1 INFO-VAX	Tue, 12 Oct 2004	Volume 2004 : Issue 566       Contents: Re: Disabling source routing Re: EVA vs. bunch of MSA?  Hardware problems??  Re: Hardware problems??  HP ignores OpenVMS & SWIFT Re: HP ignores OpenVMS & SWIFT Re: HP ignores OpenVMS & SWIFTP RE: Obsucure Inquier article about Intel mentions VMS (and not any other  OS') oP RE: Obsucure Inquier article about Intel mentions VMS (and not any other  OS') oP Re: Obsucure Inquier article about Intel mentions VMS (and not any other  OS') o Re: Rare behaviour xabrdt  Re: Sayonara storage Re: Sayonara storage Re: Sayonara storage Re: Sayonara storage Re: Sayonara storage# Re: SHOW DEVICE/FILES with FORTRAN? # Re: SHOW DEVICE/FILES with FORTRAN? # Re: SHOW DEVICE/FILES with FORTRAN? # Re: SHOW DEVICE/FILES with FORTRAN? # Re: SHOW DEVICE/FILES with FORTRAN? # Re: SHOW DEVICE/FILES with FORTRAN? # Re: SHOW DEVICE/FILES with FORTRAN? # Re: SHOW DEVICE/FILES with FORTRAN? # Re: SHOW DEVICE/FILES with FORTRAN? 0 Re: SYS$PARSE: prevented translation of logicalsB The H-P Compaq Merger Two Years Out:  Still Waiting for the Upside1 Wonder how this would work on Alpha/VMS clusters? 5 Re: Wonder how this would work on Alpha/VMS clusters? 5 RE: Wonder how this would work on Alpha/VMS clusters? 2 Re: [OT]: Software Disasters Often People Problems2 Re: [OT]: Software Disasters Often People Problems2 Re: [OT]: Software Disasters Often People Problems2 Re: [OT]: Software Disasters Often People Problems2 Re: [OT]: Software Disasters Often People Problems2 Re: [OT]: Software Disasters Often People Problems  F ----------------------------------------------------------------------  + Date: Mon, 11 Oct 2004 22:43:47 +0000 (UTC) ? From: Graham Burley <burley.not-this@encompasserve-or-this.org> % Subject: Re: Disabling source routing 9 Message-ID: <416B0C7F.67E838A1@encompasserve-or-this.org>    stitzman wrote:  >   E > Anyway, is there anywhere else to find information about this?  I'm D > trying to follow a third-party security recommendation that sourceA > routing be turned off, but I don't have the original assessment G > results, so I don't know how they even discovered that source routing ! > was a problem on these servers.   < If you can find docs on the sysconfig utility it might help.   Could it be this?   ' $ mc tcpip$sysconfig -q inet ipsrcroute  inet:  ipsrcroute = 1  3 Hopefully one of the TCPIP engineers will chime in.    ------------------------------  % Date: Tue, 12 Oct 2004 10:16:23 +0400 & From: Valentin Likoum <me@privacy.net>" Subject: Re: EVA vs. bunch of MSA?+ Message-ID: <2t17l8F1pqa41U1@uni-berlin.de>    Joel Loveless wrote:F > On Fri, 08 Oct 2004 16:39:16 +0400, Valentin Likoum <me@privacy.net> > wrote: >  > @ >>  What else functionality, except snapshot, has impressed you? >  > C > Monitoring of systems has always been important to every system I @ > worked with. I am an old school admin with DEC Console ManagerF > experience. The Management appliance allows SNMP trap forwarding for@ > event notification services. My last company used Netcool fromF > Micromuse as thier corporate event notification services. I was ableD > to setup the SNMP traps to be forwarded and displayed for operatorF > respose. They include a template file on the management appliance toC > help format this functionality. The appliance also supports email G > notification services. All notification services can be configured to D > filter at the appliance as well. I am not aware of what monitoring, > capabilities are available on the MSA1500. [snip]  :    Thank you and everyone who reponded. It's a pity after < last bad news from HP storage division my research seems to  become pure academic.    --  
 Best regards, 
   Valentin)   valentin.likoum at ncc dot volga dot ru    ------------------------------   Date: 12 Oct 2004 02:34:40 GMT) From: tomarsin2015@aol.com (Tomarsin2015)  Subject: Hardware problems??: Message-ID: <20041011223440.24357.00001753@mb-m16.aol.com>   Hello  We have the following config Alpha 8400 5/350 Alpha 4100 5/466 VMS 6.2 I We just added some RZ1ED-VW drives connected  to the Alphas using HSJ50s. N The drives are set up as shadow sets. Before we added the new drives when ever2 you would do show one of the shadow sets, it wouldN list both HSJs as host name and alternate host name. How that we added the new4 10K drives under the alternate host name is the 4100  O Disk $1$DUA246:, device type MSCP served SCSI disk, is online, member of shadow M     set DSA244:, served to cluster via MSCP Server, error logging is enabled.   O     Error count                    0    Shadow member operation count   1129412 O     Host name                "HSJ21"    Host type, avail              HSJ5, yes O     Alternate host name        "ZIA"    Alt. type, avail AlphaServer 4100 5/466  4MB, yes$     Allocation class               1  O Disk $1$DUA444:, device type MSCP served SCSI disk, is online, member of shadow M     set DSA244:, served to cluster via MSCP Server, error logging is enabled.   O     Error count                    0    Shadow member operation count   4539254 O     Host name                "HSJ07"    Host type, avail              HSJ5, yes O     Alternate host name      "HSJ18"    Alt. type, avail              HSJ5,  no $     Allocation class               1  A This has also cost some performance problems. I was wondering if  L 1. The new 10k drives are seeing the 4100 (KZPAC-CA) controllers as a better performance path. K 2. Is there anyway of getting the HSJs to take control enstead of the 4100.   I In a way I feel that the system is doing what it should be doing, the new G drives see the KZPAC-CA offering better I/O then the HSJ, so the drives J are connecting to the controllers on the 4100. But management is screaming6 about performance issues and how long it takes to do aL shadow copy. I should note that HP has looked into the issue and they cannot/ find any problems. If you need more info please $ e-mail me, because we are at a lost. Thanks phillip    ------------------------------  + Date: Tue, 12 Oct 2004 04:42:15 +0000 (UTC) 7 From: moroney@world.std.spaamtrap.com (Michael Moroney)   Subject: Re: Hardware problems??( Message-ID: <ckfnb7$609$1@pcls4.std.com>  + tomarsin2015@aol.com (Tomarsin2015) writes:    >VMS 6.2  G So much of DUDRIVER has been rototilled since then I really can't guess E what the problem might be.  My best advice: Update!  DUDRIVER is much  better now.   H A MSCP served host being the alternate path really isn't a problem.  TheI IOs go through the primary path.  If you ever see the MSCP served host as J primary path when you have a (valid) direct path, then you have a problem.     --   -Mike    ------------------------------  % Date: Mon, 11 Oct 2004 22:12:22 -0400 # From: "John Smith" <a@nonymous.com> # Subject: HP ignores OpenVMS & SWIFT , Message-ID: <ed-dncIQ_dEToPbcRVn-uA@igs.net>  9 http://www.hp.com/hpinfo/newsroom/press/2004/041011a.html     G HP Launches Financial Services Industry Platform for SWIFT Connectivity   I Solution delivers lower costs and higher availability through HP ProLiant . server running Microsoft clustering technology  " ATLANTA, SIBOS 2004, Oct. 11, 2004L ---------------------------------------------------------------------------- ----  L HP today announced a new server platform that enables financial institutionsL and large corporations to access the SWIFT financial messaging network whileL taking advantage of the lower costs and higher availability of the clustered Microsoft Windows environment.   I Companies that communicate with each other via the SWIFT global financial K messaging network need client interface software, such as the SWIFTAlliance 2 range of interfaces, for security and reliability.  A HP's solution includes software developed by HP that supports the 8 SWIFTAlliance interfaces - and makes them cluster-aware.  E The solution, called the "SWIFTAlliance Cluster Enabler for Microsoft @ Windows," runs on multiple HP ProLiant servers and leverages theJ high-availability clustering technology of the Microsoft Windows operatingK system. In a Windows-based cluster, if one server fails, the other picks up H the workload automatically so that data is protected and the application continues to run.   