ALPHA. E > None on two different ALPHAs yet, but that will probably come soon.o > I > Which type of shadow set could benefit from MINIMERGE? From MINICOPY?   C > Is it clear from which node the DISMOUNT would have to be issued.a  C      Well the VAXes can't do minicopy at all.  The shadow sets thati@ have both members on a single VAX should not be a problem since,E presumably, when that VAX goes down, those shadow sets get dismountedlD cleanly and won't need a copy.   Get a less-than-clean dismount, and$ you'll need to merge (but not copy).  F      For the case where shadow set members are on different VAX nodes,C you may be able to do a Dismount/Policy=Minicopy _from an Alpha_ ofvD the shadow set member attached to the shutting-down VAX.  The manualF didn't list any restrictions on doing a minicopy of MSCP-served disks,F so the Alpha _should_ be able to do the minicopy when the VAX reboots.E For a full-copy, you'd want to avoid such a "remote shadow copy", butwB for a minicopy, there are relatively few disk blocks that actually/ need to be copied so it's really not a problem.a  E      For the shadow sets with both members attached to a given Alpha,mC it seems this is the same as shadow sets with both members attacheda? to the same VAX.  Get a clean dismount of the shadow set duringbD shutdown and you're set.  Get a less-than-clean dismount, and you'll need to merge.  D      VAXes can't do HBMM either so a full merge of shadow sets whereD both members are attached to the same VAX would be a sort of a worstB case.  If you install the HBMM kit on your Alphas, and if at leastE one Alpha stays up when another cluster member goes down, it _should_lB be able to do minimerges on the available shadow sets (if both/all@ members are still available, not for those where one member gets9 dropped when its host leaves the cluster; that's a copy).o  ?      Personally, I'd expect this to work, Alphas minimerging ortC minicopying shadow sets with members attached to VAXes, but I'm notsC sure VMS Engineering has tested it; we haven't).   Also, I may haveiF missed some key restriction in the docs (since we weren't "interested"B in VAXes during our testing).  I think it's well worth reading theC HBMM Release Notes and the Volume Shadowing manual, and then try it-* out if it's not explicitly disallowed. :-)  F      Note that HBMM (and minicopy, IIRC) requires your Alphas to be at. V7.3-1 or above, and your VAXes to be at V7.3.          -Kenr -- m6 I don't speak for Intel, Intel doesn't speak for me...  
 Ken Fairfieldm! D1C Automation VMS System Support " who:   kenneth dot h dot fairfield where: intel dot com   ------------------------------  # Date: Tue, 05 Apr 2005 21:52:36 GMT / From: brooks@cuebid.zko.dec.nospam (Rob Brooks)  Subject: Re: shadow minicopy- Message-ID: <l20926uoqoTf@cuebid.zko.dec.com>o  . Ken Fairfield <my.full.name@intel.com> writes: > A >      Personally, I'd expect this to work, Alphas minimerging or E > minicopying shadow sets with members attached to VAXes, but I'm not E > sure VMS Engineering has tested it; we haven't).   Also, I may havehH > missed some key restriction in the docs (since we weren't "interested"D > in VAXes during our testing).  I think it's well worth reading theE > HBMM Release Notes and the Volume Shadowing manual, and then try it , > out if it's not explicitly disallowed. :-)  G As long as the devices are not mounted on the hosting VAX (or any othere? VAX in the cluster), served devices can take advantage of HBMM.9  tH >      Note that HBMM (and minicopy, IIRC) requires your Alphas to be at0 > V7.3-1 or above, and your VAXes to be at V7.3.  K HBMM is first supported on OpenVMS Alpha V7.3-2, with the appropriate patchoK kits.  Minicopy first appeared on OpenVMS Alpha V7.3, and was backported tooN V7.2-2 (which was released AFTER V7.3).  VAXes running OpenVMS V7.3 can act asO minicopy "clients" in that their writes may be tracked, but a VAX cannot host ai minicopy bitmap.  8 HBMM will not work on a device that is mounted on a VAX.   -- t  M Rob Brooks    VMS Engineering -- I/O Exec Group     brooks!cuebid.zko.dec.come   ------------------------------  % Date: Tue, 05 Apr 2005 19:58:46 +0100 # From: issinoho <issinoho@gmail.com>  Subject: Site down?s; Message-ID: <1112727527.3556.0@spandrell.news.uk.clara.net>o   Anyone know what's happened to C. http://www.pi-net.dyndns.org/anonymous/kits/ ?  I I've been trying to get the latest version of MySQL for several days now.    ------------------------------   Date: 5 Apr 2005 11:28:55 -0700e From: tadamsmar@yahoo.comr" Subject: Re: Time change questionsC Message-ID: <1112725735.230292.165410@o13g2000cwo.googlegroups.com>r   Hoff Hoffman wrote:o3 > In article <hUf4e.2913$Jr7.468@news.cpqcorp.net>,p5 hammond@not@peek.ssr.hp.com (Charlie Hammond) writes:-F > :In article <1112631490.299361.264790@z14g2000cwz.googlegroups.com>, > : tadamsmar@yahoo.com writes:e >fG > :>1.  The FAQ seems to imply that I would have to shutdown NTP during  thet > :>time change. Why is that?e >rE >   Depending on the revision of NTP, it may not contend happily withxF >   the change -- current revisions do, I'm told.   The cut-over wouldE >   have been around OpenVMS V7.3.  (Piles of time-keeping got betternE >   with V7.3, as that was a major overhaul in OpenVMS time-keeping.)r >eF >   The FAQ is fairly conservative here, as there can be problems withG >   processes -- NTP prefers to drift the time, and that is not how thewD >   folks that decided on this mess based on an idea from an EnglishE >   Golfer (William Willet) who reputedly wanted to get more time forsF >   his evening golf game (I'm not kidding) or derived from an idea of? >   Frankin's, or whatever, daylight change-over is intended to  operate.? >   It's supposed to be one big one-hour ker-chunk, and not ther drifting# >   that NTP would want to provide.e  F Back in the 1980s I led a team that created a big VMS application that ran withoutaG failure for 6 months in production.  Then they decided to put some sorteF of controlling clock on the system that set the time.  The application$ crashed due to time going backwards.  E I did some investigation and found that it is common for non-computer ; systems to avoid having time go backwards.  The trains play F catch-up in the spring and wait at the station for an hour in the fall to get back on schedule.  F The have a stack of pennies on Big Ben's pendulum.  They add or remove pennies to adjust the rate.E   >i? > :>I never had to do that before since the system time and thel& > :>tdf were changed at the same time. >a' >   Congratulations, it worked for you.  > E > :>2.  The system doc implies that I should use sys$set_system_evento toC > :>look for sysevt$c_tdf_change event to synchronize with the time  change.c > :>Is that correct? > @ >   If your application wants to know when the TDF changed, yes. >nG >   I tend to prefer to use UTC or GUIDs/UUIDs or other such operationsoD >   or values for unique identities, and relative timers and such --A >   relative timers get adjusted automatically, and UTC obviouslyN avoids! >   the "fun" of the change-over..  @ I don't think I want to do change my application to go to UTC or relativeD timers.  Too much work and/or might lose reliability without lots of testing.   > G > :>Is there an easy way to have a command procedure to run just before/ theg > :>time change? > D >   Easy?  Not that I am aware of.  There are certainly some ways toF >   do this, but I'm not aware of any TDF-is-about-to-change event norE >   any customization procedure.  (This would have to be run some set1F >   time prior to the change, to allow the operation to complete -- itD >   would have to be a TDF-will-change-in-delta event or task or...)D >   The ways involve some brute-force coding -- the easiest of theseF >   is a SUBMIT/AFTER=six-months-or-so (ugly, but functional -- and itE >   does not take into account any decreed "unscheduled" TDF changes)u  ! What are unscheduled TDF changes?T   >   orE >   a periodic job that watched for the approaching switch-over based  on >   the TZ rules.y >! > 2 >  ---------------------------- #include <rtfaq.h> -----------------------------e3 >     For additional, please see the OpenVMS FAQ --b www.hp.com/go/openvms/faq.4 >  --------------------------- pure personal opinion ---------------------------mG >         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.comP   ------------------------------  # Date: Tue, 05 Apr 2005 19:07:19 GMT # From: hoff@hp.nospam (Hoff Hoffman)m" Subject: Re: Time change questions0 Message-ID: <HlB4e.3019$HW.752@news.cpqcorp.net>  _ In article <1112725735.230292.165410@o13g2000cwo.googlegroups.com>, tadamsmar@yahoo.com writes:i" :What are unscheduled TDF changes?  F   Those dictated by changes in applicable regulations, by the physicalE   relocation of the system, by problems with the clock hardware -- anrF   ancient and thankfully now rare trigger was a power-line based clockE   synchronization scheme used by some organizations and by some clock H   hardware; by high-frequency signals injected onto the local power gridG   to cause the local clocks to reset the displayed time -- or those TDFeJ   changes triggered by erroneous or otherwise "creative" system management   input.  N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faqoN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.coms   ------------------------------  # Date: Tue, 05 Apr 2005 20:22:48 GMTc3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond)f" Subject: Re: Time change questions1 Message-ID: <ssC4e.3028$r%.1746@news.cpqcorp.net>2  D In article <1112725735.230292.165410@o13g2000cwo.googlegroups.com>,  tadamsmar@yahoo.com writes:e   ..C >           ... trains play catch-up in the spring and wait at the  9 >station for an hour in the fall to get back on schedule.i  7 (1) This statement is biased to the northern hemisphere   D (2) In the specific case of Amtrack, the trains always play catch-up0 and never succeed.  O.K., _ALMOST_ never...  ;-)     -- eJ       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  % Date: Tue, 05 Apr 2005 13:51:07 -0700i, From: Ken Fairfield <my.full.name@intel.com>" Subject: Re: Time change questions+ Message-ID: <d2utns$4a6$1@news01.intel.com>   3 Replying to Davids's post only because the originals hasn't shown up on my server...a   David J Dachtera wrote:  > tadamsmar@yahoo.com wrote: > 4 >>(My springforward.com failed, apparently due to my1 >>earlier upgrade to 7.3.2. The command procedure , >>sys$manager:utc$configure_tdf is obsolete. >>. >>I disabled ntp and just used SET TIME to get >>the times close for now.)m >> >>What is the replacement for: >>( >>@sts$manager:utc$configure_tdf -240 60   Presumably you meant:   ,   @sys$manager:utc$configure_tdf SET -240 60$                                  ^^^  ?      I just wanted to note that, while Charlie Hammond has gonee9 to great lengths to discourage use of this procedure,thise; procedure still operates as expected if you comment-out thee$ "EXIT 2" following his warning text.  =      We still use this procedure for "business reasons".  Ouro? requirements are (1) to be able to disable time synchronizationi> (via DTSS in our case) prior to the actual time change, (2) to: avoid having the clock jump following the time change when@ we restart DTSS, and (3) to be able to accomplish this via batch: jobs (no interactive queries).  UTC$CONFIGURE_TDF.COM does@ this for us (the step to reset the TDF so the clock doesn't jump< when DTSS is restarted) and at the time it was obsoleted, we= didn't have a good alternative.  (There's no way I'm going toE> run UTC$TIME_SETUP.COM interactively on 8 separate clusters in the wee hours of the morning!)  7      It may very well be that David's suggestion to useU1 Sys$Startup:TDF$UTC_STARTUP.COM after deassigning.: SYS$TIMEZONE_RULE accomplishes what we need, but I have no: experience with it and would need to do exhaustive testing@ before using it for the time change in a production environment.  	      -Ken  -- o6 I don't speak for Intel, Intel doesn't speak for me...  
