1 INFO-VAX	Tue, 10 Apr 2001	Volume 2001 : Issue 200       Contents:' Re: - OpenVMS ever to be on Intel chip?  Re: CD backups.  CD writer for a VaxStation Re: CD writer for a VaxStation Re: CD writer for a VaxStation+ Re: CI based cluster for max I/O perf. How? + RE: CI based cluster for max I/O perf. How? + Re: CI based cluster for max I/O perf. How? + Re: CI based cluster for max I/O perf. How? ? clrref on Alpha, or the deadly queue bug survives Alpha upgrade C Re: clrref on Alpha, or the deadly queue bug survives Alpha upgrade C Re: clrref on Alpha, or the deadly queue bug survives Alpha upgrade C Re: clrref on Alpha, or the deadly queue bug survives Alpha upgrade C Re: clrref on Alpha, or the deadly queue bug survives Alpha upgrade C Re: clrref on Alpha, or the deadly queue bug survives Alpha upgrade C Re: clrref on Alpha, or the deadly queue bug survives Alpha upgrade  Re: Compaq & VMS Careers3 Re: Compaq quitting with Alpha Workstations ???????  Re: Corba on OpenVMS AXP Re: Corba on OpenVMS AXP  Re: DEC Server software question error installing patch on vax?? # Re: error installing patch on vax?? 2 Re: Flipping from big-endian to little endian in C
 Re: FMS files  From the trenches  Re: FTP hijacking of VMS sites6 Re: Galaxy Network Interfaces/cards and TCP/IP Sockets. Re: Help needed with RMS Character Field IndexA Re: How to tell if a system is booting immediately after a crash. " Re: More VMS ads spotted in the UK, Re: NSLOOKUP: *** Can't initialize resolver., Re: NSLOOKUP: *** Can't initialize resolver. RE: OT: "The Puzzle Palace"  RE: OT: "The Puzzle Palace"  RE: OT: "The Puzzle Palace"  RE: OT: "The Puzzle Palace"  Re: OT: "The Puzzle Palace" " Re: Problems with CSWS and CGI-BIN3 Re: Q: Converting a VAX 6000-420 to a VAX 6000-620? 3 Re: Q: Converting a VAX 6000-420 to a VAX 6000-620? * Re: Running a Diagnostic on a Network Card3 Re: Solaris tmpfs (was Re: VMS-Related: Affordable) 3 Re: Solaris tmpfs (was Re: VMS-Related: Affordable)  Re: Spice for OpenVMS Alpha? Re: Spice for OpenVMS Alpha? Re: VMS-Related: Affordable  Re: VMS-Related: Affordable  Re: VMS-Related: Affordable  Re: VMS-Related: Affordable  [Q] Adding Memory to a MV 3100" Re: [Q] Adding Memory to a MV 3100" Re: [Q] Adding Memory to a MV 3100  F ----------------------------------------------------------------------   Date: 9 Apr 2001 21:13:59 GMT & From: davidsen@tmr.com (bill davidsen)0 Subject: Re: - OpenVMS ever to be on Intel chip?9 Message-ID: <9at8mn$h3i$1@newssvr05-en0.news.prodigy.com>   9 In article <3acf59c3$0$830$8eec23a@newsreader.tycho.net>, , Bill Pedersen <pedersen@ccsscorp.com> wrote: | Motivation is: | H |     1) the continued support of existing OpenVMS VAX based applicationI | without port to OpenVMS Alpha - especially long term given there are no , | longer any VAX systems being manufactured. | 2 |     2) small, portable development environments. | C | These are the primary draw that I see for the product at present.  | I | There does seem to be some interest in it, in talking with the folks at L | Software Resources International they see between 100 and 200 downloads of | the hobbyiest version DAILY. | J | SMP is not the direction being pursued at the present time.  It uses SMPH | already in that each thread in the emulator (each device) is a thread.  G   That's what I was asking, if the emulator runs faster on N processors 5 for larger values of N. And it sounds like yes to me.   G   I do know some people who have written large systems for VMS, I'm not H sure if they have source, given that they lost the support documentation@ I can't believe they do. They are probably a potential customer.   --  <   bill davidsen <davidsen@tmr.com>  CTO, TMR Associates, IncG "I am lost. I am out looking for myself. If I should come back before I ; return, please ask me to wait."  -seen in a doctor's office    ------------------------------  $ Date: Mon, 9 Apr 2001 22:49:31 +0100; From: "Leigh G. Bowden" <LGBowden@bowdenfamily.fsnet.co.uk>  Subject: Re: CD backups./ Message-ID: <9atanq$1d6$1@newsg3.svr.pol.co.uk>   $ Got plenty of disks less than 650MB.  K Raided some old VAXstation 4000 VLC machines that had Seagte ST3600N disks. K So I've got one in a VAXstation 4000/60 and four more "spare" should I need 1 them? I knew those VLC's were good for something!   - All the other VLC's have only got the RZ23's.   ! John Vottero wrote in message ... 6 >Good luck finding a hard disk that's less than 650MB. > I >It seems easier to use one of the virtual disk drivers (VD or LD).  Then  you H >can create a virtual disk that's any size you need and you can skip theJ >mount/copy part.  Another advantage is that you can create a virtual disk5 >only the size you need so that the burns go quicker.  > G >"Leigh G. Bowden" <LGBowden@bowdenfamily.fsnet.co.uk> wrote in message * >news:9aqi1r$id2$1@newsg1.svr.pol.co.uk...0 >> The way I read it I have to do the following: >># >> Attach a CD writer to a machine. H >> Put what ever I want to preserve onto a HDD which is less than 650MB. >> Mount this HDD as /FOREIGN.< >> Then do a COPY disk: another_disk:[000000.cds]cd_disk.rom9 >> This generates a few whinges but these can be ignored. J >> Use CDRECORD freeware utility to burn the CD with the file cd_disk.rom.J >> Ensure that the machine doing the burning is doing nothing else to stop" >> interrupting the write process. >>J >> Leigh G. Bowden wrote in message <9afu0b$ec4$1@newsg3.svr.pol.co.uk>...L >> >I suspect that this has been covered before but I can't find it. We haveD >> >quite a few remote sites with master/slave pairs of uVAX 3100/nn machinesJ >> >all running VMS6.2. Each has three SCSI 2.1GB disks and a CD. I'll not goK >> >into the history but there is no way of doing backups of these machines H >> >easily - there's certainly no tape drives and nobody to really put a tape >> in  >> >reliably from day to day.  >> >J >> >Should a disk fail it is fairly easy to recreate it's data on the data >> disksJ >> >from it's mirror but the system disk is harder to replace. There isn't@ >> >enough space on any of the other disks for a system saveset. >> >L >> >I was wondering therefore about using the CD. I know VMS doesn't supportK >> >CD-W type devices. If I create a saveset on VMS machine and then FTP it  >in J >> >binary mode to a PC and then blat it on a CD also in binary mode would >this 	 >> >work?  >> >J >> >The idea being to have standalone backup on all the other HDD and boot >from L >> >them. Mount the CD and restore the contents. It sounds to easy there has >to  >> >be a catch.  >> > >> > >> >> >  >    ------------------------------  % Date: Sun, 08 Apr 2001 21:38:46 +0100 . From: Roger@natron.demon.co.uk (Roger Barnett)# Subject: CD writer for a VaxStation - Message-ID: <158762902wnr@natron.demon.co.uk>   @ Can anyone recommend a CD writer for use with a VaxStation 3100 0 [ and if so, whether it will work with VMS 6.2 ]   Thanks  
