1 INFO-VAX	Tue, 30 Dec 2003	Volume 2003 : Issue 723       Contents:N Alphaserver 2100/VMS 7.1-1H2 - Move IP from FDDI Card to Ethernet Card (DE101) CTPOP and MVI instructions ? Re: DISK DRIVE PERFORMANCE Re: DISK DRIVE PERFORMANCE Re: DISK DRIVE PERFORMANCE Re: DISK DRIVE PERFORMANCE Re: DISK DRIVE PERFORMANCE, Re: Floating point arithmetic support in DCL, Re: Floating point arithmetic support in DCL, Re: Floating point arithmetic support in DCL, Re: Floating point arithmetic support in DCL, Re: Floating point arithmetic support in DCL8 Re: Linux kernel security bug ... VMS kernel rock solid!8 Re: Linux kernel security bug ... VMS kernel rock solid!8 RE: Linux kernel security bug ... VMS kernel rock solid! Re: My Holiday Wishes  Re: recommendation for NPAGEVIR ( Re: Singapore Exchange to run on OpenVMS( Re: Singapore Exchange to run on OpenVMS VMS 5.5-2H4 and TZ87? ! Re: VMS runs well on HP Superdome ! Re: VMS runs well on HP Superdome ! RE: VMS runs well on HP Superdome ! Re: VMS runs well on HP Superdome ! Re: VMS runs well on HP Superdome / Re: What Andrew and sun can't stand is that ...   F ----------------------------------------------------------------------    Date: 30 Dec 2003 09:21:34 -0800 From: barry_mc@usa.net (Baz)W Subject: Alphaserver 2100/VMS 7.1-1H2 - Move IP from FDDI Card to Ethernet Card (DE101) = Message-ID: <5dd78dcd.0312300921.2222c461@posting.google.com>   E We have a legacy operational Digital Alphaserver 2100 (5/300) running B OpenVMS 7.1-1H2.  It is used sporadically for accessing historicalE data and will be made redundant mid/late-2004  consequently there is & no budget for investment in this area!  E This is on it's own FDDI ring, the only remaining device on this ring F which I want to decommission.  The server has also a separate ethernetB card which I want to migrate to but I don't know how to, hence theF query.  I come from a Cisco background and am the 'network guy' here -& consequently I know zip about VAX/VMS.  F The box is running IP only and is on the same subnet as other EthernetE devices.  Consequently, I believe that I only want to reconfigure the > TCPIP stack to use the ethernet interface in place of the FDDIB interface, moving the IP address, subnet mask and default gateway.  F We have a 'VAX/VMS guy' here who drives, waters, feeds etc the machineB from an apps point of view, but is unsure what to do network-wise.  & I've looked at the cards on the back:-  . FDDI Card - operational with solid green lightD Ethernet Card - 1 RJ45 and 1 AUI - both empty as expected.  RJ45 hasD solid green light beside it and AUI has light which flashes every 12E seconds. 'DE101' is stamped on the card which I assume is the type of 
 network card.   D I would appreciate any assistance available  again please assume no= Alpha/VMS experience.  These are the network devices listed:-   E Device EWA0:, device type TULIP, is online, network device, device is  a template only.; Device EWA2:, device type TULIP, is online, network device. ; Device EWA3:, device type TULIP, is online, network device. < Device EWA10:, device type TULIP, is online, network device.  " Thanks in advance for any replies!   ------------------------------  # Date: Tue, 30 Dec 2003 12:50:32 GMT 8 From: Peter Ljungberg <SYSTEMNOSPAM@pcmk2.data.telia.se>% Subject: CTPOP and MVI instructions ? 4 Message-ID: <sweIb.44204$dP1.175916@newsc.telia.net>   Hi all,   D I'm experimenting with  performance related issues and wondering if 
 anyone has an = example using CTPOP (count population)  and/or the MVI Alpha   instructions, for VMS.   Happy new VMS-year! L For those of us who still have the luck to use and work with a very good OS.  