 Ken Fairfield ! D1C Automation VMS System Supporti" who:   kenneth dot h dot fairfield where: intel dot com   ------------------------------  % Date: Tue, 05 Apr 2005 19:01:12 -0500I2 From: David J Dachtera <djesys.nospam@comcast.net>" Subject: Re: Time change questions) Message-ID: <425326C8.854371@comcast.net>i   Ken Fairfield wrote: > [snip]9 >      It may very well be that David's suggestion to usen3 > Sys$Startup:TDF$UTC_STARTUP.COM after deassigning.< > SYS$TIMEZONE_RULE accomplishes what we need, but I have no< > experience with it and would need to do exhaustive testingB > before using it for the time change in a production environment.  D There appears to be a "gotcha" with this regarding Cerner stuff. Not2 sure yet what the issue was, but am investigating.   -- o David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page:a" http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/0  " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/m   Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------   Date: 5 Apr 2005 16:40:52 -0500h- From: Kilgallen@SpamCop.net (Larry Kilgallen) " Subject: Re: VMS Command language.3 Message-ID: <BnP1wqzVJqEu@eisner.encompasserve.org>n  c In article <4252406E.12270.42684CB@localhost>, "Stanley F. Quayle" <squayle@insight.rr.com> writes:a+ > On 4 Apr 2005 at 17:27, Tom Linden wrote: ) >> Was there ever any alternative to DCL?  > F > The GNV tools include the bash shell.  It's not set up to be a CLI, C > but it's easy to write a short LOGIN.COM that drops them into it   > without any way out...  @ But that does not provide the essential quality of a VMS commandA language interpreter - maintaining context in the same process as A the programs it runs.  It makes the shell run more like shells doo5 on Unix, which is probably good enough for Unix fans.    ------------------------------    Date: 06 Apr 2005 00:19:33 +0200( From: Andreas Davour <ante@update.uu.se>" Subject: Re: VMS Command language.4 Message-ID: <cs94qek6f4a.fsf@Psilocybe.Update.UU.SE>  / Kilgallen@SpamCop.net (Larry Kilgallen) writes:s  e > In article <4252406E.12270.42684CB@localhost>, "Stanley F. Quayle" <squayle@insight.rr.com> writes:e- > > On 4 Apr 2005 at 17:27, Tom Linden wrote: + > >> Was there ever any alternative to DCL?w > > H > > The GNV tools include the bash shell.  It's not set up to be a CLI, E > > but it's easy to write a short LOGIN.COM that drops them into it l > > without any way out... > B > But that does not provide the essential quality of a VMS commandC > language interpreter - maintaining context in the same process aseC > the programs it runs.  It makes the shell run more like shells dos7 > on Unix, which is probably good enough for Unix fans.   D For some kinds of perverted unix fans there have been a port of tcsh( around as well. I have used it, shortly.   /andreas   -- eA A: Because it fouls the order in which people normally read text. ' Q: Why is top-posting such a bad thing?i A: Top-posting.e; Q: What is the most annoying thing on usenet and in e-mail?n   ------------------------------  % Date: Tue, 05 Apr 2005 19:30:57 -0400<( From: Bill Todd <billtodd@metrocast.net>( Subject: VMS:  nothing else can compare?= Message-ID: <PKWdnYI6L-cvgs7fRVn-gg@metrocastcablevision.com>r  E Well, one might have been tempted to say something similar about the lB Tandem NonStop architecture - at least I would have been, since I H respect it immensely.  But HP appears to have persuaded a customer that H open-source software running on 'industry-standard' hardware is just as  good:e  W http://www.computerworld.com/developmenttopics/development/story/0,10801,100797,00.htmlS  ? "The total cost of ownership didn't warrant keeping it on that  B [proprietary] platform when the same availability and uptime were ( available with open-source technologies"  F Now, you know and I know that that latter phrase in this statement is A simply not true, but HP certainly seems to have suggested to its dD customer that it was (though for an admittedly non-mission-critical 