D For reliability and compatibility, HP has tested and benchmarked itsI solution on a cluster of HP ProLiant platforms running Microsoft Windows. K Similarly, SWIFT has tested the SWIFTAlliance interface on the HP clustered E platforms, for the SWIFT connectivity packs 1, 2, 3 and 4, supporting * traffic up to 40 mail messages per second.  F Along with the solution itself, HP can provide customers with completeK support and maintenance services for three years. HP is responsible for all C support services via cross-agreements with SWIFT and Microsoft - so G customers enjoy the advantages of a single point of accountability with  rapid response times.   G "Financial institutions and large corporations increasingly rely on the K SWIFT network and IP messaging platform, SWIFTNet, for their operations. HP F is making access to the network much less costly for customers - whileI allowing the high availability needed for high stakes transactions," said F Joe Lott, director, payments, Financial Services Industry, HP. "The HPH SWIFTAlliance solution leverages the accumulated clustering expertise of both Microsoft and HP."   F "Our mission is to provide the most cost-effective financial messagingH service without compromising security or reliability," said Alain Drese,I head of Customer Integration, SWIFT. "HP's offering contributes to making H the benefits of SWIFTNet readily available to financial institutions and their corporate clients."   K The HP SWIFTAlliance Cluster Enabler for Microsoft Windows is now available J worldwide. Estimated U.S. street pricing ranges from $4,500 to $12,000 perF server configuration. Estimated pricing for the HP support offering is $1,500 per server, per year.(1)   L More information about the SWIFTAlliance Cluster Enabler and other offerings3 for the financial services industry is available at 0 http://www.hp.com/hps/financial/fi_payment.html.  F SWIFT is the industry-owned cooperative supplying secure, standardisedL messaging services and interface software to 7,650 financial institutions inK more than 200 countries. The SWIFT community includes banks, broker/dealers C and investment managers, as well as their market infrastructures in ) payments, securities, treasury and trade.   K Over the past 10 years, SWIFT message prices have been reduced more than 70 J percent, and system availability approaches 5x9's reliability - 99.999% ofJ uptime. SWIFT's new IP messaging platform, SWIFTNet, has been launched and( will be fully rolled out by end of 2004.   ------------------------------  % Date: Mon, 11 Oct 2004 23:37:01 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> ' Subject: Re: HP ignores OpenVMS & SWIFT , Message-ID: <416B5143.B575E250@teksavvy.com>   John Smith wrote:  > ; > http://www.hp.com/hpinfo/newsroom/press/2004/041011a.html  > I > HP Launches Financial Services Industry Platform for SWIFT Connectivity   J Why should it mention VMS ? SWIFT widthdrew support for VMS years ago whenE Palmer told them VMS had no future.  SWIFT supports Unix and Windows.   G And judging from the prominence of "Microsoft" in that press release, I K suspect that HP got some help from Microsoft to push that platform since it L generates sales and good image for Microsoft running really important stuff.  L And while the readers of cov know better than to associate "clustering" withN Windows, the folks who matter don't know the difference and think that windows@ can provide sufficient stability and disaster recovery features.  J What is telling in that press release is that HP decided to choose WindowsL instead of HP-UX for their value added bit. Doesn't SWIFT alliance run underK HP-UX ? I know it runs on Solaris because many ex-VMS customers migrated to ( Sun when DEC "pulled the plug" on ST400.  N (I say DEC pulled the plug because VMS was a strategic platform for SWIFT, butK after being told that users shoudl migrate to Unix, SWIFT heard the message M loud and clear and dropped VMS as not only its prefered platform, but dropped  it alltogether.   L HP is in fact quite late in the game since former VMS customers have already migrated to other platforms.   ------------------------------  % Date: Tue, 12 Oct 2004 00:14:56 -0400 ( From: David Froble <davef@tsoft-inc.com>' Subject: Re: HP ignores OpenVMS & SWIFT , Message-ID: <416B5A40.9070600@tsoft-inc.com>   John Smith wrote:   ; > http://www.hp.com/hpinfo/newsroom/press/2004/041011a.html  >  > I > HP Launches Financial Services Industry Platform for SWIFT Connectivity  > K > Solution delivers lower costs and higher availability through HP ProLiant 0 > server running Microsoft clustering technology > $ > ATLANTA, SIBOS 2004, Oct. 11, 2004N > ---------------------------------------------------------------------------- > ---- > N > HP today announced a new server platform that enables financial institutionsN > and large corporations to access the SWIFT financial messaging network whileN > taking advantage of the lower costs and higher availability of the clustered! > Microsoft Windows environment.     ; Oh goodie!  History repeating itself again.  Let's see ....   P DEC has VAX customers, and wants more Alpha customers.  Great solution, rob the 2 VAX customer base to acquire more Alpha customers.  L HP now owns VMS which was a prefered platform for Swift.  HP provides a new O Swift system which allows them to pay licenses to Microsoft, replacing the VMS  B based systems which have no need to pay license fees to Microsoft.  J At least there is consistancy.  Nothing else good I can see.  Problem is,  consistantly stupid!   Dave   --  4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com 170 Grimplin Road  Vanderbilt, PA  15486    ------------------------------  % Date: Mon, 11 Oct 2004 22:46:19 -0400 ' From: "Main, Kerry" <kerry.main@hp.com> Y Subject: RE: Obsucure Inquier article about Intel mentions VMS (and not any other  OS') o R Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB45D892@tayexc19.americas.cpqcorp.net>   > -----Original Message-----9 > From: Larry Kilgallen [mailto:Kilgallen@SpamCop.net]=20   > Sent: October 11, 2004 5:33 AM > To: Info-VAX@Mvb.Saic.Com ? > Subject: RE: Obsucure Inquier article about Intel mentions=20 3 > VMS (and not any other OS') other OS') other OS')  >=20 > In article=20 @ > <FD827B33AB0D9C4E92EACEEFEE2BA2FB45D86E@tayexc19.americas.cpqc5 > orp.net>, "Main, Kerry" <kerry.main@hp.com> writes:  > >=20 > >> -----Original Message----- > > >> From: Larry Kilgallen [mailto:Kilgallen@SpamCop.net]=3D20$ > >> Sent: October 10, 2004 10:52 PM > >> To: Info-VAX@Mvb.Saic.ComD > >> Subject: RE: Obsucure Inquier article about Intel mentions=3D206 > >> VMS (and not any other OS') other OS') other OS')	 > >>=3D20  > >> In article=3D20C > >> <FD827B33AB0D9C4E92EACEEFEE2BA2FB45D86D@tayexc19.americas.cpqc 8 > >> orp.net>, "Main, Kerry" <kerry.main@hp.com> writes:	 > >>=3D20 C > >> > for that matter. However, I know of two major Customers=3D20  > >> that wanted to < > >> > convert OpenVMS applications and both went down in=20 > flames big time.	 > >>=3D20 . > >> The Customers or the conversion efforts ?	 > >>=3D20  > >=20 > >=20B > > Well, I am sure there was a lot of heat when these projects=20
 > failed, but @ > > these are big Customers and you will not hear about these=20 > in the press.  >=20> > I did not expect that.  What I was asking was which "when=20 > down in flames".9 > That is, what is the antecedent of "both" - the word=20  > "customers" or the > word "applications". >=20  F Case #1 - Original 2 year mission critical application port was VAX toH UNIX. End result was 6 month successful VAX to Alpha project. Runs great today.  F Case #2 - 2 year project to move business logic on OpenVMS to Windows,F leave back end on OpenVMS. Jury still out, but I expect a "make better> use of what we have" project using web services on OpenVMS.=20  : Both Customers are not the type that goes out of business.   Regards   
