1 INFO-VAX	Thu, 01 Sep 2005	Volume 2005 : Issue 487       Contents: Re: Another PHP solution... % Assigning value of a logical variable  crmpsc INVALIDARG  DCPS generic queue form problem # Re: DCPS generic queue form problem # Re: DCPS generic queue form problem " Re: Dropping TCPIP$SMTP's priority: fwd: Oracle Rdb Field Test 3 for I64 & Alpha Now Available
 Re: Gamers0 Re: Hp Technology forum (New Orleans sept 12-15) Re: Itanium Solutions Alliance RE: Itanium Solutions Alliance Re: Itanium Solutions Alliance Re: Itanium Solutions Alliance Re: Itanium Solutions Alliance" Re: Katrina and disaster tolerance" Re: Katrina and disaster tolerance" Re: Katrina and disaster tolerance" Re: Katrina and disaster tolerance" Re: Katrina and disaster tolerance" Re: Katrina and disaster tolerance" Re: Katrina and disaster tolerance" Re: Katrina and disaster tolerance0 Mixed interconnect VAX VMS 6.2 cluster challenge4 Re: Mixed interconnect VAX VMS 6.2 cluster challenge2 Re: MMS V3.5 v. ODS5: Unexpected case-sensitivity? Re: NTP Under UCX Questions 2 Re: OT: .com spending no longer and porting issues2 Re: OT: .com spending no longer and porting issues2 Re: OT: .com spending no longer and porting issues" Re: Pass arguments to a .com file.# RE: Possible reasons of a file loss # Re: Possible reasons of a file loss # Re: Possible reasons of a file loss # Re: Possible reasons of a file loss  Re: problem with SLS? Setting 'default' DECWindows display with multi-heads on rx26xx C Re: Setting 'default' DECWindows display with multi-heads on rx26xx  Re: SFTP in batch- password? Re: SFTP in batch- password?  Re: simh with logical networking StorageWorks is alive!* Re: traditional DECwindows: save workspace* Re: traditional DECwindows: save workspace Re: VAXen as a plural of VAX Re: VAXen as a plural of VAX Re: VAXen as a plural of VAX Re: VAXen as a plural of VAX Re: VAXen as a plural of VAX Re: VAXen as a plural of VAX  F ----------------------------------------------------------------------   Date: 1 Sep 2005 00:08:26 GMT  From: healyzh@aracnet.com $ Subject: Re: Another PHP solution...+ Message-ID: <df5gpq27n9@enews1.newsguy.com>   2 Michael Austin <maustin@firstdbasource.com> wrote:L > So, I found TR Newsportal WebNewsReader and installed it on my VMS server J > at home.  I made the appropriate edits to the config file to add my ISP    URL please.   0 Also, does this allow you to post, or just read?   	Thanks,	 	    Zane    ------------------------------  $ Date: Thu, 1 Sep 2005 08:43:39 +03000 From: "MUSTAFA ATAKAN" <matakan@inteltek.com.tr>. Subject: Assigning value of a logical variableL Message-ID: <F014DACB8BE63442993543B780A2F0180163BEF7@asteriks.inteltek.ist>    Hi,  B    Is there any way to assign the value of a logical variable to aG symbol variable? Or, can we print the value of a logical variable (I do E not mean "show log xxx", I want to print just the value it contains)?    Best Regards...    ------------------------------    Date: 31 Aug 2005 22:44:29 -0700' From: "saee" <aditi.hatwalne@gmail.com>  Subject: crmpsc INVALIDARGB Message-ID: <1125553469.520804.41620@f14g2000cwb.googlegroups.com>   Hi,   D I am using crmpsc to create section for existing data file.The flags> used are  SEC$M_GBL | SEC$M_EXPREG .It successfully does this.F Then I do delete global section and deltva call. I again do crmpsc forF other file. Return address for this second crmpsc are same as previous one. This is okay. Now consider second situation.. E I do crmpsc first time with same flags. Now I do not do delete global E section and deltva for this memory. I call crmpsc for second time for D different file.For both crmpsc return addresses are different.that'sC fine. Now I call delete global section and deltva for first memory. > Then I call crmpsc for third file...return address for this is8 different.Why it doesn't use address that it just freed?F Now If try to use  return addresses of first crmpsc's for third crmpscE wihout SEC$EXPREG flag..it gives INVALID ARGUMENTS. I don't why it is B giving this error when other parameters are same as previous ones. Whether I am missing anything? TIA, Aditi    ------------------------------  % Date: Wed, 31 Aug 2005 18:08:59 -0700 4 From: Alan Frisbie <Usenet02_REMOVE@Flying-Disk.com>( Subject: DCPS generic queue form problem% Message-ID: <1125536816.334119@smirk>   = I am having a problem with a DCPS generic (ANSI/ASCII) queue. : I'm not sure if I have just failed to RTFM enough, or if I# am just headed down the wrong path.   8 The problem starts with a third-party application, which8 allows the user to select a print queue for reports, but= without any qualifiers/parameters.   These are all 132-column ' reports, so landscape mode is required.   > One printer is an old-fashioned, ASCII-only printer.   We have> a default form associated with its queue which specifies a top& and bottom margin.   No problems here.  = The new printer is Postscript, so we use DCPS (v2.4) with it. < Besides the default Postscript execution queue, we also haveA an ANSI generic queue which feeds to the Postscript queue -- just > as shown in the DCPS manuals.   We use it for all the portrait mode output with no problems.   ? When the user requested a way to print her reports in landscape < mode, I created a new DCPS generic ANSI queue with a default? parameter of PAGE_ORIENTATION=LANDSCAPE.   This works fine, but > with one minor detail: it starts printing too close to the top@ of the page for 3-hole punching.   I tried associating the queueA with the same default form as the old print queue, but was unable 	 to do so.   A The problem is that (as far as I can tell), generic queues cannot F have forms associated with them.   I can manually enter a PRINT /FORM== command and it works just fine.   However, when printing from : the user's application, there is no way to specify a form.  8 Can someone give me a hint of how to solve this problem?   Thanks Alan   ------------------------------  % Date: Wed, 31 Aug 2005 22:20:22 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> , Subject: Re: DCPS generic queue form problem, Message-ID: <4316653B.34158D2D@teksavvy.com>   Alan Frisbie wrote: C > The problem is that (as far as I can tell), generic queues cannot H > have forms associated with them.   I can manually enter a PRINT /FORM=? > command and it works just fine.   However, when printing from < > the user's application, there is no way to specify a form.    ) You can work with layup definition files.    HELP PRINT_PARA LAYUP     Example setup for generic queue:) $ @SYS$STARTUP:DCPS$GENERIC_QUEUE -DRAFT_ ! 	DOCS - ! P1 - Generic queue name 9 	"LN03R_TTB4,LN03R_TTB7" - ! P2 - Execution queue name(s) 8 	"LAYUP=LPS$SINGLEHOLES" ! P3 - Default queue parameters  ( Take a look and DCPS$LAYUP:*.LUP  files.  F The PDF document for DCPS also documents the LAYUP options files where' you define the parameters. (User Guide)   C The layup allows you to specify borders. You can also specify layup G options directly in the print parameters (as opposed to storing them in  a file).   ------------------------------  % Date: Wed, 31 Aug 2005 20:23:40 -0700 4 From: Alan Frisbie <Usenet02_REMOVE@Flying-Disk.com>, Subject: Re: DCPS generic queue form problem% Message-ID: <1125544895.140594@smirk>    JF Mezei wrote:  > Alan Frisbie wrote:  > C >>The problem is that (as far as I can tell), generic queues cannot H >>have forms associated with them.   I can manually enter a PRINT /FORM=? >>command and it works just fine.   However, when printing from < >>the user's application, there is no way to specify a form.  + > You can work with layup definition files.  >...:  > "LAYUP=LPS$SINGLEHOLES" ! P3 - Default queue parameters >...* > Take a look and DCPS$LAYUP:*.LUP  files. >...E > The layup allows you to specify borders. You can also specify layup I > options directly in the print parameters (as opposed to storing them in 
 > a file).  C I haven't tried LAYUP=LPS$SINGLEHOLES yet, but I did try specifying @ borders.   It enclosed the print inside a box, which isn't quite5 what I wanted.   I will try your suggestion tomorrow.   E I'm a bit puzzled why VMS prohibits default forms for generic queues.  Does anyone know why this is?    Thanks,  Alan   ------------------------------  + Date: Wed, 31 Aug 2005 13:12:24 -0500 (CDT) * From: sms@antinode.org (Steven M. Schweda)+ Subject: Re: Dropping TCPIP$SMTP's priority 2 Message-ID: <05083113122437_20404687@antinode.org>  % Date: Sun, 31 Jul 2005 20:05:40 -0400   
    I'm quick.   - From: JF Mezei <jfmezei.spamnot@teksavvy.com>   N > Low and behold, I found that the account TCPIP$SMTP's priority was set to 8.N > So when a spammer wanted to send me stuff, he would monopolise my system (orF > when legitimate users sent legitimate messages that were quite big). > > > I changed the UAF's priority to 3. Seems to be smoother now.  G    I set mine to 4, as I figured that incoming e-mail had as much right ? to use my system as I did, but last Saturday someone apparently C associated with a Web site in China selling various pharmaceuticals F decided to use "antinode.org" as the (forged) sender domain in a large distribution of junk e-mail.  <    Since then, my poor AlpSta 200 4/233 had been pretty muchD overwhelmed, trying to cope with zillions of bounce message deliveryG failure notifications.  They arrive at a much higher rate than the rate E at which a slow CPU can check the sender IP adresses against a fairly  large Bad-Clients list.   H    So, to avoid complete paralysis, I lowered the TCPIP$SMTP priority toB 3, and decreased the SMTP service limit to 8.  Lots of the uselessC bounce messages are still being rejected ("INTERnet ACP SMTP Reject E Request - service limit - [...]"), but the ones (thousands) which get E through sre pretty educational.  Apparently, lots of SMTP servers out G there are perfectly willing to accept e-mail from a hijacked Windows PC E which has no valid reverse DNS look-up.  Lots more fail to reveal the A original sender's IP address in their bounce messages.  Amazing.   Depressing, but amazing.   > Any pitfalls ?  E    SETI@home (also priority 3) now competes with incoming e-mail, but  it's a pretty shallow pit.  H ------------------------------------------------------------------------  4    Steven M. Schweda               (+1) 651-699-98183    382 South Warwick Street        sms@antinode-org     Saint Paul  MN  55105-2547    ------------------------------  % Date: Wed, 31 Aug 2005 12:14:13 -0600 4 From: Norman Lastovica <norman.lastovica@oracle.com>C Subject: fwd: Oracle Rdb Field Test 3 for I64 & Alpha Now Available * Message-ID: <4315F375.321BBB4D@oracle.com>  G          Oracle is pleased to announce the availability of Field Test 3 G     for it's continuing Rdb 7.2 Beta Program.  