 application).   H In skimming the article, I just assumed that this was yet another minor H HP Itanic success story.  But had I read it more carefully I would have @ found nary an Itanic in sight, as I eventually learned from the  accompanying sidebar (  f http://www.computerworld.com/softwaretopics/os/story/0,10801,100799,00.html?source=NLT_LIN&nid=100799  ).  > So the idea that HP is completely relying upon Itanic for its J back-room/back-end systems may be starting to fray a bit around the edges.   - bill   ------------------------------   Date: 5 Apr 2005 19:57:35 -0500e+ From: young_r@encompasserve.org (Rob Young)a, Subject: Re: VMS:  nothing else can compare?3 Message-ID: <vSKqpHAYwKda@eisner.encompasserve.org>t  h In article <PKWdnYI6L-cvgs7fRVn-gg@metrocastcablevision.com>, Bill Todd <billtodd@metrocast.net> writes:  G > Well, one might have been tempted to say something similar about the sD > Tandem NonStop architecture - at least I would have been, since I J > respect it immensely.  But HP appears to have persuaded a customer that J > open-source software running on 'industry-standard' hardware is just as  > good:o > Y > http://www.computerworld.com/developmenttopics/development/story/0,10801,100797,00.html- > A > "The total cost of ownership didn't warrant keeping it on that  D > [proprietary] platform when the same availability and uptime were * > available with open-source technologies" > H > Now, you know and I know that that latter phrase in this statement is C > simply not true, but HP certainly seems to have suggested to its  F > customer that it was (though for an admittedly non-mission-critical  > application).t >    	"Skimming" further...  Y http://www.computerworld.com/developmenttopics/development/story/0,10801,100797p5,00.html   N  Continental and EDS are also in the process of partitioning the TPF mainframeL and moving some subsystems to a distributed environment of cheaper commodityL Linux servers, McDonald says. The subsystems include pricing, scheduling andL seat inventory. "The TPF systems have been taxed out so much in the last fewO years that we're running out of capacity on some of those mainframes," McDonald E says. "You have to move some of that off [the mainframe], or at leaste distribute it."F  : 	Cheaper , faster.  Good business to move off those costly 	old mainframes.  G Money-generating transactions, such as booking and ticketing, remain oni( Continental's tried-and-true mainframes.   	Makes sense for now.   I Even though the more stable 2.6 Linux kernel is now supported by the most I popular commercial Linux distributions from Red Hat and Novell Inc., manyoN companies remain cautious about migrating important systems to the open-source operating system.>  I Gary Hein, an analyst at Burton Group, says clients tell him, "What's theaK motivation for me to put my neck on the line, when Oracle on Solaris is thehL core of my business and it functions just perfectly?" But he also finds thatK once users have a good experience, they're more inclined to take the plungeu again.  O "Success breeds success," Hein says. "They're hesitant to do the first one. Bute9 after they do, they say, 'Wow. That's hard not to do.' "    D 	Right.  Early adopters - go ahead.  The "collective we" will followB 	after we study your mistakes/disasters.  Good business practices.  @ 	As we see above, some of these functions make total sense.  Two; 	segments:  it can be down infrequently , it can't be down rD 	(or shouldn't).  Some of those make Linux sense, and the mainframe B 	won't go away any time soon.  The mainframe is just getting more : 	narrowly defined in business.  Filling a niche.  For now.   				Rob    ------------------------------  % Date: Tue, 05 Apr 2005 22:58:26 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com>w, Subject: Re: VMS:  nothing else can compare?B Message-ID: <1112756303.ce7aff60436ecab1ada9b067ccaf2e4f@teranews>   Rob Young wrote: >         "Skimming" further...0C >         Cheaper , faster.  Good business to move off those costly  >         old mainframes.b    E Skimming ever further, that article reveals that the really importantF$ stuff is still on the IBM mainframe.  D And they also reveal that the Tandem solution would have been reallyH good, but someone decided it was "overkill" and that cheap systems couldG do the job. And this is very important for VMS. If you ]position VMS aseA high end "big iron" stuff, you shut it out of many many potentialtF applications/custoemrs for those same reasons.  HP needs to market VMSF agressively for all markets so that VMS doesn't get typecasted for big? iron only, a bit like Tandem is typecasted for rolls-royce onlyp' application (where money is no object).p  E VMS' clustering is anothert example of overkill. For vast majority of2E POTENTIAL VMS customers, its clustering capabilities are overkill and<D they will not pay any premium for clustring capabilities they do not really need.   ------------------------------  % Date: Tue, 05 Apr 2005 23:05:32 -0400 ( From: Bill Todd <billtodd@metrocast.net>, Subject: Re: VMS:  nothing else can compare?= Message-ID: <nqKdncrQ2bthzM7fRVn-vg@metrocastcablevision.com>    Rob Young wrote:j > In article <PKWdnYI6L-cvgs7fRVn-gg@metrocastcablevision.com>, Bill Todd <billtodd@metrocast.net> writes: >  > G >>Well, one might have been tempted to say something similar about the  D >>Tandem NonStop architecture - at least I would have been, since I J >>respect it immensely.  