 Kerry Main Senior Consultant  HP Services Canada Voice: 613-592-4660  Fax: 613-591-4477  kerryDOTmainAThpDOTcom (remove the DOT's and AT)=20  $ "OpenVMS has always had integrity .. Now, Integrity has OpenVMS .."   ------------------------------  % Date: Mon, 11 Oct 2004 22:46:19 -0400 ' From: "Main, Kerry" <kerry.main@hp.com> Y Subject: RE: Obsucure Inquier article about Intel mentions VMS (and not any other  OS') o R Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB45D891@tayexc19.americas.cpqcorp.net>   > -----Original Message-----< > From: Soterro [mailto:soterroatyahoocom@address.hp.com]=20  > Sent: October 11, 2004 9:43 AM > To: Info-VAX@Mvb.Saic.Com ? > Subject: Re: Obsucure Inquier article about Intel mentions=20 A > VMS (and not any other OS') other OS') other OS') other OS')=20 8 > other OS') other OS') other OS') other OS') other OS') >=20 > Main, Kerry wrote:/ > >>From: John Smith [mailto:a@nonymous.com]=20 B > >>You aren't directly to blame for this and I know you'd like=20 > >>to keep your@ > >>job, but just how many times have you and your colleagues=20 > >>written carly(tm) . > >>or curly or GQ Bob about this in the past? > >=20> > > Actually, I work with UNIX, Windows as well as OpenVMS.=20 > You're value to J > > any corporation goes up exponentially if you know more than one OS.=20 >=20I > I think JS asked something else: is it so out of this world to think=20 ? > that the insiders could point their superiors to some things?  >=20 > S  >   0 And you think perhaps that this does not happen?  F Course, there are many ways to offer different opinions within a largeD company, but keep in mind that every OS or platform within a companyG will always try and say it needs more $'s to promote/enhance itself.=20   E Take a big company like IBM - what do you think an AIX field person's H chances are of convincing their Mgmt to quit promoting Linux as AIX is a better system?  E Or even Sun - what would someone like Andrew's chances be to convince C his Mgmt that investing in AMD is a mistake as SPARC is the only HW  solution to focus on?   H The wise man knows when to pick his battles and when to bend when he has to.    :-)   
 Kerry Main Senior Consultant  HP Services Canada Voice: 613-592-4660  Fax: 613-591-4477  kerryDOTmainAThpDOTcom (remove the DOT's and AT)=20  $ "OpenVMS has always had integrity .. Now, Integrity has OpenVMS .."   ------------------------------  % Date: Mon, 11 Oct 2004 23:44:39 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> Y Subject: Re: Obsucure Inquier article about Intel mentions VMS (and not any other  OS') o , Message-ID: <416B530D.FFFEA4AC@teksavvy.com>   "Main, Kerry" wrote:H > Course, there are many ways to offer different opinions within a largeF > company, but keep in mind that every OS or platform within a companyF > will always try and say it needs more $'s to promote/enhance itself.  M If you go back through the archives of cov, you will find many Digits arguing 8 AGAINST advertising of VMS, giving all sorts of reasons.  M So, one must wonder if the requests for advertising money are even made to HP  by the VMS group.   M And at a time when the fear to lose one's job is permanent (has been the case I since 1992, so it is pretty much permanent, isn't it ?), I also fear that M there are few few Digits who are willing to rock the boat and tell their ever L changing management that they are making mistakes and constantlty forgetting VMS etc etc.  F Has the Ascention of Rich Marcello to a higher level given VMS anotherK "Renaissance" period where ther was actually some marketing ? ? ? ?  He has L probably had to learn to be politically correct and also omit mention of VMSJ so as to not rock the boat. His own job security is rightly more important than one product.   N What is needed is one kakamaze manager to rampage through HP and make a lot ofN noise about VMS and force change  (and this has to be done before he gets toldF to leave). In fact, it needs a sacrificial lamb. This way, the jobs of- Marcello Gorham etc would not be jeoperdized.    ------------------------------  % Date: Mon, 11 Oct 2004 23:38:37 -0400 - From: "heuveltjes" <heuveltjes@email.msn.com> " Subject: Re: Rare behaviour xabrdt* Message-ID: <416b52d5@usenet01.boi.hp.com>  9 "Wim Versteeg" <w.j.versteeg@wanadoo.nl> wrote in message 6 news:740a9987.0410031241.b17aa74@posting.google.com...J > A program which uses the xab$q_rdt field of the xabrdt seems not to work under I > 7.3-1 with ODS-5 disks. Is it 7.3-1 or ODS-5 ? I assume the latest, but  why ? L > Under 7.1-2 and 7.2-1 with ODS-2 it works just fine. The date, supplied in the L > xab$q_rdt field just before the $CLOSE is in fact the revision date of theK > file. But on 7.3-1 with ODS-5 the supplied date is ignored and the actual  date* > and time are used for the revision date.  & Oh, I didn't see this note untill now.' Check out 'access date semantics. Like:   C     $set vol <ods-5_device-name> /volume_characteristic=access_date   	 See also:   K http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=547669    Met vriendelijk sport groeten, Hein.    ------------------------------  # Date: Mon, 11 Oct 2004 20:32:19 GMT & From: jlsue <jefflsxxxz@sbcglobal.net> Subject: Re: Sayonara storage 8 Message-ID: <a6rlm05sjtli79hho82vmhilo90at9fl1b@4ax.com>  I On Mon, 11 Oct 2004 08:50:38 -0600, John Nebel <john.nebel_vms@csdco.com>  wrote:     > F >The CEO is I and it is not worth the time writing an idiot.  You can G >tell that I'm not happy.  It's her decision(s) to destroy HP in large   >and small increments. > I >The EVA is a nice, fast system and the numbers on the GS1280 are amazing  > 7 >http://www.cs.virginia.edu/stream/top20/Bandwidth.html  > C >especially since the application is sensitive to memory bandwidth.  >   H John, please do not take this inquirer article at face value.  From whatJ I've been able to learn it is a *complete falsification*, possibly planted by competitors.   J Our next generation EVA is on track and NOT being cancelled, as claimed in the article.  J I am told that HP will be releasing something soon to counter this pack of falsehoods.    --- jls 0 The preceding message was personal opinion only.6 I do not speak in any authorized capacity for anyone,  and certainly not my employer.- (get rid of the xxxz in my address to e-mail)    ------------------------------  % Date: Mon, 11 Oct 2004 17:07:29 -0400  From: norm.raphael@metso.com Subject: Re: Sayonara storage Q Message-ID: <OFA24A1A64.BA8DE469-ON85256F2A.007363AB-85256F2A.0074073A@metso.com>   A jlsue <jefflsxxxz@sbcglobal.net> wrote on 10/11/2004 04:32:19 PM:    [snip    > J > John, please do not take this inquirer article at face value.  From whatD > I've been able to learn it is a *complete falsification*, possibly planted  > by competitors.  > I > Our next generation EVA is on track and NOT being cancelled, as claimed  in > the article. > I > I am told that HP will be releasing something soon to counter this pack  of
 > falsehoods.  >   H The real trajedy is of course that the whole thing is become credible to	 too many. : There is at least one lesson lurking in this discourse....  	 > --- jls 2 > The preceding message was personal opinion only.7 > I do not speak in any authorized capacity for anyone,   > and certainly not my employer./ > (get rid of the xxxz in my address to e-mail)    ------------------------------  % Date: Mon, 11 Oct 2004 17:58:02 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com>  Subject: Re: Sayonara storage , Message-ID: <416B01E3.E70CFAD6@teksavvy.com>   jlsue wrote:J > John, please do not take this inquirer article at face value.  From whatL > I've been able to learn it is a *complete falsification*, possibly planted > by competitors.  > L > Our next generation EVA is on track and NOT being cancelled, as claimed in > the article.  N The way I read the airticle, it didn't mention specific product cancellations.N What it mentioned was that HP was lowering quality of support to a point whereG customers would no longer buy those products in quatities sufficient to 7 justified continued R&D and production of that product.   K This tidbit by itself doesn't seem so significant. But when you put all the N tidbits together, it gives an impression that HP has a strategy of downsiziongN its enterprise business to a point where it sells only 8086 based servers, andN then focusing most of its energies on consumer commodity items (ink, printers,+ cameras, rebranded glorified walkmans etc).   J If Carly didn't decide on this strategy, then the way the company is being, managed may still lead it in that direction.   ------------------------------  % Date: Mon, 11 Oct 2004 23:56:46 -0400 ( From: David Froble <davef@tsoft-inc.com> Subject: Re: Sayonara storage , Message-ID: <416B55FE.7010204@tsoft-inc.com>   norm.raphael@metso.com wrote:    > C > jlsue <jefflsxxxz@sbcglobal.net> wrote on 10/11/2004 04:32:19 PM:  >  > [snip  >  > J >>John, please do not take this inquirer article at face value.  From whatD >>I've been able to learn it is a *complete falsification*, possibly >>	 > planted  >  >>by competitors.  >>I >>Our next generation EVA is on track and NOT being cancelled, as claimed  >> > in >  >>the article. >>I >>I am told that HP will be releasing something soon to counter this pack  >> > of > 