  >>> ^P.Lj   ------------------------------  % Date: Tue, 30 Dec 2003 07:45:24 +0100  From: Dirk Munk <munk@home.nl># Subject: Re: DISK DRIVE PERFORMANCE 2 Message-ID: <bsr7gb$k0g$1@news2.tilbu1.nb.home.nl>  D Now if you could just tell us where you got this nice IOT tool......   Regards, Dirk   Rob Young wrote:^ > In article <03122915165251@dscis6-0.dalsemi.com>, brandon@dalsemi.com (John Brandon) writes: > : >>DS-RZ1DB-VW   	 9.1-GB  Ultra SCSI Drive, 7.2k RPM, SBB : >>DS-RZ1ED-VW   	 18.2-GB Ultra SCSI Drive, 10k  RPM, SBB : >>DS-RZ1FC-VW   	 36-GB   Ultra SCSI Drive, 10k  RPM, SBB  >>I >>Does anyone have the performance information for the above disk drives?  >>- >>Looking for transfer rate and I/O capacity.  >>& >>Or where can I find the information? >> >  > ) > 	One example .. because it takes time.   >  > 	Google:  DS-RZ1ED-VW ? > 	finds compaq part number.  I found this on the second google A > 	page.  I know I'm looking for a numeric and usually a -b21 or   > 	-bnn at the end:  > 1 > http://www.ebiztechcorp.com/specials/deals.html  >  > DS-RZ1ED-VW (aka 380589-B21  > 8 > 	Searching 380589-b21 and looking at a number of hits: > 6 > http://www.computeranimal.com/harddrives/475984.html > G > 	40 MByte/sec external transfer rate.  IO / sec (IOPS)?  Good luck.   @ > 	But here is a link, it has been debated in comp.arch.storage. > g > http://groups.google.com/groups?selm=KaEnRv%2BJQp14%40eisner.encompasserve.org&oe=UTF-8&output=gplain  > C > 	10K drive , lightly loaded (queue depth 1 or so) does about 110  C > 	random IO /sec.  But as you can see from that link above you can B > 	drive a 15K drive with multiple queues to get 200 IO/sec out of5 > 	it.  But as Bill points out, your turnaround drops  > 	off, as this graph shows: > 7 > 	Looking at a bunch of different stream combinations:  > ; > IO Strm	IO Rte	Elpsed Time	IO Duration	IO Spread	Worst IO - > 1	143/s	27.82 sec	6.9 ms		2.0 ms		29.2   ms . > 2	147/s	27.14 sec	13.5 ms		2.9 ms		44.9   ms. > 3	175/s	22.82 sec	16.9 ms		9.3 ms		110.3  ms/ > 4	190/s	21.04 sec	20.7 ms		14.4 ms		129.8  ms / > 5	201/s	19.81 sec	24.3 ms		17.8 ms		195.3  ms 5 > 8       230/s	17.37 sec	33.8 ms		28.2 ms		259.7  ms 5 > 16      238/s	16.79 sec	65.9 ms		32.2 ms		368.1  ms  > > > 	2 IO streams, can only service one at a time so must be at D > 	least 1 pending IO, etc.  But as you can see, even though you canF > 	get 230 IOPS with 8 IO streams, on average it takes 33.8 ms per IO. > @ > 	I've used two different tools over the years.   This next bit > 	is using that other tool. > E > 	Here is a measure of 111 1K IO/sec out of a 10K drive, queue depth  > 	of 1: > \ > http://groups.google.com/groups?selm=2hF0K83lfcFv%40eisner.encompasserve.org&output=gplain > F > iot /count=4000 /pattern=random/queue=1/size=2 $1$dga222:/show=histoP > _$1$DGA222: (DGX00 DG)  4000 xfer,    35.940 sec,111.3 xfer/sec,    111 Kb/sec > 9 > 	Here is a measure of 110 8K IO/sec out of a 10K drive:  > G > iot /count=4000 /pattern=random/queue=1/size=16 $1$dga222:/show=histo P > _$1$DGA222: (DGX00 DG)  4000 xfer,    36.190 sec,110.5 xfer/sec,    884 Kb/sec >  >  > 	Queue depth of 4: > G > iot /count=4000 /pattern=random/queue=4/size=16 $1$dga222:/show=histo P > _$1$DGA222: (DGX00 DG)  4000 xfer,    14.140 sec,282.9 xfer/sec,   2263 Kb/sec > D > 	you jump to 283 IO/sec.  But this is very misleading as it is NOTA > 	a single drive but a mirror so please keep that in mind if you G > 	pour through that link above.  Queue depth of 1 is a legitimate test H > 	as it will go to 1 drive of the mirror pair.  The 15K drive stuff is G > 	a single drive off a SCSI controller.  The 10K stuff is a Symmetrix  D > 	hyper (mirror).  There are other refs that show 10K drives do 110G > 	IOPS out there.  But here is a 10K drive that does 130 IOPS and this J > 	is probably legit as it has a faster seek time (4.5 ms versus a typicalI > 	value of 5.2 or 5.4 ms for "older" 10K drives, so a 15-20% improvement B > 	in seek would/should translate into a 15-20% increase in IOPS): > S > http://www.cdconsult.net/Maxtor/products/SCSI%20Drives/atlas_10k_iii_ultra320.htm  > 	 > 				Rob  >    ------------------------------    Date: 30 Dec 2003 07:56:15 -0600+ From: young_r@encompasserve.org (Rob Young) # Subject: Re: DISK DRIVE PERFORMANCE 3 Message-ID: <fWYbXIS0+tGW@eisner.encompasserve.org>   S In article <bsr7gb$k0g$1@news2.tilbu1.nb.home.nl>, Dirk Munk <munk@home.nl> writes:   F > Now if you could just tell us where you got this nice IOT tool......  