But HP appears to have persuaded a customer that J >>open-source software running on 'industry-standard' hardware is just as  >>good:n >>Y >>http://www.computerworld.com/developmenttopics/development/story/0,10801,100797,00.htmlr >>A >>"The total cost of ownership didn't warrant keeping it on that lD >>[proprietary] platform when the same availability and uptime were * >>available with open-source technologies" >>H >>Now, you know and I know that that latter phrase in this statement is C >>simply not true, but HP certainly seems to have suggested to its sF >>customer that it was (though for an admittedly non-mission-critical  >>application).r >> >  >  > 	"Skimming" further...  H If you just plowed on with the rest of the article, you missed the main E reason why I posted this:  it was the assertion above that "the same :F availability and uptime [which NonStop systems provide are] available  with open-source technologies".c  G Not only is that assertion simply wrong (and potentially dangerous for tF anyone believing it, as HP seems to have encouraged them to do here), H but it's pure death to any attempt to revive VMS, because the very real G differences in such areas are about the only remaining reasons to care s
 about VMS.  I If HP is telling customers this, it says quite a lot about how important  < they believe their existing, mature enterprise products are.  ; Can you say "affinity all over again"?  I knew you could...l   - bill   ------------------------------   Date: 6 Apr 2005 00:33:57 -0500o+ From: young_r@encompasserve.org (Rob Young)e, Subject: Re: VMS:  nothing else can compare?3 Message-ID: <veimigFOH8Rp@eisner.encompasserve.org>h  h In article <nqKdncrQ2bthzM7fRVn-vg@metrocastcablevision.com>, Bill Todd <billtodd@metrocast.net> writes: > Rob Young wrote:k >> In article <PKWdnYI6L-cvgs7fRVn-gg@metrocastcablevision.com>, Bill Todd <billtodd@metrocast.net> writes:d >> e >> hH >>>Well, one might have been tempted to say something similar about the E >>>Tandem NonStop architecture - at least I would have been, since I tK >>>respect it immensely.  But HP appears to have persuaded a customer that cK >>>open-source software running on 'industry-standard' hardware is just as i >>>good: >>>uZ >>>http://www.computerworld.com/developmenttopics/development/story/0,10801,100797,00.html >>>-B >>>"The total cost of ownership didn't warrant keeping it on that E >>>[proprietary] platform when the same availability and uptime were p+ >>>available with open-source technologies"h >>>hI >>>Now, you know and I know that that latter phrase in this statement is  D >>>simply not true, but HP certainly seems to have suggested to its G >>>customer that it was (though for an admittedly non-mission-critical e >>>application). >>>t >> c >> o >> 	"Skimming" further...A > J > If you just plowed on with the rest of the article, you missed the main G > reason why I posted this:  it was the assertion above that "the same iH > availability and uptime [which NonStop systems provide are] available ! > with open-source technologies".w  , 	Of course it is.  Just ask eBay and others.   > I > Not only is that assertion simply wrong (and potentially dangerous for  H > anyone believing it, as HP seems to have encouraged them to do here), J > but it's pure death to any attempt to revive VMS, because the very real I > differences in such areas are about the only remaining reasons to care   > about VMS. >   ? 	But in large infrastructures it is not uncommon to have 2 or 3s? 	tiers - as you know.  The first tier is a layer of Web serversn 	typically running IIS.  .  3 http://www.eweek.com/article2/0,1759,1640234,00.asp-  L Job One was to eliminate the single point of failure in the huge, monolithic system.   > 	[that sun cluster that was dying all the time and sooooo very? 	much comp.os.vms fun!!  Ah Andy we miss ya!  Hey... maybe this06 	will get him back:  Zinc Whiskers!  One could hope! ]  C Then, reconstruction of the application to ensure that we had fault:O isolation and that processes and tasks of like size and cost weren't congesting4I and competing with each other. We disaggregated the monolithic system andf" ensured scale and fault tolerance.    . So you went from how many servers to how many?  