 Roger Barnett . Natron Software Maintenance Ltd, York, England   ------------------------------  # Date: Mon, 09 Apr 2001 22:53:29 GMT 2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)' Subject: Re: CD writer for a VaxStation 2 Message-ID: <J3rA6.713$fB6.18649@news.cpqcorp.net>  ^ In article <158762902wnr@natron.demon.co.uk>, Roger@natron.demon.co.uk (Roger Barnett) writes: : A :Can anyone recommend a CD writer for use with a VaxStation 3100    H   Please search for "(CD-R)" in the OpenVMS FAQ, and follow the links...  N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Mon, 09 Apr 2001 20:29:44 -0500 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> ' Subject: Re: CD writer for a VaxStation ' Message-ID: <3AD26208.FBCAC026@fsi.net>    Roger Barnett wrote: > A > Can anyone recommend a CD writer for use with a VaxStation 3100 2 > [ and if so, whether it will work with VMS 6.2 ]  D Take a look near the "bottom" of Brian Schenkenberger's CD-ROM page:   http://www.tmesis.com/cdrom/  D ...for some important notes re: CD-R on MicroVAX/VAXstation 3100. HeE doesn't actually recommend a specific CD-R drive on that page, though 0 he's mentioned Yamaha's CRW4416S in other posts.   --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/   F This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings is to be expected.  @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.    ------------------------------   Date: 9 Apr 2001 09:03:10 -0500 + From: young_r@encompasserve.org (Rob Young) 4 Subject: Re: CI based cluster for max I/O perf. How?3 Message-ID: <2e$oMTdUOu7d@eisner.encompasserve.org>   \ In article <87g0fia208.fsf@prep.synonet.com>, Paul Repacholi <prep@prep.synonet.com> writes:/ > young_r@encompasserve.org (Rob Young) writes:  > E >>      cache.  Finally, you will need to go to two stars with HSJ80s C >>      as they will saturate a single star or so I have been told. 4 >>      Also, they have TWO sets of ports (A and B).3 >                        ^^^^^^^^^^^^^^^^^^^^^^^^^^  > C > For clarification Rob, By the above I assume you mean two sets of C > dual-ports, ie 4 CI cables. 2 A cables and 2 B cables, to use the  > traditional A-B names. >   = 	Yes , Compaq talks about "Dual CI ports for overall improved  	performance":  V http://www.compaq.com/products/storageworks/array-and-scsi-controllers/HSJ80index.html  G 	Hard for me to be precise that late (early?) in the day.  And I still  C 	don't have a term (nor like how they describe it as it is somewhat 1 	confusing to a novice).  But I count 8 cables.     & 	"Set 1" or "CI Port 1" or "CI Card 1"   		Port A receive
 		Port A send  		Port B receive
 		Port B send   & 	"Set 2" or "CI Port 2" or "CI Card 2"   		Port A receive
 		Port A send  		Port B receive
 		Port B send   ? 	Ports is ports I guess is what I struggle with.  Maybe talking = 	about adding support for dual-CI cards, CIPCA or CIXCD would  	help me out a bunch.   E 	One other issue in there that comes to mind.  Config 2 is crying out E 	for shadowing across separate controller pairs.  The situation could G 	be that you take a CIPCA hit early in the a.m. and have to grind along H 	(via MSCP) until an emergency downtime early p.m., that may or may not C 	be acceptable.  I do enjoy the fibre stuff as your pathing is more H 	dynamic, less management, etc. and as Dirk highlights also much higher  	bandwidth.    				Rob    ------------------------------  % Date: Mon, 09 Apr 2001 19:02:28 -0500 + From: "Main, Kerry" <Kerry.Main@compaq.com> 4 Subject: RE: CI based cluster for max I/O perf. How?R Message-ID: <DC4745D1A85CA04180C83CDC706A9D180D94F2@cthexc02.americas.cpqcorp.net>   Lyndon,   J A few folks have already provided some good feedback and I will just agree) with Rob and and add a few more tidbits -   @ - from a technology perspective, SAN's (using the FW and OpenVMSK versions/maint updates available today and especially V7.3) technology will K outperform CI clusters in almost all situations of straight IO performance. K The CI does have some things the SAN stuff does not such as all SCS traffic L must be handled by separate adapter such as NIC, MC, fddi etc. However, withJ load balanced GbE, ATM type NIC connections in V7.3, the SCS traffic will, imho, become less of an issue.  H - from a very general RAID perspective, assuming cost is not the biggestJ issue, my $.02 is to use RAID5 for capacity and RAID 0+1 for performance.   L - For higher availability environments where cost is definately not an issueF as much, then a combination of HW RAID and Host based Volume shadowingH (HBVS) can be used. For multi-site clusters, this allows a drive to failJ without impacting the local server or network link when it is repaired and, needs to be brought back into the shadowset.  L - since losing a drive with a page file could impact a system, if pricing isL not the biggest issue, I would recommend using RAIDx for ALL drives and thatJ includes system, page/swap drives as well. The sites I have config'ed haveL ALL drives on the SAN with a local drive for scratch type stuff (always niceH to have a backup copy of the OS on the local drive to troubleshoot with)  K - latest on SAN and OpenVMS info (including a just updated WP on multi-site  clusters) can be found at:F http://www.openvms.compaq.com/openvms/fibre/index.html See the updated) documentation at the end of the web page.    Regards,  