C 	If you happen to stumble upon it, make sure it is recent.  I found @ 	(through testing) the random number generator wasn't so random.   				Rob    ------------------------------    Date: 30 Dec 2003 08:02:25 -0600+ From: young_r@encompasserve.org (Rob Young) # Subject: Re: DISK DRIVE PERFORMANCE 3 Message-ID: <lFHFO1gUNmu8@eisner.encompasserve.org>   S In article <bsr7gb$k0g$1@news2.tilbu1.nb.home.nl>, Dirk Munk <munk@home.nl> writes: F > Now if you could just tell us where you got this nice IOT tool...... >   G 	Another thing, TESTDEV is on freeware somewhere and it is functionally @ 	equivalent.  The thing about TESTDEV is you pass in a seed.  So8 	I've cobbled together driver .COMS that look like this:  e http://groups.google.com/groups?selm=el%2BLvp8tKs7o%40eisner.encompasserve.org&oe=UTF-8&output=gplain      				Rob    ------------------------------  % Date: Tue, 30 Dec 2003 15:33:35 +0100 - From: "Martin Vorlaender" <mv@pdv-systeme.de> # Subject: Re: DISK DRIVE PERFORMANCE 7 Message-ID: <bss2bs$rcgm$1@ID-56200.news.uni-berlin.de>    Rob Young wrote:H > Another thing, TESTDEV is on freeware somewhere and it is functionallyA > equivalent.  The thing about TESTDEV is you pass in a seed.  So 9 > I've cobbled together driver .COMS that look like this:  >  > L http://groups.google.com/groups?selm=el%2BLvp8tKs7o%40eisner.encompasserve.o rg   Just a note:  / DCL is good enough to produce random seeds; see L http://groups.google.com/groups?selm=95Aug17.125413mdt.20484%40twiki.spire.c om   cu,    Martin --F   OpenVMS:                | Martin Vorlaender  |  VMS & WNT programmer3    The operating system   | work: mv@pdv-systeme.de F    God runs the           |   http://www.pdv-systeme.de/users/martinv/:    earth simulation on.   | home: martin@radiogaga.harz.de   ------------------------------    Date: 30 Dec 2003 09:29:34 -0600+ From: young_r@encompasserve.org (Rob Young) # Subject: Re: DISK DRIVE PERFORMANCE 3 Message-ID: <zE0cIiGobPb+@eisner.encompasserve.org>   g In article <bss2bs$rcgm$1@ID-56200.news.uni-berlin.de>, "Martin Vorlaender" <mv@pdv-systeme.de> writes:  > Rob Young wrote:I >> Another thing, TESTDEV is on freeware somewhere and it is functionally B >> equivalent.  The thing about TESTDEV is you pass in a seed.  So: >> I've cobbled together driver .COMS that look like this: >> >>N > http://groups.google.com/groups?selm=el%2BLvp8tKs7o%40eisner.encompasserve.o > rg >  > Just a note: > 1 > DCL is good enough to produce random seeds; see N > http://groups.google.com/groups?selm=95Aug17.125413mdt.20484%40twiki.spire.c > om >   : 	We know.  I've been through the Random Number wars.  Many2 	battle scars, including one from one of the best:  = From: gavron@spades.aces.com (Ehud Gavron 602-885-7700x.2546)  Newsgroups: comp.os.vms 3 Subject: Re: Random number in DCL (was Questions??)  Date: 14 Feb 92 09:49:00 GMT Sender: news@sunquest.UUCP  Reply-To: gavron@Spades.ACES.COM  F In article <kpmg21INNpun@cs.widener.edu>, YOUNG@tattoo.cs.widener.edu  (Rob Young) writes... M #In <9202132327.AA28511@ucbvax.Berkeley.EDU> TJK@rmcs.cranfield.ac.uk writes:  # 9 #> Does anyone know of a random number generator in DCL ?  # L #Well... I claim to know nothing of random number generation (seems to touchJ #a number of hot spots).  And the example that follows probably could have5 #been done with a call to MTH$RANDOM, but how boring.  #  #rand.c is followed by rand.com   > 	120 lines of C code and	15+ lines of DCL deleted for brevity.   	Try: % 		$ rnd = f$element(1,".",f$cvtime())   6 	This will return a two-digit random number, which you3 	may massage as you like, or get more digits by the - 	same method and then concatenate together...    	    #				Rob5 #                                                      young@tattoo.cs.widener.edu      	Ehud    ---    	I submit this one is better:   e http://groups.google.com/groups?selm=FK9ZaWBpPVX%24%40eisner.encompasserve.org&oe=UTF-8&output=gplain      			Rob   ------------------------------  # Date: Tue, 30 Dec 2003 14:05:35 GMT 3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond) 5 Subject: Re: Floating point arithmetic support in DCL 3 Message-ID: <PCfIb.11327$yS1.9261@news.cpqcorp.net>   7 In article <3FF0F788.7862539F@NeOaSrPtAhMlNiOnWk.net>,  B "David J. Dachtera" <djesys.nospam@NeOaSrPtAhMlNiOnWk.net> writes: >Charlie Hammond wrote:  ..K >> Likewise if a5 = 5 and a4 = 4, then if a5 / a4 .eq. 1.25 and not just 1,  >> the same problem exists.  > C >Rather depends on whether A5 and A4 were "declared" (implicitly or ! >explicitly) as integer or float.  ..   Exactly the problem.    @ I "A = 5 / 4" results in an implicit declaration of A as float,  then existing code breaks.   --  J       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------    Date: 30 Dec 2003 08:31:54 -08001 From: keithparris_NOSPAM@yahoo.com (Keith Parris) 5 Subject: Re: Floating point arithmetic support in DCL = Message-ID: <cf15391e.0312300831.6578788d@posting.google.com>   s spamsink2001@yahoo.com (Alan E. Feldman) wrote in message news:<b096a4ee.0312291020.79fb54e0@posting.google.com>... D > NOTE: There exists at least one calculator app that keeps track of< > rational operations. You can divide 1 by 3 and it displaysD > 0.33333333333333333333333333333333. But you can then multiply that& > result by 3 and you get exactly 1. !  @ I doubt the calculator app is actually keeping track of rational operations.   @ Commonly, calculators keep a number of additional "guard digits"F beyond what the display width is (one I used had 4 extra digits), thenF use those guard digits to round all results to fit the display window.D  With 4 additional guard digits which end up with a value of 9999 in? this example, the result is that the display gets rounded up to " 1.00000000000000000000000000000000   ------------------------------    Date: 30 Dec 2003 11:55:34 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) 5 Subject: Re: Floating point arithmetic support in DCL 3 Message-ID: <N5PHGO2NQZ7+@eisner.encompasserve.org>   o In article <bseosf$c9rmt$4@ID-152801.news.uni-berlin.de>, Michael Unger <spam.to.unger@spamgourmet.com> writes:  > H > Yes, floating point arithmetic would be great for many of them; but itI > would break most of them if the default behaviour (i.e., integers only) > > would be changed. So the F$MATH approach would be preferred.  @    I think we've already established that floating point supportF    would have to be added in a way that is upward compatable, i.e. NOT    the default behaviour.   @    VMS has been adding things in an upward compatable manner forC    decades.  I see no reason why this would be different, or why we %    wouldn't expect Guy to make it so.    ------------------------------    Date: 30 Dec 2003 11:59:04 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) 5 Subject: Re: Floating point arithmetic support in DCL 3 Message-ID: <RSKQYXp$Btya@eisner.encompasserve.org>   W In article <1031226022919.998B-100000@Ives.egh.com>, John Santos <JOHN@egh.com> writes:  > F > I think Peter is refering to the fact that only the 1st 4 charactersB > are significant (usually?) in DCL, so an existing procedure withF > "ELSEIF" in it will act just like it said "ELSE".  If ELSEIF becomes8 > a separate keyword, the existing procedure will break.  C    Since at least VMS 2.2 the docs have recommended not abreviating D    in production-quality command procedures.  In the VMS 6.0 ReleaseC    Notes a specific warning was issued that the 4 character parsing     limit might change.  F    I see no reason DCL can't do ELSEIF.  If it breaks your procedures,    RTFM.   ------------------------------    Date: 30 Dec 2003 11:59:51 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) 5 Subject: Re: Floating point arithmetic support in DCL 3 Message-ID: <bZ+C$$folPEa@eisner.encompasserve.org>   U In article <3FECB1C8.38F8FEB3@pacbell.net>, Don Sykes <anonymous@pacbell.net> writes:  > J > For my $.02 I'd like to have an implicit symbol assignment, like we have > with > symbol := "the string"	 > maybe,   > float_symbol %= 3.17 >     I like it, I like it!   ------------------------------    Date: 30 Dec 2003 08:59:57 -08001 From: keithparris_NOSPAM@yahoo.com (Keith Parris) A Subject: Re: Linux kernel security bug ... VMS kernel rock solid! = Message-ID: <cf15391e.0312300859.29616d6f@posting.google.com>    Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote in message news:<brcu1n$m70$1@new-usenet.uk.sun.com>... 1 > The answer is that AlphaServers running OpenVMS ' > did ~64 million dollars last quarter.   @ IDC Server Quarterly Tracker numbers put OpenVMS server customer) revenues at USD $75 Million for Q3CY2003.   * > The words blip and radar spring to mind.  E This is only server revenues.  Not Services, not Storage or SANs, not > workstations, not networking hardware or software, not layered products, etc.  E One major benefit of the OpenVMS platform is its excellent investment E protection, which means OpenVMS customers don't HAVE to spend as much B on server hardware.  They can keep everything they already had andC just add capacity by adding nodes in a cluster, instead of throwing D out their old box and doing a fork-lift upgrade when they run out of	 capacity.   B And with the Itanium platform, what we're seeing is that thanks toB sharing the hardware platforms and development costs with Windows,E Linux, and HP-UX, it's starting to become evident that Itanium system F prices will be only 40 to 70% of what an equivalent Alpha system cost.D  This trend will reduce server prices further for OpenVMS customers.   ------------------------------    Date: 30 Dec 2003 09:04:14 -08001 From: keithparris_NOSPAM@yahoo.com (Keith Parris) A Subject: Re: Linux kernel security bug ... VMS kernel rock solid! = Message-ID: <cf15391e.0312300904.71afbdcb@posting.google.com>    Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote in message news:<brs6pi$1bf$1@new-usenet.uk.sun.com>... 7 > In Q401 which was immediately prior to the compaction 5 > OpenVMS server revenues were ~95 million a quarter.   O Q401 was the quarter in which 9/11 occurred, and the economy tanked after that.    ------------------------------  % Date: Tue, 30 Dec 2003 09:07:50 -0800 # From: "Tom Linden" <tom@kednos.com> A Subject: RE: Linux kernel security bug ... VMS kernel rock solid! 9 Message-ID: <NDEMLKKEBOIFBMJLCECIAEEDCJAA.tom@kednos.com>    /* -----Original Message----- ; /* From: Keith Parris [mailto:keithparris_NOSPAM@yahoo.com] + /* Sent: Tuesday, December 30, 2003 9:00 AM  /* To: Info-VAX@Mvb.Saic.ComD /* Subject: Re: Linux kernel security bug ... VMS kernel rock solid! /* /*$ /* Andrew Harrison SUNUK Consultancy; /* <Andrew_No.Harrison_No@nospamn.sun.com> wrote in message / /* news:<brcu1n$m70$1@new-usenet.uk.sun.com>... 4 /* > The answer is that AlphaServers running OpenVMS* /* > did ~64 million dollars last quarter. /*C /* IDC Server Quarterly Tracker numbers put OpenVMS server customer , /* revenues at USD $75 Million for Q3CY2003.  L Keith, I feel like stuck record.  What is included in that?  Does it include
 (1) HW sales,  (2) SW license fees, and (3) Support fees ?  ? My guess is only (1).  So what are the numbers for (2) and (3)?   - /* > The words blip and radar spring to mind.  /*H /* This is only server revenues.  Not Services, not Storage or SANs, notA /* workstations, not networking hardware or software, not layered  /* products, etc.  /*H /* One major benefit of the OpenVMS platform is its excellent investmentH /* protection, which means OpenVMS customers don't HAVE to spend as muchE /* on server hardware.  They can keep everything they already had and F /* just add capacity by adding nodes in a cluster, instead of throwingG /* out their old box and doing a fork-lift upgrade when they run out of  /* capacity. /*E /* And with the Itanium platform, what we're seeing is that thanks to E /* sharing the hardware platforms and development costs with Windows, H /* Linux, and HP-UX, it's starting to become evident that Itanium systemI /* prices will be only 40 to 70% of what an equivalent Alpha system cost. G /*  This trend will reduce server prices further for OpenVMS customers.  /* /* ---) /* Incoming mail is certified Virus Free. = /* Checked by AVG anti-virus system (http://www.grisoft.com). D /* Version: 6.0.553 / Virus Database: 345 - Release Date: 12/18/2003 /* --- & Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).A Version: 6.0.553 / Virus Database: 345 - Release Date: 12/18/2003    ------------------------------  % Date: Tue, 30 Dec 2003 15:25:47 +0000 O From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com>  Subject: Re: My Holiday Wishes0 Message-ID: <bss5ds$9pf$1@new-usenet.uk.sun.com>   Bill Gunshannon wrote: > A > This has obviously gone way OT, but as a Theology Major with an ? > interest in Traditional Catholic Liturgy I find the number of B > people in this small group who are familiar with (and apparently8 > still understand) these Latin snippets rather curious. >   ; When I was at school Latin was a compulsory subject in most  fee paying schools.    Regards  Andrew Harrison    ------------------------------    Date: 30 Dec 2003 07:47:09 -0800. From: spamsink2001@yahoo.com (Alan E. Feldman)( Subject: Re: recommendation for NPAGEVIR= Message-ID: <b096a4ee.0312300747.33ceeeca@posting.google.com>   | helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) wrote in message news:<bsqai3$qlf$1@online.de>...! > NPAGEVIR parameter information: F > *** WARNING *** The value of NPAGEVIR (12888064) computed by AUTOGENJ >                 exceeds 33% of physical memory.  Please consider whether? >                 this should be overridden by a smaller value.  >   K > I got the above message after running AUTOGEN on a system which has been  E > in the cluster for a couple of years and on which I've run AUTOGEN  F > periodically.  I've never seen the message before.  The machine has $ > 32 MB and is a VAXstation 4000/60. > D > What should I hard-code the value (or bound by MIN and MAX) to be?  @ I would try to halve it, assuming it's about 4 times the size ofF NPAGEDYN. It only needs to be as large as nonpaged dynamic memory will> ever expand to. It is basically an emergency shutoff valve for  nonpaged dynamic pool expansion.  D Keep an eye on Nonpaged Dynamic Memory. NPAGEDYN is its initial size< upon booting and NPAGEVIR is its maximum allowable size uponD expansion. There is a small cost in memory for expansion -- about 4%D of the amount of expansion, IIRC. Read about these parameters in the/ manuals. (Discussion based on VMS 6.1,6.2 VAX.)   ! What VMS version are you running?    Disclaimer: JMHO Alan E. Feldman    ------------------------------  % Date: Tue, 30 Dec 2003 14:30:11 +0000 O From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> 1 Subject: Re: Singapore Exchange to run on OpenVMS 0 Message-ID: <bss25j$8pa$1@new-usenet.uk.sun.com>   Keith Parris wrote:  > "David J. Dachtera" <djesys.nospam@NeOaSrPtAhMlNiOnWk.net> wrote in message news:<3FEBBA1C.B1960573@NeOaSrPtAhMlNiOnWk.net>... > H >>...and, as I've asked multiple times but you've consistently attemptedI >>to skirt the issue, how's VMS doing now as compared to three years ago? & >>...five years ago? ...ten years ago? >  > D > VMS revenues are definitely lower at present compared with 5 or 10F > years ago.  There was a lot of damage done by the "port to UNIX" and+ > "port to Windows NT" waves under Digital.  > D > But I'm far more interested in where VMS is now, and the directionA > revenues are heading right now.  And that looks quite positive.   9 Who can say, a rise of 4 million dollars is so tiny as to  be largely irrelevant.   Regards  Andrew Harrison    ------------------------------    Date: 30 Dec 2003 09:20:55 -08001 From: keithparris_NOSPAM@yahoo.com (Keith Parris) 1 Subject: Re: Singapore Exchange to run on OpenVMS = Message-ID: <cf15391e.0312300920.73f37176@posting.google.com>   Z JF Mezei <jfmezei.spamnot@istop.com> wrote in message news:<3FEF56D9.FC2BBCF@istop.com>...M > Yeah, very positive when you hear comments from HP execs to the effect that K > the reductions in VMS revenues were slightly less than had been expected.   A Not about VMS -- about Alphaservers.  One comment, by one HP exec ! (Wayman) -- not comments, plural.   $ I'll explain Wayman's comment again:  F Wayman was not talking about VMS attrition. He was talking about Alpha% sales. The quote from The Register at A http://www.theregister.co.uk/content/61/34091.html is: "Alpha and B Tandem revenues declined 'but a little less than what we had built: into our model.' said Wayman." The HP Quarterly Results atF http://www.hp.com/hpinfo/investor/financials/quarters/2003/q4.html sayA "Both NonStop and Alpha declined in the mid-single digits."  