H We went from one huge back-end system and four or five very large searchN databases. Search used to update in 6 to 12 hours from the time frame in whichN someone would place a bid or an item for sale. Today, updates are usually lessJ than 90 seconds. The front end in October '99 was a two-tiered system withI [Microsoft Corp.] IIS [Internet Information Services] and ISAPI [InternetfO Server API]. The front ends were about 60 [Windows] NT servers. Fast-forward toiL today. We have 200 back-end databases, all of them in the 6- to 12-processorH range, as opposed to having tens of processors before. Not all those areO necessary to run the site. We have that many for disaster recovery purposes andl for data replication.   & What hardware platform are you on now?  M We use Sun [Microsystems Inc.] systems, as we did before. We use Hitachi DataHN Systems [Corp.] storage on Brocade [Communications Systems Inc.] SANs [storageN area networks] running Oracle [Corp.] databases and partner with Microsoft forN the [Web server] operating system. IBM provides front and middle tiers, and weO use WebSphere as the application server running our J2EE code.the stuff that ishK eBay. The code is also migrated from C++ to Java, for the most part. Eighty 4 percent of the site runs with Java within WebSphere.  O Would it make more sense to go with a commodity Intel [Corp.]-architecture back  end?  N We're in a continuous state of re-evaluation, and we're not afraid to swap outM where necessary. With the help of Sun people, we've tuned the applications tosM take advantage of the benefits of their systems. There would be a bit of worksJ to change, but should the time come where we believe there's a significant* benefit, we would probably make that move.  K Sun's recent strategy, with their alliance with AMD [Advanced Micro DevicesaL Inc.], shows that they are willing to move toward the commodity space. We'reK actually running a pilot of that architecture, AMD-based Sun servers with ae Linux variant.    B 	You see, they are piloting Linux on x86 and will throw Sun a bone 	and buy their AMD servers.  i  < 		1)  They are evaluating dumping Sun "proprietary" hardware1 		2)  Going with industry standard hardware (x86) ) 		3)  Going with free/low-cost OS - Linuxc  ? 	Where does the uptime come from?  RAC.  RAC on Linux clusters.l< 	Doesn't matter if the nodes drop like flies.  You can thank< 	the VMS folks that passed it to the Tru64 folks who (insert? 	best crazy conspiracy theory here) passed it to Larry who toldiD 	the troops to stuff it in Oracle.  Real Clustering.  Was that under9 	GP Bob's watch?  I forget.  Point is, earlier you state:   I > If you just plowed on with the rest of the article, you missed the mainwF > reason why I posted this:  it was the assertion above that "the sameG > availability and uptime [which NonStop systems provide are] available>! > with open-source technologies"..  C 	Of course that's true.  You would have someone not use RAC, right? = 	Or some other brain dead back-end to support your assertion?P  C 	Bet the business strategy.  You know eBay will do it.  It makes a   	great deal of sense.   K > If HP is telling customers this, it says quite a lot about how important n> > they believe their existing, mature enterprise products are. > = > Can you say "affinity all over again"?  I knew you could...s  > 	No.  You are missing the point.  There are countless examples> 	just like eBay's.  Why do you think Sun's stock is a dog?  It@ 	isn't as if they are burning up the industry like they used to.F 	Folks are ditching their high-priced - weak kit.  Sun's hope is that > 	people don't ditch their proprietary Unix at this point.  The> 	hand writing is on the wall for UltraSparky.  It isn't Itanic 	it is Ultanic (R) !!w   				Robo    B Men with walkie-talkie                  I'm home again to you babeC Men with flashlights waving             You know it makes me wonder,G Up upon the tower                       Sittin' in the quiet slipstreamY> The clock reads daylight savin'         Rollin' in the thunder  /                                  -- Neil Young T   ------------------------------   Date: 5 Apr 2005 11:43:31 -0700  From: tadamsmar@yahoo.com K Subject: Which release notes say sts$manager:utc$configure_tdf  is obsolete C Message-ID: <1112726611.257267.284020@z14g2000cwz.googlegroups.com>t  D I have a bad experience on Monday morning around 7:30.  My major appC failed on all systems.  The time had not been adjusted.   The Motif,F clock was reading one hour later than the system clock, something thatE I had never seen.  People were trying to prepare for experiments that0 started as early as 9.  E This was all due to utc$configure_tdf being obsolete after my upgradet	 to 7.3.2.n  B I am trying to do some root cause analysis/defect prevention on my
 failure here.g  F I cannot find anything about this in the 7.3.2 release notes or in any? of the online release notes for the layered products or patchesi