 Kerry Main Senior Consultant  Compaq Canada Inc. Professional Services  Voice: 613-592-4660  Fax  :  819-772-7036 Email: Kerry.Main@Compaq.com     -----Original Message-----5 From: Lyndon Bartels [mailto:lbartels@pressenter.com]  Sent: April 8, 2001 1:56 PM  To: Info-VAX@Mvb.Saic.Com 4 Subject: Re: CI based cluster for max I/O perf. How?     Hmmm.... To follow up.  H I've been looking at the manual "Guidelines for cluster configurations,"G Chapter 9. There are four configurations there that I'm interested in.    E Figure 9-1, is our current configuration. It has two CIPCAs per host, B and two HSJ per storage subsystem. All going through a single star coupler.  F Now according to this configuration, it looks like the second CIPCA is not helping performance.  H The second configuration looks a bit more promising. That introduces theC a second star coupler. This one states that the second star coupler F provides increased I/O performance and availability. The disadvantagesD here are: If a CIPCA fails disks won't fail-over to the other CIPCA.  G Is this still true? This manual was written in 1999. Is that still true 9 with the v7.2-1? Also, the manual states somewhere to use F decw$examples:prefer to set the paths of disks to load balance betweenF HSJs. Bit *NOT* to use the HSJ set preferred command because it cannot= be overwridden by a host "Prefer" or IO$_SETPRFPATH modifier.   B Configuration 3 (Figure 9-3) has the A Paths going through on starE coupler, and the B Paths going through the second coupler. This looks F like there is no improvement in I/O performance. But there seems to be more availability.     C Configuration 4 (Figure 9-4) splits the disks and HSJs in half. I'm H going to stay away from this one, because to get availability, I'll need< to do host-based shadowing, and that'll take away some of my performance.  ? So it looks like configuration 2 or 3 is to be the one to use.      E Does anyone have experience with either of these configurations? Each E CIPCA and HSJ has it's own ID. Do the IDs still have to be unique? Or E are the numbers duplicated? For example, the the CIPCAs in Host 1, In C config 2, do they have the same ID? I'm sure they must be unique in 	 config 3.     G I have questions: Is the fail-over still an issue? Is the documentation G regarding the HSJ SET_PREFER command still valid? There doesn't seem to : be any I/O performance difference between the two configs.    Next, as for the HSJs and disks:  G I'm using HSJ50s and 8 bit I/O personality modules. But "VW" drives. Am F I shakling my I/O throughput by using the 8 bit I/O modules? Would the 16 bit I/O module be better.    = I'm thinking about replacing HSJ50s (considering they're near F end-of-life) with HSJ80s. Would I then "Have to" upgrade to the 16 bit I/O modules.  @ Disk-drive layout: I need both I/O speed and availabilty. So I'mD thinking RAID 0+1. Since there are six (6) buses on a given HSJ. I'mE thinking of using 6 physical disk per storage unit. 3 members striped F and 3 more mirroring. Laying them out so that each physical is on it's own bus.  D Considering the discussion regarding pagefiles/disk loss. Would I beH better off not having locally hosted SCSI disks for page files, and moveC the page files onto CI based storage? I'm reluctant to do this. I'm = thinking I may loose paging performance.... But I'm not sure.     . Any advice, thoughts, experiences are welcome.   Thanks in advance,   Lyndon   -- uG My opinions are mine and mine alone. They seldom align with those of myb	 employer.D   ------------------------------  % Date: Mon, 09 Apr 2001 22:29:05 -0500 . From: Lyndon Bartels <lbartels@pressenter.com>4 Subject: Re: CI based cluster for max I/O perf. How?. Message-ID: <3AD237B1.335393C0@pressenter.com>  E I'm pretty sure I'll configure RAID 0+1 sets for best performance ande availability.   H One of the things I'm curious about is the "chunk size" when configuringG a stripe or raid set (controller based). The HSJ50 configuration manuall? says. "A small chunck size relative to the average request sizeeH increases the data the transfer rate...." My question is... How does oneF go about finding out about the "average request size." I looked in the? manuals under Monitor and Show, but not much help there.... Any A thoughts/ideas? I would think that a chunk size that was either aAF multiple of the average request size, or evenly divisible by that sameH number would be a good thing. Not as much wasted. And wouldn't the disksE cluster size need to be considered as well? Shouldn't there be a nice ' simple relationship between the three? t  
 So to recap:    9 How do I find the "average request size" on a given disk?dF Should the chunk size, cluster size and average request size have some correlation?    F Another question. I need to determine which drives are most in need ofD the performance gain a stripe set can give me. I need to know how toD measure disk access activity. Other than purchasing a tool are thereF some options? Reads per second, writes per second, that sort of thing.E Nothing extravagant. But just a good solid measure. When I do a "showIG dev d" I see the transaction count column message. What does that mean?tE I know it something I should already know. But I haven't been able toe@ find it in the manuals yet. And I Have been reading. Believe me.  G I know this may seem like a waste of everybody's time. There are faster H methods of accomplishing better disk performance. But this is for a testD system. Money is better spent on production performance. And besidesD this is also a theoretical discussion as well. I'm learning a lot. I3 thank you all for your patience and understanding. -    C I realize that a fibre SAN will outperform a CI based solution. I'meH trying to squeeze the most performance out of an aging machine as I can.F When the hardware is already bought and paid for, and more importantlyC still works extremely well, why get rid of it? (That's a rhetoricalnH question, it doesn't need answering. ;-)) I seem to recal a tidbit aboutD Seymour Cray (sp?) When asked to donate a Cray I to the smithsonian.9 "It's faster than anything anyone else makes.... so why?"e  # Anyway, thanks again for your time,    Lyndon -- sG My opinions are mine and mine alone. They seldom align with those of myr	 employer.    ------------------------------  % Date: Tue, 10 Apr 2001 00:28:38 -0400 2 From: rdeininger@mindspring.com (Robert Deininger)4 Subject: Re: CI based cluster for max I/O perf. How?L Message-ID: <rdeininger-1004010028380001@user-2iveat9.dialup.mindspring.com>  = In article <3AD237B1.335393C0@pressenter.com>, Lyndon Bartelse  <lbartels@pressenter.com> wrote:       > When I do a "show I > dev d" I see the transaction count column message. What does that mean?uG > I know it something I should already know. But I haven't been able tosB > find it in the manuals yet. And I Have been reading. Believe me.  I Transaction count is approximately the number of open files on the devicedH from the node in question.  Slightly more precisely, it is the number ofG reasons the drive can't be dismounted on this node.  On a quiet system,cJ the transaction count on the boot disk will be several hundred or more.  AJ few (at least one) of the transactions are for the index file or somethingI like that.  Such things won't stop you from dismounting.  (This is fairly 4 off the top of my head, so don't take it as gospel.)  D The transaction count should correlate closely with SHOW DEVICE/FILE  - I don't know where this is documented either.    -- s Robert Deininger rdeininger@mindspring.com0   ------------------------------  # Date: Mon, 09 Apr 2001 16:07:04 GMT + From: rjordan@mars.mcs.net (Richard Jordan) H Subject: clrref on Alpha, or the deadly queue bug survives Alpha upgrade4 Message-ID: <I6lA6.7735$JN5.172017@news.goodnet.com>  E We have a customer who, due to both some antique code and procedures tF still in use, has an unusual printing setup for things like checks andE other special documents that cannot risk overprinting by another job.AF Instead of using forms/stock types, etc, the process that is doing theE special print runs will stop a queue, wait for it to quiesce, despoolnF the port, allocate the port, and write its output directly to the portE after having the user verify the correct stock is in the printer.  On E completion the user would replace the standard stock, and the programv1 would deallocate, respool, and restart the queue.i  C This had its birth in the need to very carefully align ths stock in_E old dot matrix or line printers, though the customer is in fact using> lasers for everything now.  G On their old system (MV3100-95, VMS V5.5-2H4, all relevant patches), weiD would see a problem every few weeks where the reference count on theG actual terminal device feeding the queue would be wrong; sometimes whensC the queue was stopped a SHOW DEV/FULL would still list the symbiont G and its PID as owners with a refcnt of 1.  Sometimes we'd get everytingpD stopped and despooled and see a refcnt of 65534.  After spending theB better part of 2 months on the phone with CSC trying to fix it, we= got the 'clrref' utility and just used it whenever things got-C screwed up, assuming it was a VMS issue in V5.5-2H4, since no other>D customer ever had similar problems except at that version ( which we( were stuck at there ffor other reasons).  E Well, they other reasons are history.  They just completed an upgradeeD to a DS10, 256MB, OpenVMS V7.2-1 with all relevant patches availableG thropugh February (got a couple more pending install).  Today the prob- A lem recurred.  We now have a terminal device stuck with a phantomsC process owning it and holding refcnt at 1.  No other process on the A system has this device open.  And we're not sure if trying to use E clrref is an option, since its 1993 code (I'm reviewing it but I'm no  systems programmer).  E Code on the system is BASIC, with DCL procedures used for the printer D controls (start, stop, de/respool, etc).  Queues are all LAT feedingE DECserver90L+ or MUXserver 90/DECmux 300 ports.  The problems histor-aF ically happened about 10 times on the DECserver printers to each eventE on the MUXservers.  We really had hoped and expected that the probelmaC would be eliminated by the system upgrade, but apparently it isn't i# a VAX or VMS 5.5-2 issue after all.o  B Any assistance appreciated.  Email would be great due to the awfulC reliability of winstar/mcs in providing service here, but I'll keepm& an eye through dejagoogle too.  Thanks   Rich Jordan: rjordan@mcs.netr   ------------------------------  # Date: Mon, 09 Apr 2001 20:27:48 GMTa2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)L Subject: Re: clrref on Alpha, or the deadly queue bug survives Alpha upgrade2 Message-ID: <8XoA6.705$fB6.18505@news.cpqcorp.net>  b In article <I6lA6.7735$JN5.172017@news.goodnet.com>, rjordan@mars.mcs.net (Richard Jordan) writes:G :Instead of using forms/stock types, etc, the process that is doing theeF :special print runs will stop a queue, wait for it to quiesce, despoolG :the port, allocate the port, and write its output directly to the portfF :after having the user verify the correct stock is in the printer.  OnF :completion the user would replace the standard stock, and the program5 :would deallocate, respool, and restart the queue....   H :...where the reference count on the actual terminal device feeding the L :queue would be wrong; sometimes when the queue was stopped a SHOW DEV/FULL J :would still list the symbiont and its PID as owners with a refcnt of 1.    E   Ensure that SHARE privilege is disabled whenever assigning channelseI   to devices -- the sequence you are describing is similar to that which oH   arises when a process uses SHARE privilege to connect to an otherwise I   non-shareable device, and the channel deassignment order has the SHARE .H   process deassigning its channel last.  This particular (mis)behaviour G   is (unfortunately) completely expected, and exceedingly difficult to  G   purge from OpenVMS -- having a process with SHARE privilege involved sH   makes the last channel deassign of a non-shared channel "interesting".  F   Short answer: turn OFF use of SHARE privilege, and try to duplicate.  