 >>falsehoods.     Q I'm hoping that what you're saying is the reality of the situation.  But much of  I the discussion was about support being moved from those with significant  M experience with the product(s) to 4-week wonders in India.  Are the USA jobs   intact, or not.   P Even your post raises questions.  If the article is totally false, HP shouldn't M need any time considering it's response.  They should be able to immediately  P declare the whole article false.  Taking time for a response suggests that they L are calculating how to rebut the false parts, while not mentioning the true K parts.  Or how to spin things so that it looks like they are rebutting the  / entire article, when some parts of it are true.     J > The real trajedy is of course that the whole thing is become credible to > too many. < > There is at least one lesson lurking in this discourse.... >  > 	 >>--- jls 2 >>The preceding message was personal opinion only.7 >>I do not speak in any authorized capacity for anyone,   >>and certainly not my employer./ >>(get rid of the xxxz in my address to e-mail)  >> >    Dave   --  4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com 170 Grimplin Road  Vanderbilt, PA  15486    ------------------------------  % Date: Tue, 12 Oct 2004 00:49:36 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com>  Subject: Re: Sayonara storage , Message-ID: <416B6242.21B24B33@teksavvy.com>   David Froble wrote: Q > declare the whole article false.  Taking time for a response suggests that they M > are calculating how to rebut the false parts, while not mentioning the true L > parts.  Or how to spin things so that it looks like they are rebutting the1 > entire article, when some parts of it are true.   I All the political speech writers in the USA are rather busy right now, so G Carly has to wait in line before she can have text concucted which says N absolutely nothing and denies nothing, but gives carly the ability to say thatN she forcefully responded to that article that made allegations about future of storage business.    :-)    ------------------------------  % Date: Mon, 11 Oct 2004 14:16:06 -0400 ( From: "Hein" <hein.nomail@hp.nomail.com>, Subject: Re: SHOW DEVICE/FILES with FORTRAN?, Message-ID: <416acf39$1@usenet01.boi.hp.com>  5 "John Brandon" <brandon@dalsemi.com> wrote in message + news:04093021052367@dscis6-0.dalsemi.com...  : G > > Without breaching confidences, can you explain why this needs to be 7 > > done? What is the root problem this seeks to solve?  > L > Users need to be kicked out of the application in order to upload new dataJ > files into the work-stream.  Otherwise users would continue to use stale data.   E Ah, I missed this clarification. And this is a DB like Oracle, or RMS 
 datafiles.  D If it is RMS datafiles, then this becomes an almost trivial problem.K Just $SEARCH for a critical file in use by the application, from the search I results use the DVI and FID and a piece of string ( "RMS$" ;-) to make up F the RMS file lock name. The use GETLKI with LKI$_LOCKS to get all PIDsL involved with this lock. Now blast them. Minor detail: yo need to be in exec mode for this.I The code I posted earlier as a reply has examples for most of the details 	 you need.   L Of course what you really want to do is change the application to take out aL CR look on start, with a blocking AST. When the update needs to happen, takeI out an exlusive EX or PW lock which will Blast the right users. They then J drop to NL, close files as needed, issue warning on screen, and go wait by reconverting to CR. D When they get the CR lock back, re-open and allow users to continue.    Hein.   ------------------------------  + Date: Mon, 11 Oct 2004 18:51:50 +0000 (UTC) 6 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER), Subject: Re: SHOW DEVICE/FILES with FORTRAN?1 Message-ID: <newscache$3xnf5i$r5t1$1@news.sil.at>   ] In article <cke9pr$72t$1@newslocal.mitre.org>, lewis@PROBE.MITRE.ORG (Keith A. Lewis) writes: 1 >With the test file open (by the CREATE command):  > F >$ write sys$output f$file("lumina$scratch0:[lewis]temp.tmp","locked")+ >%SYSTEM-W-ACCONFLICT, file access conflict # > \lumina$scratch0:[lewis]temp.tmp\  >  >Again after closing the file: > G >$ write sys$output  f$file("lumina$scratch0:[lewis]temp.tmp","locked")  >FALSE > E >I'm not sure how it ever gets set to TRUE, but this test seems to be * >functional as long as you trap the error.  F F$FILE item LOCKED is for the "Deaccess locked" (you know, the one you. can UNLOCK or SET FILE/UNLOCK but not LOCK ;-)  G        The SET FILE/UNLOCK command can clear the cause of the following         error message:   4        %SYSTEM-W-FILELOCKED, file is deaccess locked  B        However, this command cannot resolve the cause of the error        message:   8        %RMS-W-FLK, file currently locked by another user   --   Peter "EPLAN" LANGSTOEGER % Network and OpenVMS system specialist  E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------    Date: 11 Oct 2004 15:40:46 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) , Subject: Re: SHOW DEVICE/FILES with FORTRAN?3 Message-ID: <5JppZl6IZHAM@eisner.encompasserve.org>   j In article <newscache$3xnf5i$r5t1$1@news.sil.at>, peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) writes: > 6 >        %SYSTEM-W-FILELOCKED, file is deaccess locked  A    I thought that was an artifact of FCS (predecessor to RMS) and C    no longer a situation one could actually get into.  Can it still &    be reached via ACP (XQP) functions?   ------------------------------  % Date: Mon, 11 Oct 2004 16:55:44 -0400 ( From: "Hein" <hein.nomail@hp.nomail.com>, Subject: Re: SHOW DEVICE/FILES with FORTRAN?, Message-ID: <416af4a3$1@usenet01.boi.hp.com>  H "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote in message- news:5JppZl6IZHAM@eisner.encompasserve.org... H > In article <newscache$3xnf5i$r5t1$1@news.sil.at>, peter@langstoeger.at# (Peter 'EPLAN' LANGSTOEGER) writes:  > > 8 > >        %SYSTEM-W-FILELOCKED, file is deaccess locked > C >    I thought that was an artifact of FCS (predecessor to RMS) and E >    no longer a situation one could actually get into.  Can it still ( >    be reached via ACP (XQP) functions?  2 The functionality is still there, best I can tell:  L The XQP writes through WCB$V_DLOCK to FCH upon close if need be. FIB$V_DLOCKL will set WCB$V_DLOCK on IO$_ACCESS is so requested. IO$_MODIFY can clear the9 bit when the application believes the file is consistent.   I I personally have not seen a single application exploiting DEACCESS LOCK, E since RSX-11M days. Do you know of any VMS provided tool that use the H Deaccess lock semantics? Do you know of any ISV / 3rd party common usageI applications that use this? Do you know of any end-user applications thatp	 use this?   K I was wandering whether it would make sense to re-define F$FILE(x,"LOCKED")SJ to also mean that the file is currently being accessed: like it would show up with SHOW DEV/FILE.  L It would be a bit hokey to have a double defintion, and in the sense that weG all understand that this is a transient state. It may well change rightMI after you looked. For any serious business you just have to open the fileaI exclusively yourself. If that works, it was not, and is not, locked by ana other process :-).  J Still, I believe this to be what our customer base wants / expects to see.   