 that I added.s  # I went from 7.2.something to 7.3.2.a  @ I think that the root cause may be that I did not review all theB release notes of all the operating system versions between the old/ version and the new version that I upgraded to.w   Correct?  A Also, when you run utc$configure_tdf on 7.3.2 the initial messageoB sort of implies that it should run ok but do only part of the timeG change steps. However, it gives an error an fails completely.  I wonder0 what is going on with this?u  C And, I could have confirmed that my SPRINGFORWARD.COM had worked byvE dialing in on Sunday.  It worked last year, so I had a false sense ofe	 security.t   ------------------------------  # Date: Tue, 05 Apr 2005 19:14:17 GMTe# From: hoff@hp.nospam (Hoff Hoffman) O Subject: Re: Which release notes say sts$manager:utc$configure_tdf  is obsoleteA1 Message-ID: <dsB4e.3021$HW.2153@news.cpqcorp.net>e  _ In article <1112726611.257267.284020@z14g2000cwz.googlegroups.com>, tadamsmar@yahoo.com writes:iF :This was all due to utc$configure_tdf being obsolete after my upgrade
 :to 7.3.2.  G   Use of this or other tools has been recommended against for some timei:   now -- please see the OpenVMS FAQ for related materials.  K   I do try to keep the OpenVMS FAQ current with what I know about time and sJ   timekeeping as related to OpenVMS, as twice a year this newsgroup sees aK   barrage of various DST-related problems and problem reports.   (These duemI   to bugs in OpenVMS, due to operational problems, due to lack of currenthH   ECOs, due to changes within the timezone rules, and/or due to invalid K   assumptions made by any of the various programmers and/or system managersc   involved.)    N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faqiN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.comP   ------------------------------  # Date: Tue, 05 Apr 2005 20:39:01 GMT 3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond)cO Subject: Re: Which release notes say sts$manager:utc$configure_tdf  is obsoletec0 Message-ID: <FHC4e.3031$r%.854@news.cpqcorp.net>  D In article <1112726611.257267.284020@z14g2000cwz.googlegroups.com>,  tadamsmar@yahoo.com writes:t  $ >I went from 7.2.something to 7.3.2.  F I think you may find release notes / new features information relativeJ to this issue in the V7.3 documentation.  That is when the AUTO_DLIGHT_SAVH system parameter was introduced.  Various changes were made to make this work.   F (Section 4.2, Automatic setting of Daylight Savings Time, in the V7.3 E New Features manual points to the OpenVMS System Manager's Manual fora more information.)  A Note that as far back as V7.1 we (then Compaq) recommended using  A UTC$TIME_SETUP with the proper params, rather than UTC$TIME_SETUPt and/or UTC$CONFIGURE_TDF.e      -- mJ       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  % Date: Tue, 05 Apr 2005 17:33:34 -0400t- From: JF Mezei <jfmezei.spamnot@teksavvy.com>eO Subject: Re: Which release notes say sts$manager:utc$configure_tdf  is obsoleteaB Message-ID: <1112736805.cef0188be9903805e8f7dac63b33a56d@teranews>   Hoff Hoffman wrote:lL >   timekeeping as related to OpenVMS, as twice a year this newsgroup sees aA >   barrage of various DST-related problems and problem reports. e  F Interestingly, this year, the questions appeared well after the event. Usually they happen before !  D It seems to me that VMS was quite late in implementing automated and@ systematic tize zone and tiem change handling and it was done inB hodgepodge way. DECNET-5 seemd to have soemthing in it that did itD automatically. Now, NTP seems broken and can't do it and needs to be% shutdown while somethng else does it.>  H It would have been nice if there had been proper plannin from day 1 withF one central clock keeping set of functions/database/logicals, and thatE any other application, such as NTP would have been adapted to work onhL that and rely on that not changing because it was done right from the start.  H And the system routines such as $GETIM could have had an extra parameterA SYS$TIME_LOCAL, SYS$TIME_GMT and SYS$TIME_OFFSET which would have E provided applications with one standard and consistent way to gettinga the time they wanted.o  H When I coded message router gateway to SMTP, and from POP, I had to haveG my own GMT offsets because those were not available at the time on VMS.oB So twice a year, I need to update the configuration files on thoseG applications and because I don't really know what is "stable" on VMS in E terms of timekeeping that I could use (and also how to trigger a timea  change event in an application).   ------------------------------   End of INFO-VAX 2005.191 ************************