N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Mon, 09 Apr 2001 21:13:49 GMTY+ From: rjordan@mars.mcs.net (Richard Jordan)yL Subject: Re: clrref on Alpha, or the deadly queue bug survives Alpha upgrade4 Message-ID: <hCpA6.7746$JN5.172644@news.goodnet.com>  H >:...where the reference count on the actual terminal device feeding theL >:queue would be wrong; sometimes when the queue was stopped a SHOW DEV/FULLI >:would still list the symbiont and its PID as owners with a refcnt of 1.  > G >   Ensure that SHARE privilege is disabled whenever assigning channelssJ >   to devices -- the sequence you are describing is similar to that whichI >   arises when a process uses SHARE privilege to connect to an otherwisepJ >   non-shareable device, and the channel deassignment order has the SHARE) >   process deassigning its channel last.a   Hoff,uG      thanks for the information.  On initial inspection, the SHARE privr6 is not in use; the actual LTA devices have protection:E      S:RWPL,O:RWPL,G,W:RWPL on them (again, historical) so anyone can D open the device for write access; programs that do not have specificC stock or alignment needs simply get their output spooled into jobs.o  D      The users all have OPER priv (historical, and this is a captiveB environment so no DCL access) for queue control (changing this hasB been a priority for me but not for the customer or my management).  E      I'll get a programmer here to point me to the right source code. F Is there any implicit 'share' access that could cause a situation likeC this from BASIC open statements?  I will see what I can find there.s  D      Maybe this will finally be the impetus to get them to switch toE using locks, and setting up the queues properly.  I think that clrref-4 just may not work here anymore (even if it would ;).   Rich JordanK rjordan@mcs.net@  p   ------------------------------  # Date: Mon, 09 Apr 2001 21:28:41 GMTD2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)L Subject: Re: clrref on Alpha, or the deadly queue bug survives Alpha upgrade2 Message-ID: <dQpA6.711$fB6.18534@news.cpqcorp.net>  b In article <hCpA6.7746$JN5.172644@news.goodnet.com>, rjordan@mars.mcs.net (Richard Jordan) writes:. :     ...On initial inspection, the SHARE priv7 :is not in use; the actual LTA devices have protection:l' :     S:RWPL,O:RWPL,G,W:RWPL on them...   F   A juxtaposition...  SHARE has nothing to do with device protections,D   ACLs, or similar.  SHARE has everything to do with the ability to C   associate a second I/O channel with an existing (and non-shared) SC   I/O device.  Not only can this mess up the last-channel deassign,eB   but it can also mess up the device responses as it is not alwaysE   clear which of the applications with a channel open to a non-shared F   device should get the I/O -- assumptions around exclusive access areL   why various sorts of I/O devices are created and marked as non-shareable,    of course.  F :     I'll get a programmer here to point me to the right source code.G :Is there any implicit 'share' access that could cause a situation likehD :this from BASIC open statements?  I will see what I can find there.  H   Enable OpenVMS security alarms for use of SHARE privilege, as a start.G   Also look for anything that is enabling privileges, and particularly  G   anything that is setting (or has by default) SHARE or ALL privileges.cG   If the application has SHARE privilege enabled, then most any channeleI   assignment command (language-specific OPEN, sys$open, sys$assign, etc)  =   can silently succeed when it probably should have failed...   E :     Maybe this will finally be the impetus to get them to switch totF :using locks, and setting up the queues properly.  I think that clrref5 :just may not work here anymore (even if it would ;).s  B   The OpenVMS VAX code to clobber the kernel data structures will G   (obviously) not work on OpenVMS Alpha -- well, not without some help.sG   Use of $sndjbc or similar would be preferable, and there are ways -- oI   using the OpenVMS queue manager -- to handle the necessary print queue     paper set-up steps..  N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Mon, 09 Apr 2001 21:33:47 GMTf+ From: rjordan@mars.mcs.net (Richard Jordan)bL Subject: Re: clrref on Alpha, or the deadly queue bug survives Alpha upgrade4 Message-ID: <%UpA6.7747$JN5.172683@news.goodnet.com>  G Confirmed the following: no users/processes with SHARE priv (unless thesE symbionts themselves are enabling it; LATSYM is the only one in use).m2 BASIC open statements to print device are minimal:<   OPEN 'LTA112:OE021.L61' FOR OUTPUT AS FILE 11%, MAP PRTBUF  G so no special options in use.  I'm working though a privilege reduction F test, and will see about trying removing OPER priv and using an ACL onE the queue to control queue control operations.  Unfortunately we are  G somewhat at the mercy of murphy for a repeat performance to know if thet9 changes are working or not... trying to prove a negative.e   Rich Jordane rjordan@mcs.neta   ------------------------------  # Date: Mon, 09 Apr 2001 22:30:38 GMTg+ From: rjordan@mars.mcs.net (Richard Jordan)'L Subject: Re: clrref on Alpha, or the deadly queue bug survives Alpha upgrade4 Message-ID: <iKqA6.7751$JN5.172950@news.goodnet.com>  6 >In article <hCpA6.7746$JN5.172644@news.goodnet.com>, 5         rjordan@mars.mcs.net (Richard Jordan) writes:p( >:     S:RWPL,O:RWPL,G,W:RWPL on them...  G >  A juxtaposition...  SHARE has nothing to do with device protections,?E >  ACLs, or similar.  SHARE has everything to do with the ability to ,D >  associate a second I/O channel with an existing (and non-shared) D >  I/O device.  Not only can this mess up the last-channel deassign,C >  but it can also mess up the device responses as it is not alwaysvF >  clear which of the applications with a channel open to a non-sharedG >  device should get the I/O -- assumptions around exclusive access areaM >  why various sorts of I/O devices are created and marked as non-shareable, g