The field test of Oracle F     Rdb 7.2 is available to run on OpenVMS Industry Standard 64 for HPC     Integrity Servers and for OpenVMS for AlphaServer systems.  The C     kits are available for download on the Rdb 7.2 Beta program web A     site (http://otnbeta.oracle.com/Rdb/index.htm) for registered      Beta customers.      G          The Oracle Rdb Field Test Release 7.2 requires OpenVMS version @     8.2 or higher for Alpha systems or OpenVMS version 8.2-1 forF     Integrity servers.  Customers in the Beta program will be providedH     with an advance copy of OpenVMS version 8.2-1 for Integrity servers.     A          The Oracle Rdb 7.2 Field Test is nearing completion.  We G     anticipate shipping one final Beta update kit before our production      version ships.     H          Oracle expects to release Oracle Rdb Release 7.2 for productionE     use in the fourth quarter of calendar year 2005.  If you have not  done@     so already, you can register for the Rdb 7.2 Beta Program by visitingH     our Beta Program web site (http://otnbeta.oracle.com/Rdb/index.htm).          Ginger Vollmar#     Rdb Beta Programs Administrator   	 - - - - - 0  opinions expressed here are mine and mine alone.  and certainly are not intended in any way to 0  express or represent any opinions or commitment  of oracle corporation.   *  norman lastovica / oracle rdb engineering   ------------------------------   Date: 1 Sep 2005 00:01:02 GMT  From: healyzh@aracnet.com  Subject: Re: Gamers + Message-ID: <df5gbu17n9@enews1.newsguy.com>   . Alex van Denzel <vandenzel@hotmail.com> wrote:K > I always assumed the computer games industry is one of the major reasons  I >   the capabilities of PeeCee hardware is growing in the last ten years   > or so.  J It's not an assumption, it's a fact.  This is one of the reasons I tend toH prefer consoles these days.  Though some genre work better on a ComputerL than on a console.  For some strange reason, "Real-Time Strategy" games seem to be a good example of this.   L With the latest generation of handhelds (GB SP, GB DS, and PSP) I'mfinding I- prefer handhelds over both PC's and Consoles.   B > Like, just after you bought the newest nVATI GeDeon 9900445 XLG I > DeLuxe[1] graphics controller, you can buy games that _need_ the nVATI  D > GeDeon 990446 XLG^2 MaxiPoly[2], that won't be in the shops for a 0 > reasonable price until wednesday next year[3].  H This sounds a little to close to reality.  Only one game has come out inG recent times that I *really* want to play on my PC.  My ageing hardware K meets the spec's, barely.  Instead of text, I see square boxes, and as such  can't actually get anywhere :^(    		Zane   ------------------------------  % Date: Wed, 31 Aug 2005 14:14:00 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> 9 Subject: Re: Hp Technology forum (New Orleans sept 12-15) , Message-ID: <4315F35D.4461DA65@teksavvy.com>   Bob Kaplow wrote: . > Postponed. http://www.hptechnologyforum.com/ > N > I wouldn't bet on a quick reschedule. I'm sure that there are many cancelled@ > NO events competing for limited available resources elsewhere.    A While NO looks terrible now, I think that once flooding subsides, F progress will be very fast and significant. However, I don't think the3 city will be ready to accept tourists before 2006.    8 "reschedule" is more likely "reshedule in another city".   ------------------------------  % Date: Wed, 31 Aug 2005 15:56:57 -0400 ' From: Dave Froble <davef@tsoft-inc.com> ' Subject: Re: Itanium Solutions Alliance 0 Message-ID: <11hc2j1r6nel3c0@corp.supernews.com>   Alan Greig wrote:   K > Watching the presentations it is clear Intel is throwing everything plus   > the kitchen sink at X86-64.   H As a few predicted years ago, when Intel's main business is threatened, % all attention would be focused there.   1 I know Rob, nobody likes to hear "I told you so."    --  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: Wed, 31 Aug 2005 16:12:17 -0400 ' From: "Main, Kerry" <Kerry.Main@hp.com> ' Subject: RE: Itanium Solutions Alliance R Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB6B23B4@tayexc19.americas.cpqcorp.net>   > -----Original Message-----4 > From: Alan Greig [mailto:greigaln@netscape.net]=20  > Sent: August 31, 2005 12:31 PM > To: Info-VAX@Mvb.Saic.Com ) > Subject: Re: Itanium Solutions Alliance  >=20 >=20 >=20 > Dan Foster wrote:  >=20 > > ? > > It seems doubtful to me that they would throw money at a=20  > port to Itanium ; > > if it was expected to 'go away' around the 2007-2008=20  > timeframe, just as# > > they bring out Longhorn Server.  > >=20H > > Although, I guess it's always possible someone was paying them to do@ > > that port. There's certainly historical precedence for that. > >=20D > > Still, by doing such a port, I have to wonder if Microsoft knows# > > something the rest of us don't?  >=20A > Remember the Windows code base is common. It's not so much a=20  > matter of=20A > doing the port as supporting it forward. Windows Server 2003=20  > is already=20 > > on Itanium. The fact that R2 won't be suggests to me that=20 > Longhorn never=20 B > will be either. Of course they can't say that yet. It's worth=20C > remembering that Windows 2000 on Alpha was cancelled after its=20 G > completion but before product ship. Windows 2000 RC2 went out from=20 G > Microsoft before the plug was pulled. Microsoft dropped it because=20 J > Compaq pulled out and made clear to MS that Alpha had no future while=20% > continuing to lie to its customers.  >=20 > --=20  > Alan Greig >=20   Alan,=20  B Can you define what you mean by "Remember the Windows code base is common."  @ As I recall some of the Alpha NT discussions, one of the biggestH challenges with porting Windows NT to Alpha was that a great deal of the1 core code was written in 32bit x86 macro code.=20   H Hence, I suspect optimizing Windows for *any* 64bit platform (as opposedG to just getting it to run) will be an ongoing challenge for Microsoft - $ regardless of the 64bit HW platform.   Regards   
 Kerry Main Senior Consultant  HP Services Canada Voice: 613-592-4660  Fax: 613-591-4477  kerryDOTmainAThpDOTcom (remove the DOT's and AT)=20  4 OpenVMS - the secure, multi-site OS that just works.   ------------------------------  # Date: Wed, 31 Aug 2005 21:10:56 GMT ( From: Alan Greig <greigaln@netscape.net>' Subject: Re: Itanium Solutions Alliance = Message-ID: <A1pRe.40250$5m3.31410@fe1.news.blueyonder.co.uk>    Main, Kerry wrote:   >> > Alan,  > D > Can you define what you mean by "Remember the Windows code base is
 > common."  F If you read Showstopper - the story of Windows NT, it details Cutlers E insistence that as little code as absolutely necessary be written in  I assembler. Microsoft have also stated that a common code pool is used to  : build all the flavours of NT. Or at least that's the goal.  H If you run on Server 2003 but not on the R2 release what's the point of H that if you intend to run on Longhorn? Longhorn should have all/most of D the code in R2 incorporated as I understand it. Longhorn isn't some I complete re-write of Windows it's just the next projected major release.  * Could quite easily be called Windows 2007.  , So let's use the dates instead of codenames:    Windows 2003	 Itanium/X86/X86-64 Windows 2005(R2) X86/X86-64  Windows 2007 (Longhorn) ?   G It is worth emphasising here that R2 *is* supported on the X86-64. Are  I you really trying to tell me you can make sense of this in a manner that  * shows a rosy future for Windows on Itanic?  J > Hence, I suspect optimizing Windows for *any* 64bit platform (as opposedI > to just getting it to run) will be an ongoing challenge for Microsoft - & > regardless of the 64bit HW platform.  C Which is one reason I can't see the logic in throwing R2 code into  F Longhorn without the chance to test it on Itanic with a user release. D Unless you never intend to release again on Itanic that is. Then it  makes perfect sense.   --  
 Alan Greig   ------------------------------  % Date: Wed, 31 Aug 2005 17:13:35 -0400 2 From: "Paul A. Jacobi" <Paul.Jacobi@nospam.hp.com>' Subject: Re: Itanium Solutions Alliance , Message-ID: <43161e22$1@usenet01.boi.hp.com>  6 "Alan Greig" <greigaln@netscape.net> wrote in message 7 news:JUjRe.39331$5m3.29935@fe1.news.blueyonder.co.uk...   * > Itanic did get a brief 10 minute mention  J Did you notice the OpenVMS logo on one of Pat Gelsinger slide?  I hope we G can somehow get him to do a demo of OpenVMS on Itanium at IDF sometime.     1 http://www.intel.com/idf/us/fall2005/webcast.htm#      Paul A. Jacobi OpenVMS Systems Group 
 Nashua, NH   ------------------------------    Date: 31 Aug 2005 23:14:53 -0500+ From: young_r@encompasserve.org (Rob Young) ' Subject: Re: Itanium Solutions Alliance 3 Message-ID: <GMnMIBwulnMx@eisner.encompasserve.org>   h In article <JUjRe.39331$5m3.29935@fe1.news.blueyonder.co.uk>, Alan Greig <greigaln@netscape.net> writes: >  >  > JF Mezei wrote:  > h >> http://news.com.com/Itanium+allies+to+pool+development+efforts/2100-1006_3-5844877.html?tag=nefd.lede >>  C >> Intel is getting its IA64 customers to fund an Itanium Solutions 5 >> Alliance to help promopte the struggling alliance.  >>  K >> The article then moves on to more familiar territory, outlining why IA64 G >> just hasn't caught on. (It isn't just the inquirer that has negative  >> spin on IA64).  > I > The entire IT news industry knows that Itanium is critically ill. They  D > only differ on whether it's terminal or not. Most have just about G > forgotten about it. It was extremely telling that Intel's CEO didn't  G > mention the Itanic at all during his 1 hour long keynote at IDF last  I > week - "In terms of Itanium it was time budget. I was given 59 minutes  ; > and I hit 59 minutes." I read that as "That's all folks."  > F > The presentations are all available for replay on Intel's web site. F > Worth checking out. Itanic did get a brief 10 minute mention on day K > three. Very, very tellingly nothing was even hinted at beyond Montecito.    + 	Yawn...  stretch, harrumphh, harrumphh....   6 	Yes... very very telling nothing was hinted at beyond. 	Montecito.  Very telling indeed.  Pleeeze....  > 	If you're going to FUD something up in the Internet/Blog era,5 	you are going to have to do a lot better than that.    2 http://news.zdnet.co.uk/0,39020330,39215762,00.htm   Stephen Shankland 