I've  heard 5% for Alpha.   F Alpha sales would include hardware for both VMS and Tru64. So it's notE unreasonable that since OpenVMS did well, that would keep the decline ! of the total lower than expected.   A Wayman's figures also do not include Services or Storage or other F sources of VMS revenues.  So to interpret his statement as reflecing a+ decline in VMS revenues would be incorrect.    ------------------------------    Date: 30 Dec 2003 03:05:38 -08001 From: matthew.finbow@btinternet.com (Matt Finbow)  Subject: VMS 5.5-2H4 and TZ87?= Message-ID: <ec25d2bf.0312300305.1e917ebc@posting.google.com>   E I'm helping to look after a VMS 5.5-2H4 cluster and we want to switch B from using DAT for backups to DLT, I'm curious to know whether theD TZ87 or possibly even the TZ88 works fully on VMS 5.5-2H4, including standalone backup.  D I've looked though various sources of information, the STABACKIT.COME file on 5.5-2H4 mentions the TZ87 (but not the TZ88), but the current D SPD for VMS Alpha/VAX lists the minumum VMS version for TZ87 as 6.1,E oddly however it lists the min version of VMS for the TZ88 as 5.5-2H4   B The Systems + Options catalog for tape drives (1999) lists the min" version of VMS for the TZ88 as 6.2  E The manual for the TZ877 loader (internal TZ87) lists the restriction D for 5.5-2 as no standalone backup (doesn't mention 5.5-2H4 however).  E The TZ86 is listed in the 5.5-2 SPD, but it would be easier for us if " we could use CompactTapeIII tapes.  < So to summarize, can we just switch to a TZ87 and still have standalone backup?' Can we use TZ88 drives in the same way?    Thanks in advance,   Matthew Finbow.    ------------------------------  % Date: Tue, 30 Dec 2003 15:43:30 +0000 O From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> * Subject: Re: VMS runs well on HP Superdome0 Message-ID: <bss6f3$a48$1@new-usenet.uk.sun.com>   David Froble wrote:  > JF Mezei wrote:  > H > I believe Superdome allows hard partitions.  So you could have one or : > more VMS partitions, and optionally other OS partitions. >   < The SuperDome supports nPartitions these are hard partitions: based on Cell boards. A 64 way SuperDome can support up to 16 nPartitions.   ; Sometime in the future nPartitions which are currently hard 9 will become soft, this appears currently to be technology  that is only planned for HP-UX.    regards  Andrew Harrison  >  > Dave >    ------------------------------  % Date: Tue, 30 Dec 2003 08:17:00 -0800 / From: Greg Cagle <news@removethisgregcagle.com> * Subject: Re: VMS runs well on HP Superdome/ Message-ID: <vv3981c83jhq92@corp.supernews.com>   ( Andrew Harrison SUNUK Consultancy wrote:  = > Sometime in the future nPartitions which are currently hard ; > will become soft, this appears currently to be technology ! > that is only planned for HP-UX.   < Actually vPartitions (soft partitions) have been in place on< HP-UX for some time - a few years IRRC. You are correct that? they are HP-UX only at this time. nPartitions (hard partitions)  are not OS-specific.   - Greg --  
 Greg Cagle gregc at gregcagle dot com   ------------------------------  % Date: Tue, 30 Dec 2003 12:40:44 -0500 ' From: "Main, Kerry" <kerry.main@hp.com> * Subject: RE: VMS runs well on HP SuperdomeR Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB1E10BF@tayexc19.americas.cpqcorp.net>   > -----Original Message-----; > From: Greg Cagle [mailto:news@removethisgregcagle.com]=20 " > Sent: December 30, 2003 11:17 AM > To: Info-VAX@Mvb.Saic.Com , > Subject: Re: VMS runs well on HP Superdome >=20* > Andrew Harrison SUNUK Consultancy wrote: >=20G > > Sometime in the future nPartitions which are currently hard will=20 H > > become soft, this appears currently to be technology that is only=20 > > planned for HP-UX. >=20A > Actually vPartitions (soft partitions) have been in place on=20 A > HP-UX for some time - a few years IRRC. You are correct that=20 8 > they are HP-UX only at this time. nPartitions (hard=20" > partitions) are not OS-specific. >=20 > - Greg > -- > Greg Cagle > gregc at gregcagle dot com >=20   Greg -  ! Minor correction / clarification:   C vPars (or otherwise known as soft partitions) are also available on F OpenVMS. Rather than call them vPars, OpenVMS chose to call it Galaxy,H but the technology is basically the same i.e. the ability to dynamicallyD shift CPU's between separate OS instances. Like vPars on HP-UX, softC partitions have been available on OpenVMS for a number of years.