 >  of course.h   Hoff,/I      thanks; that much I did know, partly by reading up on its since yourrG previos post; I was just posting additional info about the environment,lD albeit in poor location and context.  It does appear that nobody is  using SHARE priv on the system.a   >bH >:Is there any implicit 'share' access that could cause a situation like" >:this from BASIC open statements? >dI >  Enable OpenVMS security alarms for use of SHARE privilege, as a start. H >  Also look for anything that is enabling privileges, and particularly H >  anything that is setting (or has by default) SHARE or ALL privileges.H >  If the application has SHARE privilege enabled, then most any channelJ >  assignment command (language-specific OPEN, sys$open, sys$assign, etc) > >  can silently succeed when it probably should have failed... >   A The apps are not installed, so no privs via that route (they will F be installed only after a full month long cycle has completed, to easeE making changes to programs needed by the Alpha transition).  The VAX  ) had them installed, but never with privs..  B Thanks for the idea; its been so long since I needed to track privD usage that I hadn't even thought of using alarms.  I'll get them set up asap.  F >:     Maybe this will finally be the impetus to get them to switch to2 >:using locks, and setting up the queues properly. >iC >  The OpenVMS VAX code to clobber the kernel data structures will aH >  (obviously) not work on OpenVMS Alpha -- well, not without some help.H >  Use of $sndjbc or similar would be preferable, and there are ways -- J >  using the OpenVMS queue manager -- to handle the necessary print queue  >  paper set-up steps. >u  C The clrref code indicated it would work on Alpha, but its age puts  E it in the V1.5 to V6.1 timeframe.  Given the magnitude of the changesiE since that time I'm not going to try it on a production V7.2-1 systemn maybe at home when I have time.e  D And we do have other customers running more (ahem) recent code usingD $sndjbc and normal queue criteria; this stuff was written in 1980 orA thereabouts and still creaks along, waiting for someone with the t authority to say 'fix it'...  6 THanks very much for taking the time to respond again.   Rich Jordanc rjordan@mcs.net    ------------------------------  % Date: Mon, 09 Apr 2001 20:23:14 -0500o1 From: "David J. Dachtera" <djesys.nospam@fsi.net>hL Subject: Re: clrref on Alpha, or the deadly queue bug survives Alpha upgrade' Message-ID: <3AD26082.93C08B98@fsi.net>t   Richard Jordan wrote:e > I > Confirmed the following: no users/processes with SHARE priv (unless theoG > symbionts themselves are enabling it; LATSYM is the only one in use).b4 > BASIC open statements to print device are minimal:> >   OPEN 'LTA112:OE021.L61' FOR OUTPUT AS FILE 11%, MAP PRTBUF  B Eeeewwww!!! A MAP on a PRINT(er) channel! Lots of trailing blanks!4 Wasted disk space! Wasted bandwidth on the network!   H Of course, some things actually expect Fixed_132 format for print files.  # Ah! The wonders of "Legacy" app.'s!n   -- . David J. Dachterao dba DJE Systemsn http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/t  F This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings is to be expected.  @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.o   ------------------------------  # Date: Mon, 09 Apr 2001 20:00:13 GMT 0 From: sander@vmsbiz.enet.dec.com (Warren Sander)! Subject: Re: Compaq & VMS Careersr2 Message-ID: <hxoA6.700$fB6.18341@news.cpqcorp.net>  $ re: compaq openvms jobs.. at compaq.  I A lot of the jobs listed have been put on 'hold' until after the layoff's H etc. There were a bunch and a lot did get filed in the past few months.   I Also don't forget the openvms.monster.com board which has several hundred- OpenVMS jobs listed.     -- iB ------------------------------------------------------------------6 Warren Sander                        OpenVMS MarketingD Compaq Computer Corporation          Work:  warren.sander@compaq.comE 200 Forest Street MR01-3/J1          Personal: sander@ma.ultranet.comk3 Marlboro, MA 01752                   (508) 467-4875n6    My opinions are my own and I only speak for myself -          Read http://www.openvms.compaq.com/  B ------------------------------------------------------------------   ------------------------------  $ Date: Mon, 9 Apr 2001 14:26:47 -04005 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>e< Subject: Re: Compaq quitting with Alpha Workstations ???????2 Message-ID: <g9nA6.691$fB6.18366@news.cpqcorp.net>  A Carl Perkins wrote in message <6APR200120235688@gerg.tamu.edu>...TE >In article <TGkz6.625$fB6.16807@news.cpqcorp.net>, "Fred Kleinsorge" ' <kleinsorge@star.zko.dec.com> writes...o     >}Other than that,D >}workstations have pretty much been servers without some of the RAS	 features.d >} >rI >I think you have forgotten the XP1000. It may not have "something to the1G >platform itself specific to graphics", but it has hever been sold in ar1 >configuration other than the XP1000 workstation.a >g    ? The XP1000 conforms to the above pretty much.  While built as a J "workstation" there isn't anything special about it other than eliminationL of some features that would have gone into a Server platform.  If it did notL exist, the DS10 design (with a faster CPU, and some other tweaks) could have( just as easily been used for the XP1000.   ------------------------------  # Date: Mon, 09 Apr 2001 22:14:31 GMTo2 From: seibel_r@localhost.localdomain (Rich Seibel)! Subject: Re: Corba on OpenVMS AXPg; Message-ID: <slrn9d4d1r.gh8.seibel_r@localhost.localdomain>o  I I apologize for the previous post.  Somehow my edit did not get included.e  F On Mon, 9 Apr 2001 09:41:23 +0200, Jakob Erber <erberj@post.ch> wrote:L >Despite all signs that OpenVMS will be there for a long time to come, thereH >is still support missing for software, I find vital for an important OS >plattform.M >0J >For example, as far as I can see, there is no Corba product available for2 >VMS, which offers proffessionell support as well. >aK >There are some very good freeware Orbs, but can you use them for a missionp >critical system?  >l  K The ACE Orb (TAO) is one such orb.  It is used by many companies world-wides" for mission critical applications.  G If you are interested in TAO for OpenVMS contact Malcolm Spence at OCI.n spence_m@ociweb.com 