Opinions? Hein.h   ------------------------------  % Date: Mon, 11 Oct 2004 18:04:43 -0400h- From: JF Mezei <jfmezei.spamnot@teksavvy.com> , Subject: Re: SHOW DEVICE/FILES with FORTRAN?, Message-ID: <416B0374.99ED3C74@teksavvy.com>   Hein wrote:RK > I personally have not seen a single application exploiting DEACCESS LOCK,eG > since RSX-11M days. Do you know of any VMS provided tool that use thek > Deaccess lock semantics?    N Not sure if that particular semantic is used, however, ALL-IN-1 uses a similar methodology to lock documents.  M Consider the case where you are editing a file. The actual RMS file is lockedPL only while you read it in or while writing it back. So ALL-IN-1 uses its ownK document index to mark the document as locked, which means that anyone elsePM trying to edit it while you are working on it, whether on a remote PC or an a-G character cell terminal will know that the document is curently in use.,N (because A1 ran on clients, they probably figured it was easier to use a fieldN in an indexd file to mark a document locked as opposed to using a RMS semanticH which wouldn't translate well when transfered to remote non-VMS clients.  N I think TPU should have made use of that RMS semantic, at least optionally, toA ensure nobody else could edit that file while you are editing it.t   ------------------------------  + Date: Tue, 12 Oct 2004 01:15:52 +0000 (UTC)h, From: lewis@PROBE.MITRE.ORG (Keith A. Lewis), Subject: Re: SHOW DEVICE/FILES with FORTRAN?. Message-ID: <ckfb87$nrl$2@newslocal.mitre.org>  | "Hein" <hein.nomail@hp.nomail.com> writes in article <416af4a3$1@usenet01.boi.hp.com> dated Mon, 11 Oct 2004 16:55:44 -0400:L >I was wandering whether it would make sense to re-define F$FILE(x,"LOCKED")K >to also mean that the file is currently being accessed: like it would showA >up with SHOW DEV/FILE.g >DM >It would be a bit hokey to have a double defintion, and in the sense that we0H >all understand that this is a transient state. It may well change rightJ >after you looked. For any serious business you just have to open the fileJ >exclusively yourself. If that works, it was not, and is not, locked by an >other process :-).  >iK >Still, I believe this to be what our customer base wants / expects to see.P  L Another interesting thing that HP could do is document (and support) the RMSL lock to system lock mapping.  Then it would be easy to write a freeware tool0 to check this kind of thing by calling $GETLKIW.  4 RMS does use the lock database for this, doesn't it?  0 --Keith Lewis              klewis {at} mitre.org> The above may not (yet) represent the opinions of my employer.   ------------------------------  % Date: Mon, 11 Oct 2004 23:42:14 -0400a- From: "heuveltjes" <heuveltjes@email.msn.com> , Subject: Re: SHOW DEVICE/FILES with FORTRAN?* Message-ID: <416b53ae@usenet01.boi.hp.com>  9 "Keith A. Lewis" <lewis@PROBE.MITRE.ORG> wrote in messaget( news:ckfb87$nrl$2@newslocal.mitre.org...6 > "Hein" <hein.nomail@hp.nomail.com> writes in articleG <416af4a3$1@usenet01.boi.hp.com> dated Mon, 11 Oct 2004 16:55:44 -0400:-  J > Another interesting thing that HP could do is document (and support) the  RMS lock to system lock mapping.   Read the source Luke!4J There is actually not much to it. Just use ANAL/SYS to poke around after a $GET with default lock. Trivial   0 > Then it would be easy to write a freeware tool2 > to check this kind of thing by calling $GETLKIW.   Has been done:F http://h71000.www7.hp.com/freeware/freeware60/rms_tools/src/blocking.c  G I also posted a deriviate from this tool, just for rms file locks, as a." potential solution to this thread.  6 > RMS does use the lock database for this, doesn't it?   Absolutely!y   Cheers,v Hein.    ------------------------------  % Date: Mon, 11 Oct 2004 23:35:56 -0400M( From: David Froble <davef@tsoft-inc.com>, Subject: Re: SHOW DEVICE/FILES with FORTRAN?, Message-ID: <416B511C.6000808@tsoft-inc.com>   Hein wrote:   J > "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote in message/ > news:5JppZl6IZHAM@eisner.encompasserve.org...0 > H >>In article <newscache$3xnf5i$r5t1$1@news.sil.at>, peter@langstoeger.at >>% > (Peter 'EPLAN' LANGSTOEGER) writes:, > 7 >>>       %SYSTEM-W-FILELOCKED, file is deaccess lockedC >>>vC >>   I thought that was an artifact of FCS (predecessor to RMS) and0E >>   no longer a situation one could actually get into.  Can it still ( >>   be reached via ACP (XQP) functions? >> > 4 > The functionality is still there, best I can tell: > N > The XQP writes through WCB$V_DLOCK to FCH upon close if need be. FIB$V_DLOCKN > will set WCB$V_DLOCK on IO$_ACCESS is so requested. IO$_MODIFY can clear the; > bit when the application believes the file is consistent.r > K > I personally have not seen a single application exploiting DEACCESS LOCK,hG > since RSX-11M days. Do you know of any VMS provided tool that use theiJ > Deaccess lock semantics? Do you know of any ISV / 3rd party common usageK > applications that use this? Do you know of any end-user applications thats > use this?u > M > I was wandering whether it would make sense to re-define F$FILE(x,"LOCKED") L > to also mean that the file is currently being accessed: like it would show > up with SHOW DEV/FILE. > N > It would be a bit hokey to have a double defintion, and in the sense that weI > all understand that this is a transient state. It may well change righttK > after you looked. For any serious business you just have to open the filefK > exclusively yourself. If that works, it was not, and is not, locked by anV > other process :-). > L > Still, I believe this to be what our customer base wants / expects to see. >  > Opinions? Hein.     5 Maybe I should automate the posting of the following:o   ;i) ;       Assign the VMS I/O channel numberm ;TA          MOVAB   WRK_Q_DEVDSC,R0         ; Set descriptor addresse/          CVTBW   NAM$T_DVI(R8),DSC$W_LENGTH(R0)d2          MOVAB   NAM$T_DVI+1(R8),DSC$A_POINTER(R0)C          $ASSIGN_S DEVNAM=(R0),CHAN=WRK_W_VMSCHN,ACMODE=#PSL$C_USER2B          BLBS    R0,300$                 ; Continue if all is okay    9 After this snippet of code, there is an open file, right?    Where's the lock?p  / You can not check for an open file using locks!c   Dave     -- v4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com 170 Grimplin Road2 Vanderbilt, PA  15486-   ------------------------------  % Date: Tue, 12 Oct 2004 00:46:02 -0400-- From: JF Mezei <jfmezei.spamnot@teksavvy.com>t, Subject: Re: SHOW DEVICE/FILES with FORTRAN?, Message-ID: <416B616C.FA2E19CA@teksavvy.com>   David Froble wrote:a+ > ;       Assign the VMS I/O channel numbera > ;bC >          MOVAB   WRK_Q_DEVDSC,R0         ; Set descriptor address 1 >          CVTBW   NAM$T_DVI(R8),DSC$W_LENGTH(R0)a4 >          MOVAB   NAM$T_DVI+1(R8),DSC$A_POINTER(R0)E >          $ASSIGN_S DEVNAM=(R0),CHAN=WRK_W_VMSCHN,ACMODE=#PSL$C_USER D >          BLBS    R0,300$                 ; Continue if all is okay > ; > After this snippet of code, there is an open file, right?e > Where's the lock?i    K No it doesn't open a file, it opens a channel to the device. That device isu the disk that holds some file.  M And just because yo don't see an obvious call to the locking routines doesn'teK mean that a lock isn't taken on a resource behind the scenes on your behalfc without you knowing it.m   ------------------------------    Date: 11 Oct 2004 12:59:51 -0700. From: spamsink2001@yahoo.com (Alan E. Feldman)9 Subject: Re: SYS$PARSE: prevented translation of logicalsr= Message-ID: <b096a4ee.0410111159.518c5b53@posting.google.com>h  s spamsink2001@yahoo.com (Alan E. Feldman) wrote in message news:<b096a4ee.0410102221.5cbf41e4@posting.google.com>...sc > JF Mezei <jfmezei.spamnot@teksavvy.com> wrote in message news:<416936EA.5B0F6D30@teksavvy.com>...iB > > I would like to have SYS$PARSE of a file specification such as_ > > SYS$LIBRARY:MYFILE.TLB keep the SYS$LIBRARY instead of expanding it to SYS$SYSROOT:[SYSLIB]o > > M > > I have tried settin the NOCONCEAL  NAM block option to 0, but it seems toaP > > conceal only logicals that are defined as devices (terminal, concealed), but7 > > logicals such as SYS$LIBRARY: still get translated.S > >  > > Is there a magic trick ? > > N > > goal: after parsing a file specification, if it involves possible multipleR > > matching files (wildcard included, or file name omitted), I then pop up a fileN > > selection dialogue, and I would like the device/directory to appear as the > > user had typed them. > > P > > So, if the user types in "SYS$LIBRARY:", with a default file spec of "*.TLB" > > fed to $PARSE. > > Z > > I want $parse to convert it to SYS$LIBRARY:*.TLB  instead of SYS$SYSROOT:[SYSLIB]*.TLB >  > 7 > The best I can come up with is to make it of the form  > % >     F$PARSE(".TLB",0::SYS$LIBRARY:)i > D > but there MUST be a colon after the device/logical name and you'll% > have to strip off the 0:: yourself.s > C > I don't think you can make the parse service do what you want, at$ > least not directly.e  