 CNET News.com  August 31, 2005, 13:30 BST  J "The industry has lined up . all but one . with Itanium as the platform ofH choice for RISC replacement and mainframe platforms for the future," PatN Gelsinger, general manager of Intel's digital enterprise group, said last week. at the Intel Developer Forum in San Francisco.  K The lone holdout Gelsinger referred to is IBM, whose Power processor is the K second horse in what Intel likes to call a two-horse race, but it should be L noted that Sun also shuns Itanium in favour of high-end systems built aroundF Sparc processors.    [But UltraSparc isn't even worth mentioning, note Sun's continued slide]   [snip]  K Intel's Itanium plans still extend years into the future. Next up will be a J model code-named Montecito, due to be released by the end of this year. ItO employs dual cores, an approach already used by high-end server chips from Sun,  HP and IBM.   M "Montecito will deliver a huge step up in performance, more than doubling the O performance capabilities of today's Madison family of products," Gelsinger said 
 last week.  N After Montecito comes a close relative called Montvale, in 2006. Then in 2007,O Intel plans to introduce a four-core Itanium called Tukwila, a design that will ; be succeeded by another four-core chip, code-named Poulson.     J >> The big question now is whether this effort to keep IA64 alive is IntelK >> driven, or driven by HP and possibly SGI to help justify their continued  >> reliance on that chip.  > 6 > SGI seem to be implementing plan "B" move to X86-64. >    	Sure.   				Rob    ------------------------------    Date: 31 Aug 2005 12:32:39 -0700# From: "WhoDat?" <whohe@whoever.com> + Subject: Re: Katrina and disaster tolerance A Message-ID: <1125516759.214460.7030@g49g2000cwa.googlegroups.com>    Keith Parris wrote:  > O'Brien Paddy wrote:I > > Well JF, you can probably bet your last dollar that anything that VMS M > > covers will never be told to the world.  (Just pre-empting John Smith :-)  > . > That choice is up to HP's customers, not HP. > G > Commerzbank chose to share their 9/11 story publicly, for which we're  > all grateful -- see : > http://h71000.www7.hp.com/openvms/brochures/commerzbank/ > and   E In which other media is this article published? Brochures on HP's web ( site don't count as "telling the world".  E > eSpeed/Cantor-Fitzgerald was another VMS-based 9/11 survivor -- see 3 > http://www.eweek.com/article2/0,3959,40716,00.asp   @ OpenVMS clustering is not mentiond. OpenVMS is mentioned once as follows:   ##C Sun Microsystems Inc. servers running Windows NT, Alpha/OpenVMS and  Solaris.   ##  C So, Alpha/OpenVMS runs on Sun Microsystems servers? That's the only 4 mention of VMS and they got it wrong! Good Grief!!!!    B > http://www.baselinemag.com/print_article2/0,2533,a=17022,00.asp.  A OpenVMS is not even mentioned, nor is even the word "clustering".   F Reading this article, I don't see that OpenVMS had anything to do with
 the recovery.    qoute: ##> The mirror site and the World Trade Center were connected by aB high-speed optical line, over which eSpeed linked the storage area@ networks at each site. Sybase data-replication software mirroredE critical databases between the sites. Half of the company's Microsoft E Exchange e-mail servers were also located full-time in Rochelle Park.    ##
 (later on) ##A Compaq delivered 100 desktop computers at 2 a.m. on Wednesday and E Verizon expedited the installation of voice lines and the transfer of F some of eSpeed's digital circuits. Cisco provided a phone system based; on Internet protocols, in a hurry. Microsoft had an NT team F continuously on hand, as eSpeed's server and desktop maintenance group had been especially hard-hit.    ##  C Compaq desktops, Verizon communications, Cisco phones and Microsoft 6 servers did the job! I guess that's the way to go, eh?  > > Capitol Health Authority shared their experience here -- seeH > http://groups.google.com/group/comp.os.vms/msg/2e4deb37038472a2?hl=en&' > Merrill Corp. also shared here -- see H > http://groups.google.com/group/comp.os.vms/msg/27d8c99a21665015?hl=en&    A Really! Didn't know c.o.v. was so vital to HP's OpenVMS marketing  efforts (oxymoron alert!).  G I know you're trying, Keith Parris, but the examples you give argue for G the side you are contesting. If only HP tried half as hard as you do to + promote VMS, you wouldn't even have to try.    ------------------------------  % Date: Wed, 31 Aug 2005 16:01:22 -0400 ' From: Dave Froble <davef@tsoft-inc.com> + Subject: Re: Katrina and disaster tolerance 0 Message-ID: <11hc2raeh18a4af@corp.supernews.com>   Keith Parris wrote:  > O'Brien Paddy wrote: > I >> Well JF, you can probably bet your last dollar that anything that VMS  L >> covers will never be told to the world.  (Just pre-empting John Smith :-) >  > . > That choice is up to HP's customers, not HP.  A It's up to HP to ask their customers to do so.  In the case of a  G disaster, HP should be aware of the customer's problems, since it will  , probably require support, new hardware, etc.  H Is somebody checking the list of customers who potentially could have a ; problem?  Will somebody be asking for the story to be told?    --  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: 31 Aug 2005 15:27:01 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen) + Subject: Re: Katrina and disaster tolerance 3 Message-ID: <L0HeoX9jcV4a@eisner.encompasserve.org>   g In article <1125516759.214460.7030@g49g2000cwa.googlegroups.com>, "WhoDat?" <whohe@whoever.com> writes:   G > In which other media is this article published? Brochures on HP's web * > site don't count as "telling the world".  0 Who said that ?   Anonymous posts don't "count".   ------------------------------  % Date: Wed, 31 Aug 2005 14:17:47 -0700  From: Z <Z@no.spam> + Subject: Re: Katrina and disaster tolerance + Message-ID: <%7pRe.17845$ih4.7387@fe02.lga>    O'Brien Paddy wrote:D > Are there any well known VMS shops in the area ? Do they have goodG > disaster tolerance plans ? It might make for interesting VMS material + > for the technical journal, marketing etc.  ... H > This is definitely an area where you not only need disaster tolerance,H > but remote sites and remote backups, with remote being more than 100km( > away. (aka: an area where VMS is best)  9 Agreed, but that doesn't mean you need clustering or VMS.   B Our Fault Tolerant subsystem (we have 3 system in the area hit by H Katrina) is Windows-based and uses a replication server supplied by the  database vendor.  D You need to stop associating FT with VMS and clustering. That was a > legitimate assumption maybe 10-15 years ago, but not any more.   ------------------------------  # Date: Wed, 31 Aug 2005 22:31:17 GMT 1 From: Keith Parris <keithparris_NOSPAM@yahoo.com> + Subject: Re: Katrina and disaster tolerance 3 Message-ID: <VcqRe.11516$0B7.8594@news.cpqcorp.net>    WhoDat? wrote: > Keith Parris wrote: G >>Commerzbank chose to share their 9/11 story publicly, for which we're  >>all grateful -- see : >>http://h71000.www7.hp.com/openvms/brochures/commerzbank/ > G > In which other media is this article published? Brochures on HP's web * > site don't count as "telling the world".   Here are some: ComputerWorld:  [ http://www.computerworld.com/securitytopics/security/recovery/story/0,10801,101249,00.html    D Datamation: http://itmanagement.earthweb.com/erp/article.php/3380341H Enterprise IT Planet, Enterprise Storage Forum, and Enterprise Security  Planet: J http://www.enterpriseitplanet.com/storage/features/article.php/3396941 andN http://www.enterprisestorageforum.com/continuity/features/article.php/3403141  and E http://www.esecurityplanet.com/trends/print.php/3396941, respectively   I Commerzbank made a public presentation (I think it was at CETS 2002, but   may have been HP World 2003): B http://www.encompassus.org/proceedings/abstract.cfm?ID=2254&conf=2  B In December of 2002, I had the privilege of hearing Zahid Kazi of H Commerzbank give a presentation on their experience at the Local User's % Group (LUG) meeting in New York City.   I Even the HP brochure about Commerzbank seems to have wider exposure than  C one might guess. Here's a reference in regard to railroads, of all  6 things: http://www2.usenetarchive.org/Dir55/File5.html  E >>eSpeed/Cantor-Fitzgerald was another VMS-based 9/11 survivor -- see 3 >>http://www.eweek.com/article2/0,3959,40716,00.asp  > C > OpenVMS clustering is not mentiond. OpenVMS is mentioned once  as   > follows:   >  > ## F  > Sun Microsystems Inc. servers running Windows NT, Alpha/OpenVMS and Solaris.  >  > ##   >F  > So, Alpha/OpenVMS runs on Sun Microsystems servers? That's the only7  > mention of VMS and they got it wrong! Good Grief!!!!   @ You'd didn't give the complete quote. The article actually said:I "The north tower impact took out eSpeed's operations on the 103rd floor,  H which included some 250 Compaq Computer Corp. and Sun Microsystems Inc. 7 servers running Windows NT, Alpha/OpenVMS and Solaris."   G Given the entire quote, it's obvious they meant some of the 250 Compaq  2 Computer Corp. servers were running Alpha/OpenVMS.  G VMS was certainly pivotal at eSpeed / Cantor-Fitzgerald. (I've visited  D at their site; I know.) Soon after 9/11 I wrote this in an internal G forum: "There's a very moving video about the Cantor-Fitzgerald family  I of companies, and their heroic recovery after losing almost 2/3 of their  G employees on 9/11, at http://www.espeed.com/, under Our Story --> View  F Our Stories. You can spot some VMS displays in the video if you watch I carefully, and if your vision isn't blurred too much by the tears." That  F video doesn't appear to be on their website now, but it was certainly  moving.    ------------------------------    Date: 31 Aug 2005 15:26:24 -0700# From: "WhoDat?" <whohe@whoever.com> + Subject: Re: Katrina and disaster tolerance B Message-ID: <1125527184.668112.63230@o13g2000cwo.googlegroups.com>   Larry Kilgallen wrote:i > In article <1125516759.214460.7030@g49g2000cwa.googlegroups.com>, "WhoDat?" <whohe@whoever.com> writes:  > I > > In which other media is this article published? Brochures on HP's web , > > site don't count as "telling the world". >  > Who said that ?   @ I wrote that. I didn't say anything, but my lips might have been moving.     > Anonymous posts don't "count".   Do too. :-)   D Anyway, the subject of disaster recovery is being discussed now moreG than usual in various media (and will be for a little while longer) and C getting the VMS word out to the public seems like a good idea to me , (whoever I am). What do you think? Bad idea?   ------------------------------    Date: 31 Aug 2005 15:55:01 -0700# From: "WhoDat?" <whohe@whoever.com> + Subject: Re: Katrina and disaster tolerance C Message-ID: <1125528901.110834.288520@g47g2000cwa.googlegroups.com>    Keith Parris wrote:  > WhoDat? wrote: > > Keith Parris wrote: I > >>Commerzbank chose to share their 9/11 story publicly, for which we're  > >>all grateful -- see < > >>http://h71000.www7.hp.com/openvms/brochures/commerzbank/ > > I > > In which other media is this article published? Brochures on HP's web , > > site don't count as "telling the world". >  > Here are some: > ComputerWorld:\ > http://www.computerworld.com/securitytopics/security/recovery/story/0,10801,101249,00.html  E Okay, I remember that one a few months ago. Looks like no one from HP F took the publicity ball and ran with it when the opportunity presented itself, though.    > F > Datamation: http://itmanagement.earthweb.com/erp/article.php/3380341I > Enterprise IT Planet, Enterprise Storage Forum, and Enterprise Security 	 > Planet: L > http://www.enterpriseitplanet.com/storage/features/article.php/3396941 andO > http://www.enterprisestorageforum.com/continuity/features/article.php/3403141  > and G > http://www.esecurityplanet.com/trends/print.php/3396941, respectively  >   G There are reporters who have beat the drum occasionally. Anyone from HP C give them any positive feedback and offer to provide the name of an C expert they could call anytime they had need of such? That would be  good if they had.   J > Commerzbank made a public presentation (I think it was at CETS 2002, but > may have been HP World 2003): D > http://www.encompassus.org/proceedings/abstract.cfm?ID=2254&conf=2 >    Didn't see it. Sorry.   C > In December of 2002, I had the privilege of hearing Zahid Kazi of I > Commerzbank give a presentation on their experience at the Local User's ' > Group (LUG) meeting in New York City.  > J > Even the HP brochure about Commerzbank seems to have wider exposure thanD > one might guess. Here's a reference in regard to railroads, of all8 > things: http://www2.usenetarchive.org/Dir55/File5.html > G > >>eSpeed/Cantor-Fitzgerald was another VMS-based 9/11 survivor -- see 5 > >>http://www.eweek.com/article2/0,3959,40716,00.asp  > > E > > OpenVMS clustering is not mentiond. OpenVMS is mentioned once  as 
 >  > follows:  >  > >  > ## H >  > Sun Microsystems Inc. servers running Windows NT, Alpha/OpenVMS and
 > Solaris. >  > >  > ##  >  >H >  > So, Alpha/OpenVMS runs on Sun Microsystems servers? That's the only9 >  > mention of VMS and they got it wrong! Good Grief!!!!  > B > You'd didn't give the complete quote. The article actually said:J > "The north tower impact took out eSpeed's operations on the 103rd floor,I > which included some 250 Compaq Computer Corp. and Sun Microsystems Inc. 9 > servers running Windows NT, Alpha/OpenVMS and Solaris."  > H > Given the entire quote, it's obvious they meant some of the 250 Compaq4 > Computer Corp. servers were running Alpha/OpenVMS. >   G Given the entire quite, it isn't clear at all what was running what and F what played what role in the D.R. Which one was running Alpha/OpenVMS?( And, just what the heck is that, anyway?  H > VMS was certainly pivotal at eSpeed / Cantor-Fitzgerald. (I've visitedE > at their site; I know.) Soon after 9/11 I wrote this in an internal H > forum: "There's a very moving video about the Cantor-Fitzgerald familyJ > of companies, and their heroic recovery after losing almost 2/3 of theirH > employees on 9/11, at http://www.espeed.com/, under Our Story --> ViewG > Our Stories. You can spot some VMS displays in the video if you watch J > carefully, and if your vision isn't blurred too much by the tears." ThatG > video doesn't appear to be on their website now, but it was certainly 	 > moving.   D When *anything* has played not just a pivotal role in a major event,E but *the* pivotal role, there should be more than a casual mention of A that *anything* and a few screen shots in the background. Sun and B Microsoft certainly got their share of mention about their role in this, wouldn't you say?   ? You know VMS is good; I know VMS good. I hope you know I wasn't D attacking you (although, who the hell I am isn't clear to me anymore since I've become "anonymous").    ------------------------------  % Date: Wed, 31 Aug 2005 19:32:06 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> + Subject: Re: Katrina and disaster tolerance * Message-ID: <43163DD6.711418@teksavvy.com>   WhoDat? wrote:   F > Anyway, the subject of disaster recovery is being discussed now moreI > than usual in various media (and will be for a little while longer) and E > getting the VMS word out to the public seems like a good idea to me . > (whoever I am). What do you think? Bad idea?    B bad idea. VMS management and HP apologists will point out that VMSB doesn't need marketing and that money spent on amrketing would notN provide financial results and that it is doing well without marketing anyways.   ------------------------------    Date: 31 Aug 2005 13:46:54 -0700' From: "syslost" <wm.reynolds@gmail.com> 9 Subject: Mixed interconnect VAX VMS 6.2 cluster challenge C Message-ID: <1125521214.303523.283590@g47g2000cwa.googlegroups.com>   B We have a VAX VMS 6.2 mixed interconnect cluster configured as so: Site 1 is 25 miles from Site 2  1     Site 1                                 Site 2 6  x______x______x__100MBTelnet/DECNet___x_____x_______xB  | x____|_x____|_x_____________________|_x___|_x_____|_x 100MB SCS8  | |    | |    | |                     | |   | |     | |8 ----   ----   ----                    ----   ----   ----9 |v1 |  |v3 |  |v4 |                   |v2 |  |v4 |  |v6 | 8 ----   ----   ----                    ----   ----   ----7   |      |      |                       |      |      | 7   |      |      |                       |      |      | 7   |      |      |                       |      |      | 7   |      |      |                       |      |      | <  ------------------ Star coupler       ------------------ SC7   |     |      |                         |     |      | 8 ----  ----   ----                      ----  ----   ----C |hsj| |hsj|  |hsj| 24 HSJ 50           |hsj| |hsj|  |hsj| 16 HSJ 50 9 |   | |   |  |   |                     |   | |   |  |   | 8 ----  ----   ----                      ----  ----   ----: Lots of rz40's, all                    Lots of rz40's, all8 dual ported, most                      dual ported, most- shadowed                               shadow     B The LavC circuit is connected via Cisco routers 2600 and 2600. The switchesE are 4006's. The 4006's have gig connections to our 6500 which run the  VTP # domain and the datacenter switches.    Stats from testing LAN: A Sending 5000, 1400-byte ICMP Echos to 10.84.200.124, timeout is 2  secondF !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!  A Success rate is 100 percent (5000/5000), round-trip min/avg/max = 	 4/4/20 ms   E When we try to set up shadow sets with one member at site 1 the other  atF site 2, the shadow copy takes about 21 hours for a RZ40 set of drives.  E That's only allowing the one shadow copy at a time (either direction)   E Using show cluster the cr_waits show about the same value whether the  two B sites are joined as one cluster, or the clusters are booted as two separate clusters.  C Each site has its own system device shared among the three nodes at 
 that site.  G The site 1 cluster is the production side of the world, site 2 is to be  the DisasterE Tolerant site.  When the two clusters are joined as one, even without  shadowing disks G between sites, the performance is noticeably slower.  With Nodes timing  out, theD rest of the cluster thinking the Node left the cluster, but the node that "left"  the cluster is still running.   D Copying 70,000 blocks from site 2 to a shadow set with one member at site 1 to 4 hours.F While copying 70,000 blocks from Site 1 the shadow set with its member at site 1 took 3~4 minutes.   On all node system params:! RECNXINTERVAL                  60 ! TIMVCFAIL                    1600 > ALLOCLASS               1 for Site 1, 2 for Site 2.  But we're=                         not using these drive for production, .                         just to test shadowing! SCSBUFFCNT                   1112 ! SCSCONNCNT                     80   B Performance graphs show the routers as running at 80 to 90% of the router CPU.   D What kind of performance should we expect to see from shadow copies?  ? We've called HP, the SCSBUFFCNT and SCSCONNCNT are set to their  recommendations.  < Need suggestions of what to look at and modify to get better performance.   thanks
 Bill Reynolds    ------------------------------  # Date: Wed, 31 Aug 2005 23:30:20 GMT 1 From: Keith Parris <keithparris_NOSPAM@yahoo.com> = Subject: Re: Mixed interconnect VAX VMS 6.2 cluster challenge 2 Message-ID: <g4rRe.11528$BD7.274@news.cpqcorp.net>   syslost wrote:D > We have a VAX VMS 6.2 mixed interconnect cluster configured as so:  G I presume it's not easy or you would have already done so, but VMS 7.3  C or later versions would allow you to take better advantage of your  G multiple links between sites. In VMS releases before 7.3, VMS can only  C transmit to a given remote node over 1 path at a time; with 7.3 or  H above, it can transmit on both paths in parallel to a given remote node.  I Given that you have 3 nodes per site, things aren't as bad as they would  H be with, say, a single node per site, because chances are good the load G will tend to be shared somewhat equally across the two available links.   H Have you seen the material on VMS DT clusters and Shadowing starting at E http://www2.openvms.org/kparris/ ? If not, you might find it helpful.     > Site 1 is 25 miles from Site 2 ...   > Stats from testing LAN:D  > Sending 5000, 1400-byte ICMP Echos to 10.84.200.124, timeout is 2	  > second I  > !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!   >D  > Success rate is 100 percent (5000/5000), round-trip min/avg/max =  > 4/4/20 ms  I Do you know your actual circuit path length(s)? Ping measures latency at  I the IP level. 4 milliseconds round-trip latency would be more indicative  : of something like 250 miles distance rather than 25 miles.  I In addition to Ping, I'd also recommend running LOCKTIME.COM from the V6  G Freeware CD, [KP_LOCKTOOLS] directory, to check the round-trip latency  F for lock requests, which use the SCS protocol, to see if that differs . from (i.e. is even worse than) the IP results.  3 >     Site 1                                 Site 2 8 >  x______x______x__100MBTelnet/DECNet___x_____x_______xD >  | x____|_x____|_x_____________________|_x___|_x_____|_x 100MB SCS: >  | |    | |    | |                     | |   | |     | |: > ----   ----   ----                    ----   ----   ----; > |v1 |  |v3 |  |v4 |                   |v2 |  |v4 |  |v6 | : > ----   ----   ----                    ----   ----   ----  H I notice two LAN adapters per node, and two inter-site links, which are D good. Are the two links bridged to one another, or do they form two H separate independent LANs? The reason I ask is that if they are bridged H together, the Spanning Tree protocol may be turning off one of your two I inter-site links, effectively cutting your inter-site bandwidth in half.  