 (314)579-0066i    
 >best regardsc >e >Jakob >a >j     -- iD --------------------------------------------------------------------D Rich Seibel, Software Engineer                 (314)579-0066 ext 220D Object Computing, Inc.                           seibel_r@ociweb.comD Need ACE training?                      See http://www.theaceorb.comD --------------------------------------------------------------------   ------------------------------  % Date: Mon, 09 Apr 2001 21:44:55 -0400e* From: Eric Ebinger <eebinger@telocity.com>! Subject: Re: Corba on OpenVMS AXPg8 Message-ID: <003f01c0c15f$d8a5eae0$e900a8c0@teamrdb.com>  F If you look at the OpenVMS stratigies and directions presentation from
 February 2001nA (http://www.openvms.compaq.com/presentations/openvmsstrategy.htm).9 You will see that Iona Orbix 2000 is being ported to VMS.o   Eric Ebinger ----- Original Message -----" From: Jakob Erber <erberj@post.ch> To: <Info-VAX@Mvb.Saic.Com>r$ Sent: Monday, April 09, 2001 3:41 AM Subject: Corba on OpenVMS AXP0    G > Despite all signs that OpenVMS will be there for a long time to come,  thereeI > is still support missing for software, I find vital for an important OS, > plattform. >tK > For example, as far as I can see, there is no Corba product available for 3 > VMS, which offers proffessionell support as well.f >nL > There are some very good freeware Orbs, but can you use them for a mission > critical system? >r > best regards >e > Jakob- >- >- >-   ------------------------------  % Date: Mon, 09 Apr 2001 17:33:49 -0600a$ From: Lee Y T Mah <lytmah@cha.ab.ca>) Subject: Re: DEC Server software question ) Message-ID: <3AD246DD.8ECF2BF5@cha.ab.ca>o  M If you feel the following image would meet your requirements and we could getnO the OK from Compaq or whoever owns the rights to the image, I could email it tot@ you.  The version is V1.3 and it is stable (Uptime of 914 days).      Directory SYS$COMMON:[DECSERVER]  E WWENG2.SYS;2                            2320  19-NOV-1993 15:32:22.00l    M Network Access SW V1.3 for DS900TM  BL80-27  ROM V5.1-0  Uptime: 914 11:29:01   G Address:   08-00-2B-XX-XX-XX   Name:   ABCDEF             Number:     0    Identification:  LAB NE An  C Circuit Timer:            80           Password Limit:            3 B Console Port:              1           Prompt:              Local>C Inactivity Timer:         30           Queue Limit:             100tC Keepalive Timer:          20           Retransmit Limit:         24rC Multicast Timer:          30           Session Limit:            64 C Node Limit:              200           Software:             WWENG2    Service Groups:   0i   Enabled Characteristics:' Announcements,  Broadcast,  Dump,  Lockh   TSM_SVR_DT4NEA> sho port all  0 Port    Access    Status        Services Offered     1     Local     Local Mode   2     Local     Local Mode   3     Local     Connectedi   4     Local     Connectede   5     Local     Connected    6     Local     Connected    7     Local     Idle   8     Local     Local Mode   9     Local     Idle  10     Local     Idle  11     Local     Idle  12     Local     Idle  13     Local     Connectedi  14     Local     Local Mode  15     Local     Idle  16     Remote    Idle  17     Remote    Idle  18     Local     Local Mode  19     Local     Local Mode  20     Local     Idle  21     Local     Idle  22     Local     Idle  23     Local     Idle  24     Local     Idle  25     Remote    Idle  26     Local     Idle  27     Local     Idle  28     Remote    Idle  29     Remote    Idle  30     Remote    Idle  31     Local     Idle  32     Remote    Idle     Patrick Massey wrote:6   > Hello,J > I've recently installed a DS900TM and found that only the first 16 of 32N > ports work.  The software it's loading WWENG2 ver 1.1a makes it think it's aL > DS700.  I really need to get the other 16 ports working.  Does anyone knowM > where I can find an updated loadfile?  I've contacted Compaq and the vendors: > I ordered the part from, neither have been very helpful. > Thanks in advance. >t
 > -Pat Masseys   -- Leer  ; Lee Y T Mah                        Capital Health Authority-? Email: lytmah@cha.ab.ca            Information Systems, RAH CSCE4 Phone:  (780) 477-4725, 477-4233   10240 Kingsway NW? Fax:      (780) 491-5119, 491-5619    Edmonton, AB, CAN  T5H3V9A   ------------------------------  % Date: Mon, 09 Apr 2001 16:44:26 -0400P& From: Patti Johnson <pjohnson@wcu.edu>( Subject: error installing patch on vax??' Message-ID: <3AD21F2A.20C05993@wcu.edu>p  L Hello, I'm new at this and am now trying to apply patches to a couple of ourJ machines.  I'm starting w/ the newer one, a VAX system (7.2).  There's not much to the directions... just  0 	@sys$update:vmsinstal patchname locationofpatch? Sounds easy enough.  But, when I do so, here's the error I get:r  6 %VMSINSTAL-I-RESTORE, Restoring product save set A ...; %BACKUP-E-READERRS, excessive error rate reading <location> + -BACKUP-E-HDRCRC, software header CRC errorl4 %BACKUP-I-SPECIFY, specify option (QUIT or CONTINUE) BACKUP>   L I'm not sure where to turn for answers... I'd assumed something was wrong w/ the saveset I'd downloaded fromtC http://ftp.support.compaq.com/patches/.new/openvms.shtml so I triedrE downloading again but the error recurred.  It's vaxupdate01_072, btw.    tia: pj -- D  M _____________________________________________________________________________aM Patti Johnson (pj)                                           pjohnson@wcu.edu5 WCU Computer Center  Cullowhee, N.C. 287236M 828-227-3224                         Time spent with a child is never wasted.    ------------------------------  # Date: Mon, 09 Apr 2001 21:37:49 GMTL2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman), Subject: Re: error installing patch on vax??2 Message-ID: <NYpA6.712$fB6.18534@news.cpqcorp.net>  P In article <3AD21F2A.20C05993@wcu.edu>, Patti Johnson <pjohnson@wcu.edu> writes: ..1 :	@sys$update:vmsinstal patchname locationofpatch/@ :Sounds easy enough.  