 Uh, make thatn  %     F$PARSE("0::.TLB","SYS$LIBRARY:")f  , but it still may not do quite what you want.   ------------------------------  % Date: Mon, 11 Oct 2004 21:35:25 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com>eK Subject: The H-P Compaq Merger Two Years Out:  Still Waiting for the Upsider, Message-ID: <416B34CA.EC82BBEE@teksavvy.com>  E > http://knowledge.wharton.upenn.edu/index.cfm?fa=viewArticle&id=1053u  H Most interesting analysis of HP with both pros and cons, comparison with IBM/Gerstner and Dell.  L Essentially, it says that the merger is still a work in progress and that inI order to realise the "sum is greater than the parts" HP needs to do more.e   ------------------------------  % Date: Mon, 11 Oct 2004 22:16:22 -0400 # From: "John Smith" <a@nonymous.com>F: Subject: Wonder how this would work on Alpha/VMS clusters?, Message-ID: <FP6dnX92bYjho_bcRVn-rA@igs.net>  9 http://www.hp.com/hpinfo/newsroom/press/2004/041008a.html:  K Oracle and HP Smash World Record for TPC-H 10 TB Data Warehousing Benchmarkg  L Oracle Database 10g on HP Integrity Superdome cluster tops IBM's best resultE on the TPC-H 10 TB Data Warehousing Benchmark, using 20 percent fewerhE processors and offering more than 33 percent better price/performanceC  2 REDWOOD SHORES and PALO ALTO, Calif., Oct. 8, 2004L ---------------------------------------------------------------------------- ----  J Oracle and HP (NYSE: HPQ) (Nasdaq: HPQ) today announced a new world recordD TPC-H 10 terabyte (TB) data warehousing benchmark result for OracleI Database 10g and Oracle Real Application Clusters running HP-UX 11i v2 ontL two HP Integrity Superdome servers. This highlights the scalability and costG leadership of Oracle Database 10g and HP Integrity Superdome servers bynF outperforming IBM DB2's best TPC-H 10 TB result using 20 percent fewerJ processors, while offering more than 33 percent better price/performance1.H Furthermore, the benchmark demonstrates the single processor performanceL superiority of the Integrity servers with the Intel Itanium 2 processor by5 outperforming the IBM Power processor by 73 percent1."  I Data warehouse performance is a requirement in today's fast moving globalrJ markets. Robust, highly scalable data warehouses allow executives and dataK analysts to scour vast repositories of customer and market information withiD decision support tools to uncover new strategic markets and identifyK opportunities to streamline costs. Only the most powerful database softwaretK and server platforms, such as those from Oracle and HP, are able to examine:B these large volumes of data, execute queries with a high degree ofD complexity, and give insight to these critical business questions as/ demonstrated by today's TPC-H benchmark result./  J First Trust Corp., one of the nation's largest independent trust companiesB for self-directed retirement plans, is a joint Oracle-HP customer.E "Performance and scalability are critical to First Trust. Our ItaniumrE 2-based HP Integrity servers running Oracle Real Application Clusters I exceeded our expectations, delivering a threefold performance improvement-K over our previous IBM RISC-based system running on AIX, while still drivingbF down our total cost of ownership," said Jeff Knight, vice president ofL Technology and Vendor Relations, First Trust. "With these benchmark results,G HP and Oracle are delivering impressive solutions that provide the best > performance per processor and scale to meet the most demanding requirements."  K Oracle and HP together set the world-record performance of 86,282.7 queriesmL per hour (QphH) @ 10 TB at a price-performance ratio of $161/QphH @ 10 TB onB the independent Transaction Processing Performance Council's TPC-HJ benchmark1. This benchmark result produced 88 percent scalability from oneF to two nodes for a 10 TB data warehouse2. This two-node server clusterK represents the industry's first publication of a commercial benchmark using I the standards-based InfiniBandTM technology. The InfiniBand offering fromFK HP, along with 128 Itanium 2 processors running at 1.5 GHz with 6M L3 cache J connected to HP StorageWorks XP128 disk arrays contributed to the superiorK scale-out performance for this benchmark. Previously, HP and Oracle set the F world record for the fastest single-system TPC-H 10 TB benchmark3. TheH unprecedented performance results of both the single system and two-nodeH cluster show the scale-up and scale-out superiority of the HP and OracleL combination. Based on this and previous achievements, HP and Oracle now holdL the world records for both non-clustered and clustered TPC-H results for the 10 and 3 TB scale factors.  L As the leading database for production data warehousing, Oracle Database 10gC provides a single, integrated database engine for scalable and high4I performing data warehousing implementations. "This achievement once again H underscores Oracle's ability to satisfy large enterprise data processingI needs," said Chuck Rozwat, executive vice president of Development Server F Technologies, Oracle Corp. "With Oracle, customers can always count on3 getting world class performance at the best value."n  K "This benchmark demonstrates that HP is committed to offering our customers2D a business intelligence platform with the greatest power at the bestC price/performance in the industry," said Rich Marcello, senior vicer? president and general manager of Business Critical Servers, HP.t  G This benchmark adds to the extensive list of HP and Oracle Database 10g0F performance world records. HP and Oracle were the first to surpass oneH million transactions per minute with the TPC-C benchmark by using the HPG Integrity Superdome server and the companies currently hold the overallrL TPC-C world record result of 1.18 million transactions per minute on a LinuxK cluster of HP Integrity rx5670 servers4.  [ wasn't there a recent Alpha/Rdbi8 benchmark that did in excess of 1MM TPC-C posted here? ]   About Oracle Database 10gc  H Oracle Database 10g is designed to be effectively deployed on everythingG from small blade servers to the biggest SMP servers and clusters of allt> sizes. It features automated management capabilities for easy,L cost-effective operation. Oracle Database 10g's unique ability to manage all? data from traditional business information to XML documents andsF spatial/location information makes it the ideal choice to power Online? Transaction Processing, Decision Support and Content Managementr