=20   F One of best demo's (imho) at the recent HP World event in St Louis wasH with the person running the booth dynamically shifting cpu's to handle a) different workload on a big Alpha GS1280.   H Oh - and by the way, the tool he was using to drag-n-drop CPU's from one+ OS instance to another was a handheld iPaq.    :-)    Regards   
 Kerry Main Senior Consultant  HP Services Canada Voice: 613-592-4660  Fax: 613-591-4477  Email: kerryDOTmainAThpDOTcom . (remove the DOT's and AT for email address)=20   ------------------------------  % Date: Tue, 30 Dec 2003 17:22:16 +0000 O From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> * Subject: Re: VMS runs well on HP Superdome0 Message-ID: <bssc88$btr$1@new-usenet.uk.sun.com>   Greg Cagle wrote: * > Andrew Harrison SUNUK Consultancy wrote: > > >> Sometime in the future nPartitions which are currently hard< >> will become soft, this appears currently to be technology" >> that is only planned for HP-UX. >  > > > Actually vPartitions (soft partitions) have been in place on> > HP-UX for some time - a few years IRRC. You are correct thatA > they are HP-UX only at this time. nPartitions (hard partitions)  > are not OS-specific. >   < That isn't what the Integrity White paper says, it indicates8 that vPartitions will be supported in a release of HP-UX1 after 11i 2 which is the current Itanium version.   7 It is quite possible that vPartitions do exist on HP-UX  HP-PA.   Regards  Andrew HArrison  > - Greg   ------------------------------  % Date: Tue, 30 Dec 2003 09:58:51 -0800 / From: Greg Cagle <news@removethisgregcagle.com> * Subject: Re: VMS runs well on HP Superdome/ Message-ID: <vv3f70eb5op6e4@corp.supernews.com>   ( Andrew Harrison SUNUK Consultancy wrote:   > Greg Cagle wrote:  > + >> Andrew Harrison SUNUK Consultancy wrote:  >>? >>> Sometime in the future nPartitions which are currently hard = >>> will become soft, this appears currently to be technology # >>> that is only planned for HP-UX.  >> >> >>? >> Actually vPartitions (soft partitions) have been in place on ? >> HP-UX for some time - a few years IRRC. You are correct that B >> they are HP-UX only at this time. nPartitions (hard partitions) >> are not OS-specific.  >> > > > That isn't what the Integrity White paper says, it indicates: > that vPartitions will be supported in a release of HP-UX3 > after 11i 2 which is the current Itanium version.  > 9 > It is quite possible that vPartitions do exist on HP-UX  > HP-PA.  < It wasn't clear you were only referring to Integrity systems in your comments.    - Greg --  
 Greg Cagle gregc at gregcagle dot com   ------------------------------  % Date: Tue, 30 Dec 2003 10:22:32 +0000sO From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com>o8 Subject: Re: What Andrew and sun can't stand is that ...0 Message-ID: <bsrjl8$4p1$1@new-usenet.uk.sun.com>   Alan E. Feldman wrote:h > Greg Cagle <news@removethisgregcagle.com> wrote in message news:<vuks8eigsasq48@corp.supernews.com>... >   H > Also, "Slowaris" is being used in direct response to a Solaris person.D > I certainly hope cov participants aren't sending mail to HP peopleH > using terms like Itanic and Chompaq. I guess it's partly just a matter > of common courtesy.i >   8 Possibly more importantly its being used in a discussion6 about OS security where it has absolutely no relevance< except as an indication that the person using the derogatory9 word (if that is what it is) has run out of collateral ton support his case.H  9 I don't really care if Bob calls Solaris Slowaris it sayss0 rather more about him than it does about the OS.   Regardse Andrew HarrisonnE > There are probably more reasons but I think this is enough for now.  >  > Alan E. Feldman    ------------------------------   End of INFO-VAX 2003.723 ************************