I If the two inter-site links are isolated from one another, there are two  	 benefits: E 1) VMS can use both paths at once, since Spanning Tree won't tend to  ( disable one of the inter-site links, and@ 2) If the equipment used to form each of the two LANs is really G independent and separate, then because the Spanning Trees on each path  E are completely independent, you can run RECNXINTERVAL at the default  I value of 20 seconds. If the two LANs are part of the same Spanning Tree,  I then whenever the Spanning Tree has to be reconfigured, that can take 35  G seconds or so, longer than the default value for RECNXINTERVAL (unless  I you use the new 802.1w Rapid Spanning Tree convergence algorithm instead  " of the original 802.1d algorithm).  G Since you have redundancy in your LAN, I'd highly recommend setting up  G LAVC$FAILURE_ANALYSIS. EDIT_LAVC.COM from the V6 Freeware CD directory  F [KP_CLUSTERTOOLS] can do the heavy lifting for you in setting this up.I It will generate OPCOM messages if LAN adapters, switches, or links fail.   : > ----  ----   ----                      ----  ----   ----E > |hsj| |hsj|  |hsj| 24 HSJ 50           |hsj| |hsj|  |hsj| 16 HSJ 50 ; > |   | |   |  |   |                     |   | |   |  |   | : > ----  ----   ----                      ----  ----   ----  D Are you using write-back caching? That can help Shadowing full-copy  performance.  D > The LavC circuit is connected via Cisco routers 2600 and 2600. The
 > switchesG > are 4006's. The 4006's have gig connections to our 6500 which run the  > VTP % > domain and the datacenter switches.   G Are you using something like L2TP Tunneling between the Cisco routers,   or just a simple VLAN setup?  E I've seen a lot of lost packets with Cisco 4000-series switches. Use  E REXMT.COM from the V6 Freeware [KP_CLUSTERTOOLS] directory to see if  E you're seeing retransmitted packets because packets are getting lost   and/or corrupted.   G > When we try to set up shadow sets with one member at site 1 the other  > atH > site 2, the shadow copy takes about 21 hours for a RZ40 set of drives. > G > That's only allowing the one shadow copy at a time (either direction)   A That sounds extremely high, especially for a 9 GB drive with two   100-megabit inter-site links.   G > Using show cluster the cr_waits show about the same value whether the  > two D > sites are joined as one cluster, or the clusters are booted as two > separate clusters.  @ $SHOW CLUSTER and the CR_WAITS field displays credit waits on a G per-SYSAP basis. If credit waits are on the VMS$VAXcluster SYSAP, that  B would typically be involved with traffic like locking; the SYSGEN H parameter CLUSTER_CREDITS can be raised in that case. For MSCP-serving, I credit waits would show up on the MSCP$DISK SYSAP. The associated SYSGEN  H parameter would be MSCP_CREDITS, on the MSCP-serving node. Also look at F $MONOTIR MSCP and $SHOW DEVICE/SERVED and check for buffer waits; you $ may also need to raise MSCP_BUFFERS.  E > Each site has its own system device shared among the three nodes at  > that site.  I Without an inter-site Fibre Channel link, that's what you must do, since  I you can't share a shadowed system disk across sites if you must use MSCP  F Serving to access the remote shadowset members, and having a separate F system disk per site is a good idea even if you DO have an inter-site C Fibre Channel link, because it can help you survive things like an  - errant $DELETE [...]*.*;* on one system disk.   I > The site 1 cluster is the production side of the world, site 2 is to be  > the DisasterG > Tolerant site.  When the two clusters are joined as one, even without  > shadowing disks 6 > between sites, the performance is noticeably slower.  D One thing that might contribute to this is if you're doing a lot of I directory lookups. If all the nodes have equivalent values of LOCKDIRWT,  D then as soon as the remote nodes join the cluster, they would start E handling about half of the directory lookups, statistically speaking.   ' This would be exacerbated by a bad LAN.    > With Nodes timing out, theF > rest of the cluster thinking the Node left the cluster, but the node+ > that "left" the cluster is still running.    Why should nodes time out?  C I'd suspect some sort of LAN problem, where you're getting lost or   corrupted packets.  F > Copying 70,000 blocks from site 2 to a shadow set with one member at > site 1 to 4 hours.H > While copying 70,000 blocks from Site 1 the shadow set with its member > at site 1 took > 3~4 minutes.  6 I assume this is while the 2 sites are in one cluster. Sounds broken to me.0 I think this again points to a probable bad LAN.   > On all node system params:# > RECNXINTERVAL                  60 # > TIMVCFAIL                    1600 @ > ALLOCLASS               1 for Site 1, 2 for Site 2.  But we're? >                         not using these drive for production, 0 >                         just to test shadowing$  > SCSCONNCNT                     80   These all look acceptable.  C I assume all the HSJs at Site 2 have an allocation class of 2 also.   # > SCSBUFFCNT                   1112   H A shortage of this would show up as BDT_WAIT counts under $SHOW CLUSTER.  D > Performance graphs show the routers as running at 80 to 90% of the
 > router CPU.   H Sounds like potential saturation there. In general, for any resource, I 7 try to keep resource utilization below 70% if possible.   F > What kind of performance should we expect to see from shadow copies?  E I'd guess a 9 GB drive should be able to shadow-copy in no more than  H 2-to-4 hours in your configuration, and you should be able to full-copy B many such shadowsets (probably all of your shadowsets at once) in F parallel, given your inter-site link bandwidth. (You'd likely have to J raise SHADOW_MAX_COPY on all nodes to shadow copy all shadowsets at once.) --- H One side-effect of a 2-site configuration like this is that some of the H traffic which used to go on CI in a single-site environment will now go H over Fast Ethernet, and all the LAN interrupts go to the Primary CPU in I   6.2. So one more thing to watch out for is Primary CPU interrupt-state  > saturation. $MONITOR MODES/CPU=0/ALL allows you to check that.   ------------------------------  + Date: Wed, 31 Aug 2005 19:18:16 +0000 (UTC) . From: klewis@LUMINA.MITRE.ORG (Keith A. Lewis); Subject: Re: MMS V3.5 v. ODS5: Unexpected case-sensitivity? . Message-ID: <df4vpo$n0g$1@newslocal.mitre.org>   sms@antinode.org (Steven M. Schweda) writes in article <05082506523069_20A0026A@antinode.org> dated Thu, 25 Aug 2005 06:52:30 -0500 (CDT):* >: "John E. Malmberg" <wb8tyw@qsl.network>E >> Work around #1 you have found, rename the extension to upper case.  > H >   Too unpleasant.  Defeats the purpose of using an ODS5 file system to >preserve case.   G Not really.  You can preserve the names and just change the extensions. 5 But then if you move it to Unix you'll have problems.   L >> Work around #2 is to use explicit rules in the .MMS file for each target  >> with it's sources.   D Maybe adding some rules to MMS$RULES?  (I couldn't find the original@ MMS$RULES file in the product distribution, so I copied from theL documentation when I needed to change things.)  I'm thinking of source rules such as .c.OBJ, .h.TLB, etc.  0 --Keith Lewis              klewis {at} mitre.org> The above may not (yet) represent the opinions of my employer.   ------------------------------  % Date: Wed, 31 Aug 2005 17:58:51 -0400 $ From: "Chris" <an.other@not_here.ca>$ Subject: Re: NTP Under UCX Questions9 Message-ID: <BKpRe.4341$884.616763@news20.bellglobal.com>   A One more reason that UCX "is less than optimal" (read .... sucks)     0 "Hoff Hoffman" <hoff@hp.nospam> wrote in message- news:wc3Re.11425$4P6.6606@news.cpqcorp.net... G > In article <11h96j3d1bo67ab@corp.supernews.com>, nospam@nohost.no.net  writes: > > :Our setup is a VMS cluster running VMS 7.1 and UCX ver 4.1. > A >   Both of which are ancient.  NTP has seen significant upgrades @ >   in more recent TCP/IP Services versions.  V5.4 or V5.5 wouldB >   be better and more current choices, IIRC.  V5.4 and later have >   NTP V4 implementations.  > A > :I'd like to know if it's possible to set the maximum amount of J > :difference between the system time and the server time so that UCX doesG > :not change the time if it is greater than the set maximum. I believe J > :that is referred to as aperture but I haven't seen a means to set this. > : >   There is no means to set this value.  It's hard-coded.    K Can be specified if you were using a 3rd party TCP/IP product (eg. TCPware) H (I believe the logical value is called "TCPWARE_NTP_WAYTOOBIG" .... nice terminology)     > A >   If you want to see the basic NTP code, see the udel reference 7 >   implementation of NTP V4.  (The FAQ has a pointer.)  > I > :Is it also possible to configure UCX so that it does not slew or drift H > :the clock but makes instaneous corrections to the system time when it) > :determines that there is a difference?  > C >   That's not permissible under NTP.  NTP drifts.  If you want the @ >   big change, shut off NTP and use ntpdate (IIRC) to reset the	 >   time.  >   D Again slew versus step is configurable with other products (TCPware)    E >   Conversely, twice a year, the calculated time does see a one-hour D >   change for those political geographies that can and do celebrateG >   the Daylight Saving Time (DST) "fun".  That too is how the standard E >   says the environment should operate.  (NTP doesn't handle the DST D >   switch-over, that's handled automatically in more recent OpenVMSB >   versions -- the AUTO_DLIGHT_SAV mechanism is another potential= >   reason to upgrade from V7.1 to V7.3 or later, obviously.)  > D >   The usual reason I've seen these cases are tools or applicationsD >   that are using time for a database key or such, but that are not >   using UTC. >  > -- > F >   There's a whole new section of UTC calls available in OpenVMS V7.3E >   and later, see the more current utility routines reference manual C >   for documentation on the DTSS UTC calls.  (The documentation on 4 >   these calls is in the V8.2 manuals, as a start.) >  > -- > D >   I'm updating the FAQ section on time and time-keeping as part ofC >   the next edition of the FAQ, in honor of the flood of questions B >   arising from the recent US Federal legislative DST activities. >  > ( >  ---------------------------- #include' <rtfaq.h> ----------------------------- 4 >     For additional, please see the OpenVMS FAQ --  www.hp.com/go/openvms/faq , >  --------------------------- pure personal# opinion --------------------------- G >         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com  >    ------------------------------    Date: 31 Aug 2005 13:46:32 -0500/ From: brooks@cuebid.zko.dec.nospam (Rob Brooks) ; Subject: Re: OT: .com spending no longer and porting issues - Message-ID: <tGNcy$fmKQKB@cuebid.zko.dec.com>   ( John Reagan <john.reagan@hp.com> writes: > DeanW wrote: > G >> I would expect OS code or apps that get into the guts of hardware to H >> be more somewhat more involved... Note the VMS engineers didn't "just >> recompile", either.  F > I'd guess that 80-90% of VMS "just recompiled".  We had a couple of  > interesting cases.  H > I know of many, many pieces that simply recompiled and have yet to be  > touched by human hands.   L To amplify what John said -- several relatively large and twisted componentsD needed absolutely no touching whatsoever (multipath, for example) or minimal touching.   K The Shadowing driver is a good example of the later.  Written in Macro-32,  N it really does some awfully scary things, but thanks to John's work on IMACRO,J we just needed to touch up the handful of places that had made assumptions about the calling standard.    --    M Rob Brooks    VMS Engineering -- I/O Exec Group     brooks!cuebid.zko.dec.com    ------------------------------  % Date: Wed, 31 Aug 2005 14:18:27 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> ; Subject: Re: OT: .com spending no longer and porting issues + Message-ID: <4315F467.48D1CB1@teksavvy.com>    Rob Brooks wrote: N > To amplify what John said -- several relatively large and twisted componentsF > needed absolutely no touching whatsoever (multipath, for example) or > minimal touching.   H However, you have established testing procedures which you use regularly whenever updates are made.  D For a commercial customer point of view, they don't spend their timeE updating sofwtare and testing them. So for them, even if it is just a H recompile without code changes, it still requires testing and qualifying
 the software.    ------------------------------   Date: 31 Aug 2005 19:57:25 GMT( From: bill@cs.uofs.edu (Bill Gunshannon); Subject: Re: OT: .com spending no longer and porting issues + Message-ID: <3nmgd5F29artU1@individual.net>   3 In article <LcmRe.11472$Gq7.5510@news.cpqcorp.net>, ) 	John Reagan <john.reagan@hp.com> writes:  > DeanW wrote: >  >>  G >> I would expect OS code or apps that get into the guts of hardware to H >> be more somewhat more involved... Note the VMS engineers didn't "just >> recompile", either. > F > I'd guess that 80-90% of VMS "just recompiled".  We had a couple of  > interesting cases. > H > One, we didn't think we had bugcheck ready to take a crash dump.  The G > code recompiled cleanly, but we thought we would have to make *some*  I > changes.  However, somebody just hit ;C in XDELTA instead of ;P and lo  F > and behold, we had a dump file.  I think we did have to go back and H > extend the crash dumping to dump out unwind descriptors and the like, = > but we were shocked how well it worked without touching it.  > J > The code for SMP was much the same way.  We forgot to comment it out in H > the build and guess what?  It worked untouched (again, I think we did M > enhance the code later for better handling of some IA64-specific features).  > H > I know of many, many pieces that simply recompiled and have yet to be  > touched by human hands.  >  >   G We've been over this before.  For many businesses "just a recompile" is F not acceptable before code gets put into production.  It's nice to sayK in a press blurb, but once again it misses reality by a pretty wide margin.   E Although I have been unable up to this point to track down any of the G mythical thousands of DOD VMS Systems (at least among my DA peers) they & would be a prime example of the above.   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------  % Date: Wed, 31 Aug 2005 17:00:42 -0700 , From: Ken Fairfield <my.full.name@intel.com>+ Subject: Re: Pass arguments to a .com file. + Message-ID: <df5gba$ple$1@news01.intel.com>    Z wrote: > Ken Fairfield wrote: > ; >>     3) As a matter of style, I prefer to fully spell out D >> DCL verb names, e.g., DIRECTORY, inside command procedures.  ThisD >> serves the dual function of clearly documenting the procedure andB >> of avoiding possible conflicts, for the most part, with symbols' >> that may be defined in your process.  >  > J > Why not just SET SYMBOL/SCOPE to address that? Isn't that what it's for?  F      Absolutely. :-)  I use it if I need to be particularly defensive,) e.g., for the "unwashed masses".  :-) :-)   I > Users I run across seem to _love_ $DEL*ETE :== ... in their LOGIN.COMs  @ > and using the full spelling of DELETE doesn't circumvent that.  D      Right.  $ DEL*ETE == "DELETE/CONFIRM" is especially frustratingC when you're wanting to do a DELETE/ENTRY or DELETE/SYMBOL, etc. ;-p D One work around, besides using SET SYMBOL/SCOPE, is simply to defineF a local symbol, $ DELETE = "DELETE", to override the (possible) global symbol.   C      In any case, I was just trying to provide an untested, off the C cuff starter for the OP.  Having been recently burned by an obscure > procedure called during system startup that included the line,   $ set au/al/en=(b:a,logf:di)  D (no, I didn't write that), and spending an inordinate amount of timeC trying to find why my security audit settings were getting reset on 4 reboot, I thought I'd share a personal style hint...   	-Ken  --  6 I don't speak for Intel, Intel doesn't speak for me...  
 Ken Fairfield ! D1C Automation VMS System Support " who:   kenneth dot h dot fairfield where: intel dot com   ------------------------------  % Date: Wed, 31 Aug 2005 22:28:09 +0300 0 From: "MUSTAFA ATAKAN" <matakan@inteltek.com.tr>, Subject: RE: Possible reasons of a file lossL Message-ID: <F014DACB8BE63442993543B780A2F0180163BEA3@asteriks.inteltek.ist>  D I looked at it and there is no syslost directory on the device...=20   -----Original Message-----< From: Bob Kaplow [mailto:kaplow_r@encompasserve.org.mars]=20( Sent: Wednesday, August 31, 2005 8:26 PM To: Info-VAX@Mvb.Saic.Com , Subject: Re: Possible reasons of a file loss  
 In articleA <F014DACB8BE63442993543B780A2F0180163BE28@asteriks.inteltek.ist>, 2 "MUSTAFA ATAKAN" <matakan@inteltek.com.tr> writes:F > Is there any possible reason of a file loss in OpenVMS. This loss=20H > event occured two times, both on Monday's but the dates are different.  E > So I have checked the Sunday evening backup, the file was there.=20 G > However, while i have checked the Monday evening backup, the file was  absent. J > Between these two backup times, the only commands that are run on the=20I > directory of these files are "analyze /disk /repair" and "purge". Is=20 G > it possible that one of these commands may be responsible of the file  loss?   B Have you looked in the [SYSLOST] directory on the device to see if- analyze found the lost file and put it there?    --=20 D   Bob Kaplow   NAR # 18L   >>> To reply, there's no internet on Mars
 (yet)! <<<E Kaplow Klips & Baffle:	http://nira-rocketry.org/Document/MayJun00.pdf =     www.encompasserve.org/~kaplow_r/    www.nira-rocketry.org  www.nar.org   ;         We must have faith in our democratic system and our 
 Constitution, G         and in our ability to protect at the same time both the freedom  and )         the security of all Americans.=20    ------------------------------  % Date: Wed, 31 Aug 2005 21:17:08 +0200 & From: "H Vlems" <nospam@what.ever.com>, Subject: Re: Possible reasons of a file loss< Message-ID: <c2f08$431602cd$513b9a2c$22029@news.versatel.nl>  = "MUSTAFA ATAKAN" <matakan@inteltek.com.tr> schreef in bericht F news:F014DACB8BE63442993543B780A2F0180163BE28@asteriks.inteltek.ist... Hi,   G Is there any possible reason of a file loss in OpenVMS. This loss event E occured two times, both on Monday's but the dates are different. So I D have checked the Sunday evening backup, the file was there. However,D while i have checked the Monday evening backup, the file was absent.E Between these two backup times, the only commands that are run on the G directory of these files are "analyze /disk /repair" and "purge". Is it H possible that one of these commands may be responsible of the file loss?  H Or is there any way to trace the user's activity history to find out any$ deletion commands that they entered?   My OpenVMS version: 7.3-2 8 Disks are in HSG80 Compaq storage and in SAN environment Fibre SCSI patch is v0300    Best Regards...   F ----------------------------------------------------------------------  E I'm pretty sure the file is not "lost". The most likely event is that  ANALYZE/DISK found somethingL wrong with the file and put it in DJA2:[SYSLOST]. (subsitute your own device
 for DJA2:)H Purge may have something to to with it as well. Imagine a strategy where disk storage is kept to a K minimum, so /KEEP=1 is used. The owner of the file does not realize that at  a particular moment and I decides to remove the latest version, assuming that previous versions are  still there; which they aren't owing to the PURGE/KEEP=1.J Just guessing, but I've been using VMS since 1983 and have seen many files "lost" by users.F Never, let me repeat that, never, because VMS made the error, somehow, someway, after a lot of I discussions it appeared that the user might have done something wrong ;-)    Hans   ------------------------------  + Date: Wed, 31 Aug 2005 20:56:01 +0000 (UTC) . From: klewis@LUMINA.MITRE.ORG (Keith A. Lewis), Subject: Re: Possible reasons of a file loss. Message-ID: <df55h1$rrb$1@newslocal.mitre.org>   "MUSTAFA ATAKAN" <matakan@inteltek.com.tr> writes in article <F014DACB8BE63442993543B780A2F0180163BE28@asteriks.inteltek.ist> dated Wed, 31 Aug 2005 18:28:32 +0300:H >Is there any possible reason of a file loss in OpenVMS. This loss eventF >occured two times, both on Monday's but the dates are different. So IE >have checked the Sunday evening backup, the file was there. However, E >while i have checked the Monday evening backup, the file was absent. F >Between these two backup times, the only commands that are run on theH >directory of these files are "analyze /disk /repair" and "purge". Is itI >possible that one of these commands may be responsible of the file loss?  > I >Or is there any way to trace the user's activity history to find out any % >deletion commands that they entered?  >  >My OpenVMS version: 7.3-29 >Disks are in HSG80 Compaq storage and in SAN environment  >Fibre SCSI patch is v0300  & The commands are something like these.  * to enable audit logging of file deletions:- $ SET AUDIT/CLASS=FILE/AUDIT=(SUCCESS:DELETE)   3 then when you want to look at what has been logged:  $ SET AUDIT/SERVER=FLUSHC $ ANAL/AUDIT /FULL /OUT=TEMP.TMP SYS$MANAGER:SECURITY.AUDIT$JOURNAL    and if the file gets too big:  $ SET AUDIT/SERVER=NEW* $ PURGE SYS$MANAGER:SECURITY.AUDIT$JOURNAL  0 --Keith Lewis              klewis {at} mitre.org> The above may not (yet) represent the opinions of my employer.   ------------------------------  % Date: Wed, 31 Aug 2005 19:54:21 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>, Subject: Re: Possible reasons of a file loss+ Message-ID: <4316513D.9180CF16@comcast.net>    MUSTAFA ATAKAN wrote:  >  > Hi,  > I > Is there any possible reason of a file loss in OpenVMS. This loss event G > occured two times, both on Monday's but the dates are different. So I F > have checked the Sunday evening backup, the file was there. However,F > while i have checked the Monday evening backup, the file was absent.G > Between these two backup times, the only commands that are run on the I > directory of these files are "analyze /disk /repair" and "purge". Is it J > possible that one of these commands may be responsible of the file loss? > J > Or is there any way to trace the user's activity history to find out any& > deletion commands that they entered? >  > My OpenVMS version: 7.3-2 : > Disks are in HSG80 Compaq storage and in SAN environment > Fibre SCSI patch is v0300   : Some respondents seemed to be confused by the term "lost".  G In VMS-land, a "lost" file is one with no directory entry. While that's H very common during normal operations (such as files spooled for printingG to spooled devices), it is generally considered a corruption otherwise.   D Obviously, you mean to say that a file appears to have been deleted.  E Try Keith Lewis's advise about setting audits and alarms that you can  report on later.  > It would be interesting to read how you determined that a fileC disappeared since you mention the use of PURGE (it's job: to delete  "surplus" file versions).    --   David J Dachtera dba DJE Systems  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/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Wed, 31 Aug 2005 19:49:34 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net> Subject: Re: problem with SLS + Message-ID: <4316501E.2EE6E5AE@comcast.net>   
 NOSPAM wrote:  >  > Hello,D >        i am working on a VMS/AXP 7.3-1, which has a TL820 jukebox.9 > The installation of SLS 2.9C fails when running the IVP  >  (storage add operation) : > E > "SLS does not respond!  Run SLS$SYSTEM:SLS$SLSMGR.EXE on the server F > and make sure all nodes can access the server database.  Run the IVPA > again. If there still is a problem, restart SLS and run the IVP 	 > again."  > 9 > while this happen, this error is logged on the console:  > 8 > %%%%%%%%%%%  OPCOM  31-AUG-2005 16:02:52.82  %%%%%%%%%! > Message from user SLS on AXPGS3 D > SLS$TAPMGRRQ, -SYSTEM-F-REJECT, connect to network object rejected > 9 > After sys$startup:sls$startup  the processes are there:  > C > 22400137 SLS$TAPMGRDB  LEF   10   198   0 00:00:00.17  1064    97 C > 22400138 SLS$TAPMGRRQ  LEF    9   928   0 00:00:00.51  1620   297 C > 2240013A SLS$OPCOM     HIB    6    45   0 00:00:00.03   186   207  > * > the network object SLS$DBX appear be ok. > G > Running SLS$SYSTEM:SLS$SLSMGR.EXE I added AXPGS3 as authorized client F > but, the option 2 (Volume Pool Auth.) didn't work, it hangs while on0 > the console a new OPCOM like before cames out. >  > Any help about it?  7 Try a newer SLS. The current version is V2.9H, I think.    --   David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------    Date: 31 Aug 2005 13:36:21 -0700! From: kenneth.randell@verizon.net H Subject: Setting 'default' DECWindows display with multi-heads on rx26xxC Message-ID: <1125520581.958930.101000@o13g2000cwo.googlegroups.com>   F I have an rx2620 with the management card that has the Radeon 7000 VGA0 built in, as well as a Radeon 7500 card as well.  D When VMS boots, GHA0 is the Radeon 7500 and GHB0 is the Radeon 7000.1 The CDE login screen appears as expected on GHA0.   D For bizarre configuration reasons, I would like the 'default' (i.e.,E where the CDE login appears) to be on the GHB0 (7000) device, not the F GHA0 device.  I have renamed the DECW$PRIVATE_SERVER_SETUP.TEMPLATE toD *.COM.  I have also set DECW$SERVER_SCREENS to "GHB0,GHA0", but this+ option looks like it only affects XINERAMA.   < Is there a supported way of doing this, or am I out of luck?   Thanks.    ------------------------------  # Date: Wed, 31 Aug 2005 22:17:22 GMT * From: "FredK" <fred.nospam@nospam.dec.com>L Subject: Re: Setting 'default' DECWindows display with multi-heads on rx26xx3 Message-ID: <S%pRe.11514$qv7.5251@news.cpqcorp.net>    $ DECW$DEVICE == "GHB0, GHA0"     . <kenneth.randell@verizon.net> wrote in message= news:1125520581.958930.101000@o13g2000cwo.googlegroups.com... H > I have an rx2620 with the management card that has the Radeon 7000 VGA2 > built in, as well as a Radeon 7500 card as well. > F > When VMS boots, GHA0 is the Radeon 7500 and GHB0 is the Radeon 7000.3 > The CDE login screen appears as expected on GHA0.  > F > For bizarre configuration reasons, I would like the 'default' (i.e.,G > where the CDE login appears) to be on the GHB0 (7000) device, not the H > GHA0 device.  I have renamed the DECW$PRIVATE_SERVER_SETUP.TEMPLATE toF > *.COM.  I have also set DECW$SERVER_SCREENS to "GHB0,GHA0", but this- > option looks like it only affects XINERAMA.  > > > Is there a supported way of doing this, or am I out of luck? > 	 > Thanks.  >    ------------------------------  % Date: Wed, 31 Aug 2005 16:07:29 -0400 ' From: Dave Froble <davef@tsoft-inc.com> % Subject: Re: SFTP in batch- password? 0 Message-ID: <11hc36pj7m46pd2@corp.supernews.com>   DeanW wrote:: > On 31 Aug 2005 08:49:15 +0100, Peter 'EPLAN' LANGSTOEGER > <peter@langstoeger.at> wrote:  > a >>In article <3f119ada050830115858b2fa6f@mail.gmail.com>, DeanW <dean.woodward@gmail.com> writes:  >>L >>>We're trying to use SFTP to communicate to another VMS box, doing routineM >>>file transfers in batch mode, but can't figure out how to get it to take a H >>>password except interactively. Is this possible, or do we need to try >>>something else? >>9 >>The obvious suggestion is COPY over DECnet (over TCPIP) B >>and with Proxies you don't even need a password in your scripts. >  > B > Not all nodes run DECnet, there's a firewall in the middle, etc. > ! > SFTP or sneakernet, I'm afraid.   D Run a 'tunnel' between the VMS systems.  DECnet, TCP/IP, most every @ protocol should work.  In theory.  In practice may be different.   --  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: 31 Aug 2005 16:51:57 -0700" From: "Jose Baars" <peut@peut.org>% Subject: Re: SFTP in batch- password? B Message-ID: <1125532317.094690.80370@g49g2000cwa.googlegroups.com>  9 SFTP is quite horrible in general, and the TCPIP Services 2 SSH implementation on VMS is a bit, well... young.  4 Using SFTP in batch is a bit of a mess, as you might have discovered:1 password authentication is out of the window, you 1 must use key authentication, and then you have to 4 explain to the guys at the other end how they should1 convert it to OpenSSH format ( which AFAIK is not * supported by the TCPIP Services ssh_keygen/ tool, as opposed to most Unix implementations), 9 and how to set up their authorization file, and the rest.   4 Then writing the DCL is not straightforward either :/ you need all kinds of messy intermediate files, = probably you might want to recreate the ssh2 directory, etc., 2 sftp creates directories at the weirdest places to= put the random file in ( where is rn: on VMS, but what's more : important : why do most ported utilities when it is aroundG like SSH use it commonly as a reversed null device in stead of stopping   F to read after a decent amount of bytes like OpenSSL does (mostly :-( )  2 So through a application developer doing something. wonderful on Windows, I was pointed at Jsch atA http://www.jcraft.com/jsch, which provides a Java SFTP 'library'.   / After having learned enough Java to get around, 8 it was quite doable to make a few utilities that fulfill$ the most obvious SFTP tasks at hand:@ getting and putting a file using either a password or a keyfile.  6 Something like sftpchannel.put(localfile, remotefile); is no rocket science is it?  I'm quite happy with it.   ------------------------------  % Date: Wed, 31 Aug 2005 16:11:38 -0400 ' From: Dave Froble <davef@tsoft-inc.com> ) Subject: Re: simh with logical networking 0 Message-ID: <11hc3ejelllf370@corp.supernews.com>   Alan Greig wrote:  >  >  > Bob Koehler wrote: > G >> In article <1125484348.673445.211120@g43g2000cwa.googlegroups.com>,  ( >> "Galen" <gltackett@gmail.com> writes: >>= >>> I'll have to try the equivalent on my OS X 10.3.9 system.  >>> ! >>> Has anyone tried to use TUN?   >> >> >> >>    I'll bite.  What is TUN? >  > B > As far as I understand it TUN is just an IP TUNnel only logical I > interface. You are definitely not going to be able to cluster over it.  K > TAP carries all traffic across the logical LAN including SCS and you can  J > bring up a cluster over it if you have bandwidth and low enough latency. >   B  From one of Keith's posts on your topic of clustering using simh:  H In my group's labs (Multivendor Systems Engineering), we have a Layer 2 B Tunneling Protocol (L2TP) tunnel set up between a couple of Cisco : routers so we can send SCS traffic over IP, and we have a H 100-millisecond delay introduced on that IP link using a Shunra network H emulator box. We have a 7-node cluster running, split across this link. I This equipment was being staged for the hands-on workshop "Long-distance  H HP OpenVMS Clusters" slated for the (now-postponed) HP Technology Forum C in New Orleans (our thoughts and prayers are with the folks there).    Possibly a tunnel would work?    --  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: Wed, 31 Aug 2005 19:52:01 -0400 ) From: "Neil Rieck" <n.rieck@sympatico.ca>  Subject: StorageWorks is alive! 9 Message-ID: <NnrRe.4399$884.630216@news20.bellglobal.com>   4 Some of you may have already seen this but I hadn't.  G I was under the impression that Compaq had killed off, or spun off, the E StorageWorks product but it looks like HP has reversed that thinking.   " www.hp.com/go/storageworks/OpenVMS  www.hp.com/go/storageworks/Tru64" www.hp.com/go/storageworks/NonStop  www.hp.com/go/storageworks/hp-ux  www.hp.com/go/storageworks/linux" www.hp.com/go/storageworks/windows  