 applications.o   About HP Integrity Servers  D HP Integrity servers, based on the industry standard Intel Itanium 2A processor, are known for driving unique customer value from their-L unparalleled flexibility, mission-critical reliability, and highly scalable,I industry-leading performance. There is strong market momentum and growingbD solution ecosystems for Integrity servers due to their capability toK successfully solve the most challenging real-world customer problems acrossi= broad industry segments.  [ guess they never heard of Alpha ]d   About TPC-H0  G TPC-H is a decision support benchmark consisting of a suite of business:J oriented ad-hoc queries and concurrent data modifications. The performanceF metric is called the TPC-H Composite Query-per-Hour Performance MetricL (QphH@Size) and reflects multiple aspects of the capability of the system toE process queries. More information is available at http://www.tpc.org.t   ------------------------------  % Date: Mon, 11 Oct 2004 23:26:29 -0400t* From: "Bill Todd" <billtodd@metrocast.net>> Subject: Re: Wonder how this would work on Alpha/VMS clusters?= Message-ID: <eaudnQdJpKQU0_bcRVn-iA@metrocastcablevision.com>e  . "John Smith" <a@nonymous.com> wrote in message& news:FP6dnX92bYjho_bcRVn-rA@igs.net...; > http://www.hp.com/hpinfo/newsroom/press/2004/041008a.htmlt > C > Oracle and HP Smash World Record for TPC-H 10 TB Data Warehousing 	 Benchmarko >mG > Oracle Database 10g on HP Integrity Superdome cluster tops IBM's bestp resultG > on the TPC-H 10 TB Data Warehousing Benchmark, using 20 percent feweroG > processors and offering more than 33 percent better price/performancea  I Well, yuh.  IBM hasn't bothered to submit a system in this category sincebC the 1.3 GHz POWER4 (not even POWER4+, let alone POWER5) system theyeL submitted almost 2 years ago now.  And indeed, perhaps not too surprisingly,B the newest, top-of-the-line Itanics actually do manage to beat theL 2-generation-old POWER4s, running their aging database software, by the (notF all that impressive, given the nature of the match-up) amounts stated.  A Nothing but the finest HP marchitecture to admire here, as usual.n   ...   C > "This benchmark demonstrates that HP is committed to offering ourT	 customerscF > a business intelligence platform with the greatest power at the bestE > price/performance in the industry," said Rich Marcello, senior viceaA > president and general manager of Business Critical Servers, HP.o  K Surely Rich is practicing for a political career, since he's far too brightSJ not to understand that even a generation-old POWER4+ system similar to itsK POWER4 predecessor would likely match the performance of his prize Itanics,iH while a POWER5 system would leave them so far in the dust that the crowd> would have dispersed long before they crossed the finish line.   >oI > This benchmark adds to the extensive list of HP and Oracle Database 10gfH > performance world records. HP and Oracle were the first to surpass oneJ > million transactions per minute with the TPC-C benchmark by using the HP > Integrity Superdome server  F Indeed they were.  But they held that distinction only for a matter ofA weeks, until POWER4+ bettered that result using only half as many? processors.r  -  and the companies currently hold the overalltH > TPC-C world record result of 1.18 million transactions per minute on a Linuxs) > cluster of HP Integrity rx5670 servers4t  C Ah, yes - the 'cluster cheat' benchmark result.  Most manufacturerso0 sheepishly abandoned that tactic many years ago.  $ .  [ wasn't there a recent Alpha/Rdb: > benchmark that did in excess of 1MM TPC-C posted here? ]  J No:  there was a reference to over 1 million transactions per minute, IIRCI using Oracle's (but it may not have been Rdb's) new row-caching facility,aJ but no indication that they were TPC-C-style transactions (though it would. not be *all* that surprising if they were...).   - bill   ------------------------------  % Date: Tue, 12 Oct 2004 13:35:53 +0930o: From: "Barratt, Chris (FMC)" <Chris.Barratt@fmc.sa.gov.au>> Subject: RE: Wonder how this would work on Alpha/VMS clusters?P Message-ID: <E829CF9B8F94014887EBC61E1951B0CE05637640@sagemshs001.fmc.sa.gov.au>    n' > ..  [ wasn't there a recent Alpha/Rdbe< > > benchmark that did in excess of 1MM TPC-C posted here? ] > @ > No:  there was a reference to over 1 million transactions per  > minute, IIRC6 > using Oracle's (but it may not have been Rdb's) new  > row-caching facility,a< > but no indication that they were TPC-C-style transactions  > (though it would0 > not be *all* that surprising if they were...). >   L I think you are referring to some testing performed by Oracle using Rdb on aI GS1280, where they achieved 1.8 million transactions per second. This wase nothing to with TPC, however. L Rdb's row caching was one of the features being tested, and key to achieving the performance.  * A discussion can be found in this article < http://h71000.www7.hp.com/openvms/journal/v3/oracle_rdb.html   ------------------------------  # Date: Mon, 11 Oct 2004 18:39:04 GMTe& From: jlsue <jefflsxxxz@sbcglobal.net>; Subject: Re: [OT]: Software Disasters Often People Problemsf8 Message-ID: <fcklm0h9u3usf0pol675bakk2r4jundrpf@4ax.com>  K On Tue, 05 Oct 2004 17:00:20 -0400, JF Mezei <jfmezei.spamnot@teksavvy.com>8 wrote:   >John Smith wrote:M >> SAN JOSE, Calif. - New software at Hewlett-Packard Co. was supposed to getiN >> orders in and out the door faster at the computer giant. Instead, a botchedI >> deployment cut into earnings in a big way in August and executives gota	 >> fired.n >hN >My take on this is that this project was a scapegoat to divert attention fromO >other problems. I cannot believe that a corporation such as HP would be stupidbO >enough to pull the plug on the old production system until it is satisfied theF+ >fancy new system is actually working fine.m >tO >I think Carly concucted this excuse. (they probably did have SAP problems, butsG >somehow I doubt it woudl have frozen their production lines to preventlI >deliveries of large enterprise systems. (we 're not talking about volumeo& >manufacturing like wintel crap here).  K While I don't have any access to "inside" info on what the problems were, IeA can confirm that it was a *real* outage... i.e., contrary to youreJ "scapegoat" opinion, there were definitely order processing problems.  AnyG customer who had an order outstanding at the time would confirm this asg well.  g  I My guess is that this kind of thing does impact recognizable revenue, and I thus the bottom line... though I'm admitting right out it's only a guess.iB I can tell you I wasn't surprised to see it affected our earnings.  ( Thankfully these have been resolved now.  J >It is called change management. And many corporations completely lack it.  H Even the best change management programs can run into snags when budgetsK and politics are involved - though I'm not claiming this has anything to dotB with HP's problems noted in the article, as I no idea what exactly	 happened.h   --- jlse0 The preceding message was personal opinion only.6 I do not speak in any authorized capacity for anyone,  and certainly not my employer.- (get rid of the xxxz in my address to e-mail)i   ------------------------------  % Date: Mon, 11 Oct 2004 12:07:00 -0700 # From: "Tom Linden" <tom@kednos.com> ; Subject: Re: [OT]: Software Disasters Often People Problemsc( Message-ID: <opsfpzdyjtzgicya@hyrrokkin>  I On Mon, 11 Oct 2004 18:39:04 GMT, jlsue <jefflsxxxz@sbcglobal.net> wrote:6  0 > On Tue, 05 Oct 2004 17:00:20 -0400, JF Mezei    > <jfmezei.spamnot@teksavvy.com> > wrote: >a >> John Smith wrote:L >>> SAN JOSE, Calif. - New software at Hewlett-Packard Co. was supposed to   >>> geteI >>> orders in and out the door faster at the computer giant. Instead, a  , >>> botchednJ >>> deployment cut into earnings in a big way in August and executives got
 >>> fired. >>C >> My take on this is that this project was a scapegoat to divert    >> attention fromuL >> other problems. I cannot believe that a corporation such as HP would be  	 >> stupideE >> enough to pull the plug on the old production system until it is    >> satisfied the- >> fancy new system is actually working fine.  >>E >> I think Carly concucted this excuse. (they probably did have SAP  e >> problems, butI >> somehow I doubt it woudl have frozen their production lines to preventnK >> deliveries of large enterprise systems. (we 're not talking about volumei( >> manufacturing like wintel crap here). >oG > While I don't have any access to "inside" info on what the problems  d	 > were, I C > can confirm that it was a *real* outage... i.e., contrary to your4J > "scapegoat" opinion, there were definitely order processing problems.    > AnytI > customer who had an order outstanding at the time would confirm this asc > well.   - In some ways, that is even more disconcertingo >0K > My guess is that this kind of thing does impact recognizable revenue, and K > thus the bottom line... though I'm admitting right out it's only a guess. D > I can tell you I wasn't surprised to see it affected our earnings. > * > Thankfully these have been resolved now. > J >> It is called change management. And many corporations completely lack   >> it. > J > Even the best change management programs can run into snags when budgetsL > and politics are involved - though I'm not claiming this has anything to   > doD > with HP's problems noted in the article, as I no idea what exactly > happened.a > 	 > --- jlse2 > The preceding message was personal opinion only.7 > I do not speak in any authorized capacity for anyone,e  > and certainly not my employer./ > (get rid of the xxxz in my address to e-mail)r       -- sC Using Opera's revolutionary e-mail client: http://www.opera.com/m2/l   ------------------------------  # Date: Mon, 11 Oct 2004 20:27:17 GMT & From: jlsue <jefflsxxxz@sbcglobal.net>; Subject: Re: [OT]: Software Disasters Often People Problemst8 Message-ID: <c2rlm09r96r5d9kl5n3kbua05sil2uksfi@4ax.com>  H On Mon, 11 Oct 2004 12:07:00 -0700, "Tom Linden" <tom@kednos.com> wrote:  J >On Mon, 11 Oct 2004 18:39:04 GMT, jlsue <jefflsxxxz@sbcglobal.net> wrote: >r1 >> On Tue, 05 Oct 2004 17:00:20 -0400, JF Mezei  h! >> <jfmezei.spamnot@teksavvy.com>i	 >> wrote:  >> >>> John Smith wrote:hM >>>> SAN JOSE, Calif. - New software at Hewlett-Packard Co. was supposed to  h >>>> getJ >>>> orders in and out the door faster at the computer giant. Instead, a   >>>> botchedK >>>> deployment cut into earnings in a big way in August and executives gotg >>>> fired.  >>> D >>> My take on this is that this project was a scapegoat to divert   >>> attention fromM >>> other problems. I cannot believe that a corporation such as HP would be  f
 >>> stupidF >>> enough to pull the plug on the old production system until it is   >>> satisfied theo. >>> fancy new system is actually working fine. >>>lF >>> I think Carly concucted this excuse. (they probably did have SAP   >>> problems, buteJ >>> somehow I doubt it woudl have frozen their production lines to preventL >>> deliveries of large enterprise systems. (we 're not talking about volume) >>> manufacturing like wintel crap here).p >>H >> While I don't have any access to "inside" info on what the problems  
 >> were, ID >> can confirm that it was a *real* outage... i.e., contrary to yourK >> "scapegoat" opinion, there were definitely order processing problems.     >> AnyJ >> customer who had an order outstanding at the time would confirm this as >> well. >-. >In some ways, that is even more disconcerting  J Not sure how you arrive at this.  The article states it all already, so it" shouldn't be "more disconcerting."  H All I wanted to do was counter JF's (typical) bs with a dose of reality.   --- jls 0 The preceding message was personal opinion only.6 I do not speak in any authorized capacity for anyone,  and certainly not my employer.- (get rid of the xxxz in my address to e-mail)    ------------------------------  % Date: Mon, 11 Oct 2004 17:52:11 -0400a- From: JF Mezei <jfmezei.spamnot@teksavvy.com>s; Subject: Re: [OT]: Software Disasters Often People Problemsr, Message-ID: <416B0085.37C06524@teksavvy.com>   jlsue wrote:J > All I wanted to do was counter JF's (typical) bs with a dose of reality.  N I am sorry, but putting mission critical software into production before it isN properly tested and without the ability to go back if it fails is plain stupidJ and too often associated with wintel newbees who think that are masters of0 windows because they are good at space invaders.  G We are talking about a fortune 500 company jeoperdizing its order entryeN system/production/delivery/billing system for a period measured in days/weeks,N not minutes/hours. We are talking about a corporation that sells to enterpriseN and also wants to sell its own expertise to enterprise. And if customers can'tM place orders, get stuff built/delivered and get invoices, I call that missioniN critical. One wonders why they coudln't have been able to take the orders downL manually ON PAPER and send the paper to the plant by FED EX if they coudln'tN have gone back to the old system. Even airlines are able to process passengersO (albeit at much reduced rate) when reservations systems go down at the airport.s  M Bad financials for a quarter don't need musical chair reorgs or firings as iscE often done to appease the wall street casino analysts. However, sheerhG incompetence in managing such a project which has had such an impact oneM operations should have resulted in some firings. And it isn't just the personsN responsible for the project, but also the person who had chosen that person to head that project.  N One reason this whole story smells digested grass is exactly because HP waitedF for the financials to come out before mentioning this massive failure.  N Donald Trump would have a field day with its trademarked "You're fired" if he+ were to run HP the way he runs his TV show.E   ------------------------------  % Date: Mon, 11 Oct 2004 21:18:16 -0400e# From: "John Smith" <a@nonymous.com> ; Subject: Re: [OT]: Software Disasters Often People Problemsc, Message-ID: <2bCdnbgMquhDrfbcRVn-qQ@igs.net>   jlsue wrote:. > On Tue, 05 Oct 2004 17:00:20 -0400, JF Mezei' > <jfmezei.spamnot@teksavvy.com> wrote:e >s >> John Smith wrote:G >>> SAN JOSE, Calif. - New software at Hewlett-Packard Co. was supposedaC >>> to get orders in and out the door faster at the computer giant.,C >>> Instead, a botched deployment cut into earnings in a big way in $ >>> August and executives got fired. >>A >> My take on this is that this project was a scapegoat to divertwE >> attention from other problems. I cannot believe that a corporationn@ >> such as HP would be stupid enough to pull the plug on the oldB >> production system until it is satisfied the fancy new system is >> actually working fine.k >>C >> I think Carly concucted this excuse. (they probably did have SAP F >> problems, but somehow I doubt it woudl have frozen their productionG >> lines to prevent deliveries of large enterprise systems. (we 're not = >> talking about volume manufacturing like wintel crap here).i >pE > While I don't have any access to "inside" info on what the problems F > were, I can confirm that it was a *real* outage... i.e., contrary toB > your "scapegoat" opinion, there were definitely order processingB > problems.  Any customer who had an order outstanding at the time > would confirm this as well.' >AG > My guess is that this kind of thing does impact recognizable revenue,dF > and thus the bottom line... though I'm admitting right out it's onlyC > a guess. I can tell you I wasn't surprised to see it affected ourn > earnings.t > * > Thankfully these have been resolved now. >hC >> It is called change management. And many corporations completelyr >> lack it.O >tB > Even the best change management programs can run into snags whenF > budgets and politics are involved - though I'm not claiming this hasF > anything to do with HP's problems noted in the article, as I no idea > what exactly happened.    K What happened to 'running in parallel and reconciling' for a month?. That's G the way the rest of the world works with major system cut-overs. Yes itmH means a few more people with green eyeshades and rolled-up sleeves doingK overtime for a month but that's how you catch things that don't work acrosseE end-of-dat/week/month procedures in addition to the day-to-day stuff.    ------------------------------  % Date: Mon, 11 Oct 2004 21:57:00 -0400o- From: JF Mezei <jfmezei.spamnot@teksavvy.com>e; Subject: Re: [OT]: Software Disasters Often People Problemsa, Message-ID: <416B39D8.E2A8EE2D@teksavvy.com>   John Smith wrote:uM > What happened to 'running in parallel and reconciling' for a month?. That'saI > the way the rest of the world works with major system cut-overs. Yes it J > means a few more people with green eyeshades and rolled-up sleeves doingM > overtime for a month but that's how you catch things that don't work acrosszG > end-of-dat/week/month procedures in addition to the day-to-day stuff.O  F Probably the same mentality that results in some thinking that forcingK customers to migrate from alpha to ia64 systems is just a simple recompile.i  M Thing is that everyone knows that SAP is far mroe than the unix equivalent oflL VMSinstall. It means spending mega time setting things up. This si more of aJ year long project than a weekend upgrade. I still fail to see how it couldG have been implemented with such disastrous consequences instead of justu, delaying implementation until SAP was ready.   ------------------------------   End of INFO-VAX 2004.566 ************************