 Neil Rieck Kitchener/Waterloo/Cambridge,  Ontario, Canada.8 http://www3.sympatico.ca/n.rieck/links/cool_openvms.html   ------------------------------  # Date: Wed, 31 Aug 2005 22:16:16 GMT * From: "FredK" <fred.nospam@nospam.dec.com>3 Subject: Re: traditional DECwindows: save workspace 3 Message-ID: <Q_pRe.11513$7B7.9858@news.cpqcorp.net>   L "Phillip Helbig---remove CLOTHES to reply" <helbig@astro.multiCLOTHESvax.de>/ wrote in message news:df517t$c3v$1@online.de... D > With CDE, I can save the workspace, so that all DECwindows objectsH > appear at the same place the next time I log in.  With the traditionalG > interface, I can use the "automatic startup" to tell it WHAT to start H > automatically, but the placement is not what I want.  Is there any way  > to get the CDE behaviour here? >   L Not that I know of offhand.  The CDE functionality is actually a function of# a window manager "Save Me" message.   D > Something else:  I've seen it done, so I know it works, but I haveG > forgotten how.  How can I direct the entire CDE interface to a remote J > display (I've seen it being redirected to a VXT across the internet, butJ > now I want to redirect it to a VAX in the same cluster).  At the moment,I > I have a nice new ALPHAserver 1200 (actually a Digital Server 5305 with C > modified firmware---I don't care about support since it is just a G > hobbyist system), but unfortunately no SCSI controller, so except for H > the system disk and a CD-ROM drive, I'm using MSCP-served disks from aJ > VAX in the cluster.  The VAX has a nicer and bigger monitor than the oneJ > I could connect to the ALPHA, so I want to redirect the CDE-display fromI > the ALPHA to the VAX (at least as long as I have to keep the VAX in the F > cluster to serve the disks, I see no reason to use a smaller monitor# > directly connected to the ALPHA).  >   6 Offhand (just trying to fire up some dead braincells):  + SET DISPLAY/CRE/PERM/NODE=FOO/TRANSP=DECNET  MC DECW$STARTLOGIN   ------------------------------  % Date: Wed, 31 Aug 2005 19:23:36 -0400 . From: JF Mezei <jfmezei.spamnot@vaxination.ca>3 Subject: Re: traditional DECwindows: save workspace - Message-ID: <43163BD9.DA5300C4@vaxination.ca>   / Phillip Helbig---remove CLOTHES to reply wrote:  > D > With CDE, I can save the workspace, so that all DECwindows objectsH > appear at the same place the next time I log in.  With the traditionalG > interface, I can use the "automatic startup" to tell it WHAT to start H > automatically, but the placement is not what I want.  Is there any way  > to get the CDE behaviour here?  G DECW$CLOCK handles the placement. If you look at its .dat file, you see J a clock.geometry setting in it that records placement of window on screen.     You could add  XmNxE        Specifies the x-coordinate of the upper left outside corner of C the widget's window. The value is relative to the upper left inside  corner of the parent        window.   XmNyE        Specifies the y-coordinate of the upper left outside corner of C the widget's window. The value is relative to the upper left inside  corner of the parent        window.    H resources in your .dat files for the various applications to force their( window to appear in a specific location.   ------------------------------  % Date: Wed, 31 Aug 2005 21:08:28 +0200 & From: "H Vlems" <nospam@what.ever.com>% Subject: Re: VAXen as a plural of VAX < Message-ID: <c725d$431600c5$513b9a2c$21263@news.versatel.nl>  K "Tim ffrench-Lynch" <tim-DOT-ffrench-HYPHEN-lynch@selex-sas.com> schreef in / bericht news:4315BA48.5F9AA97A@selex-sas.com... H > Is VAXen a well enough established plural of VAX to use in a technical > document?  >  > Tim   D If VAX is considered to be a noun then in dutch you'd be correct. InK english, a foreign language for me, I'd hesitate. IMHO the plural of VAX is  VAX systems.  I That said, modern use of the english language is confusting to the extent # that VAXen won't even register ....    Hans   ------------------------------    Date: 31 Aug 2005 13:00:46 -0700# From: "WhoDat?" <whohe@whoever.com> % Subject: Re: VAXen as a plural of VAX B Message-ID: <1125518446.009281.46310@g14g2000cwa.googlegroups.com>   Steve Lionel wrote: 7 > On Wed, 31 Aug 2005 15:10:16 +0100, Tim ffrench-Lynch 5 > <tim-DOT-ffrench-HYPHEN-lynch@selex-sas.com> wrote:  > I > >Is VAXen a well enough established plural of VAX to use in a technical  > >document? > P > Back around 1979, there was a (humorous) attempt to answer this question, withK > a number of alternatives proposed.  My favorite at the time was "VAXoth", ! > which was alleged to be Hebrew.  > H > Of course, the lawyers would want you to say "VAX systems", (with (TM)K > acknowledgement), and if I were writing a technical paper, I would choose P > something like that rather than trying to be clever and pluralize a trademark. > Steve   F VAX is trademarked (in the US), but not by Digital, Compaq or HP. See:  > < http://tess2.uspto.gov/bin/gate.exe?f=doc&state=p6nnrr.3.7 >   ## VAX A IC 025. US 022 039. G & S: WEARING APPAREL, NAMELY, SHIRTS, HATS, G SHORTS, JACKETS, AND PANTS. FIRST USE: 20010323. FIRST USE IN COMMERCE:  20010323 ## <:^)   ------------------------------    Date: 31 Aug 2005 13:13:12 -0700# From: "WhoDat?" <whohe@whoever.com> % Subject: Re: VAXen as a plural of VAX C Message-ID: <1125519192.302253.112320@g14g2000cwa.googlegroups.com>    WhoDat? wrote: > Steve Lionel wrote: 9 > > On Wed, 31 Aug 2005 15:10:16 +0100, Tim ffrench-Lynch 7 > > <tim-DOT-ffrench-HYPHEN-lynch@selex-sas.com> wrote:  > > K > > >Is VAXen a well enough established plural of VAX to use in a technical  > > >document? > > R > > Back around 1979, there was a (humorous) attempt to answer this question, withM > > a number of alternatives proposed.  My favorite at the time was "VAXoth", # > > which was alleged to be Hebrew.  > > J > > Of course, the lawyers would want you to say "VAX systems", (with (TM)M > > acknowledgement), and if I were writing a technical paper, I would choose R > > something like that rather than trying to be clever and pluralize a trademark.	 > > Steve  > H > VAX is trademarked (in the US), but not by Digital, Compaq or HP. See: > @ > < http://tess2.uspto.gov/bin/gate.exe?f=doc&state=p6nnrr.3.7 > >  > ## > VAX C > IC 025. US 022 039. G & S: WEARING APPAREL, NAMELY, SHIRTS, HATS, I > SHORTS, JACKETS, AND PANTS. FIRST USE: 20010323. FIRST USE IN COMMERCE: 
 > 20010323 > ## > <:^)  F Oops. Sorry, our type of VAX (tm) is still LIVE, renewed by Compaq forG 6 years on 2002/03/07 -- on page two along with the VAX vacuum cleaner.    ------------------------------  % Date: Wed, 31 Aug 2005 14:35:03 -0600 - From: "Jim MacKenzie" <jim@dusykbarlow.sk.ca> % Subject: Re: VAXen as a plural of VAX + Message-ID: <4316146e$1@news.accesscomm.ca>   2 "H Vlems" <nospam@what.ever.com> wrote in message 6 news:c725d$431600c5$513b9a2c$21263@news.versatel.nl... > F > If VAX is considered to be a noun then in dutch you'd be correct. InK > english, a foreign language for me, I'd hesitate. IMHO the plural of VAX   > is > VAX systems.  B If "VAX systems" is the plural, then "VAX system" is the singular.  J If we can use "VAX" as a singular, then it must have a plural form.  "ax" F (sometimes spelled "axe") is "axes" in the plural, but "ox" is "oxen".  D If it were my paper, I'd be inclined to use "VAXen" - it seems well L accepted - but I'd insert a footnote at the first iteration, explaining why  I chose it.    Jim    ------------------------------  % Date: Wed, 31 Aug 2005 16:58:52 -0400 3 From: Steve Lionel <Steve.Lionel@REMOVEintelME.com> % Subject: Re: VAXen as a plural of VAX 8 Message-ID: <b56ch1d9eq3k26ac9tif1pasa5lecc7g7v@4ax.com>  C On 31 Aug 2005 13:00:46 -0700, "WhoDat?" <whohe@whoever.com> wrote:   G >VAX is trademarked (in the US), but not by Digital, Compaq or HP. See:  > ? >< http://tess2.uspto.gov/bin/gate.exe?f=doc&state=p6nnrr.3.7 >  >  >##  >VAXB >IC 025. US 022 039. G & S: WEARING APPAREL, NAMELY, SHIRTS, HATS,H >SHORTS, JACKETS, AND PANTS. FIRST USE: 20010323. FIRST USE IN COMMERCE:	 >20010323  >##  ><:^)   L DEC never registered VAX as a trademark - its philosophy was not to registerM trademarks in general.  Also, trademarks are product-line specific.  When the E DEC VAX was introduced, there was a VAX vacuum cleaner line in the UK L ("Nothing sucks like a VAX!")  For a time in the 1980s, Sears imported theseM into the US, and I see now that Circuit City (I think) is selling VAX-branded  vacuums.  K There is no conflict here, though DEC and the VAX vacuum people did discuss K the matter and come to some sort of agreement.  There is Cadillac dog food, G for example.  As long as the product lines are distinct enough to avoid I confusion, different companies can share a trademark.  (Consider, though, L Apple Music's lawsuit against Apple Computer for violating its agreement not to enter the music business...)   J Nevertheless, you should acknowledge trademarks claimed by others when youL know of them, at least when writing formally.  A common way of doing this isJ to use an asterisk and then add a footnote that trademarks are property of& their respective owners, or some such.   Steve Lionel Software Products Division Intel Corporation 
 Nashua, NH  8 User communities for Intel Software Development Products"   http://softwareforums.intel.com/ Intel Fortran Support 7   http://developer.intel.com/software/products/support/    ------------------------------    Date: 31 Aug 2005 15:29:06 -0700$ From: "AEF" <spamsink2001@yahoo.com>% Subject: Re: VAXen as a plural of VAX B Message-ID: <1125527346.583792.77300@o13g2000cwo.googlegroups.com>   Jim MacKenzie wrote:3 > "H Vlems" <nospam@what.ever.com> wrote in message 8 > news:c725d$431600c5$513b9a2c$21263@news.versatel.nl... > > H > > If VAX is considered to be a noun then in dutch you'd be correct. InL > > english, a foreign language for me, I'd hesitate. IMHO the plural of VAX > > is > > VAX systems. > D > If "VAX systems" is the plural, then "VAX system" is the singular. > K > If we can use "VAX" as a singular, then it must have a plural form.  "ax" H > (sometimes spelled "axe") is "axes" in the plural, but "ox" is "oxen". > E > If it were my paper, I'd be inclined to use "VAXen" - it seems well M > accepted - but I'd insert a footnote at the first iteration, explaining why 
 > I chose it.  >  > Jim     F I don't see why people tend to like the term "VAXen". It reminds me ofF fauna, centaurs, oxen, and fairy-tale space -- not exactly what I like  to think of when I think of VAX.  F (What a cool name: VAX. Then there's the ever popular VAX/VMS. And theF older VAX-11. Hard to beat! How can you beat -- VAX -- or -- VAX/VMS?)  B Do you pay taxen and send faxen to order VAXen which come in boxen5 dragged by oxen? Do you cut open the boxen with axen?    I prefer VAXes or VAX systems.     Hey, different strokes...    JMHO   ------------------------------   End of INFO-VAX 2005.487 ************************