But, when I do so, here's the error I get: :u7 :%VMSINSTAL-I-RESTORE, Restoring product save set A ...l< :%BACKUP-E-READERRS, excessive error rate reading <location>, :-BACKUP-E-HDRCRC, software header CRC error5 :%BACKUP-I-SPECIFY, specify option (QUIT or CONTINUE)  :BACKUP> :dF :...I'd assumed something was wrong w/ the saveset I'd downloaded fromD :http://ftp.support.compaq.com/patches/.new/openvms.shtml so I triedF :downloading again but the error recurred.  It's vaxupdate01_072, btw.  #   The file is apparently corrupted.   =   Ensure you are using BINARY transfer mode when downloading.rA   (I prefer to use the ftp area, not web access -- different web  C   browsers have different ideas of which file should be binary and gD   which should be non-binary.)  The FAQ has pointers to the FTP URL.  C   Also ensure that you are NOT shoveling the files through multipled@   systems -- that's certainly possible, but involves great care.1   Easier to simply avoid it, at least to start...   @   When posting, please remember to include the specific commandsE   used -- in this case, the commands used to download and to transfer A   the files to the target OpenVMS VAX system, as well as the full G   command used for the installation.  (I have learned that assumptions oH   that I make about the commands/steps/versions/etc involved can differ G   (widely!) from the assumptions made by others.  No offense intended.)   N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------   Date: 9 Apr 2001 19:48 PDT) From: rankin@eql.caltech.edu (Pat Rankin)-; Subject: Re: Flipping from big-endian to little endian in Cm. Message-ID: <9APR200119482935@eql.caltech.edu>  , In article <3ad1e207$1@newsfeed.vitts.com>,\7  "John Gemignani, Jr." <john@ossc.DELETE.net> writes...3  >> >     outbuf[i+1] = inbuf[i];G >> >                           000C          movb    (r1)[r0],1(r2)[r0]h  >> >     outbuf[i] = inbuf[i+1];G >> >                           0012          movb    1(r1)[r0],(r2)[r0]a > 	 >    Bug.1 >OL >    The result would be two bytes of inbuf[i], as you didn't preserve [i+1] >    before overwriting it.g  >      Thanks for playing, but the judges have just disqualified? you for not reading the code you're commenting about.  There is.> no overwriting taking place; both the source and the generated? instructions illustrate that.  (You'll have to take it on faith @ that the caller isn't passing the same array for both arguments.? I don't know how many times I'm going to have to say that using ? separate input and output buffers lets you avoid copying thingsc? into and back out of a temporary; this is the third so far....)e  =      And regarding an earlier response:  the Fortran compiler > does generate better code for comparable source.  And the fact< that it is allowed to assume that inbuf[] and outbuf[] don't< overlap but the C compiler isn't appears not to matter here.< However, I am using quite old versions of both compilers, so* either of them could have improved by now.  2                 Pat Rankin, rankin@eql.caltech.edu   ------------------------------  % Date: Mon, 09 Apr 2001 18:30:37 -0400s- From: JF Mezei <jfmezei.spamnot@videotron.ca>p Subject: Re: FMS files, Message-ID: <3AD23806.1C1BC917@videotron.ca>   Tim Llewellyn wrote:T > FMS is available on the Condist CD's still. It is no longer supportable, so if you > haveH > the original licence I doubt there will be any upgrade licence issues   J Because FMS is used by some Compaq software packages such as Office ServerL (aka: AllIN1), it is still supported. However, it is a mature product and noJ enhancements are made to it, but I beleive it is still tested for each new release of VMS.    ------------------------------  % Date: Mon, 09 Apr 2001 13:39:22 -0600l% From: Dean Woodward <deanw@rdrop.com>A Subject: From the trenches) Message-ID: <3AD20FEA.2FFA5C1A@rdrop.com>S  A A friend (not one of my customers) on a non computer related listi% sent this- I couldn't ask for better:y  @ > Anyway, somebody asked why I was "stuck" with the VAX/VMS-only> > system at work, and the correct answer is "Because it works.= > Very well. No crashes in the 15 years it's been installed."e   'nuff said.e   ------------------------------   Date: 10 Apr 2001 04:44:21 GMT2 From: mathog@seqaxp.bio.caltech.edu (David Mathog)' Subject: Re: FTP hijacking of VMS sites , Message-ID: <9au335$ot4@gap.cco.caltech.edu>  V In article <9alg6r$bt7$1@hecate.umd.edu>, bleau@umtof.umd.edu (Lawrence Bleau) writes:H >Hello, folks.  I just noticed a lot of oddly named files - directories,N >actually - in my system's anonymous ftp area.  They weren't created by anyoneC >associate with our site.  One associate of mine gave me his guess:  >e@ >i don't know if this is the case here, but i read articles from@ >people who's anonymous ftp areas have been "hijacked" by people@ >distributing large files like movies.  typically, a 1MB file isE >uploaded to check the speed.  if the hijacker thinks the speed is okr> >lots of directories are created and movie parts are uploaded. >r- >see http://www.macintouch.com/ftphijack.htmlp >sN >I think this subject deserves some discussion.  It's the first I heard of it,0 >and I've been managing VMS systems for a while. >e( >Has anyone experienced this phenomenon?  J Yes.  Leave any anonymous FTP site open for uploads and it will eventuallyE be hit.  I had to restrict access on one such site to just the local nG subnets to get rid of these leeches.  Before doing that I had tried to nF control it by sending email to the (many) sites that were originating K accesses.  But it soon became obvious that that was a lost cause.  We could A not restrict upload sizes as the common machine was being used to-F transfer images to a slide imager, and the images were often over 16M.K It's been a year since I tightened up that ftp server and I'm still seeing a% access attempts to those directories.e  $ It wasn't movies though but "WareZ".    >Has is cropped up at VMS sites?  J Probably less than elsewhere since one doesn't normally leave an anonymousK FTP upload open on VMS.  In our case the affected server was running on thenI PC attached to the slide imager.  It never even crossed my mind to allow U) anonymous FTP uploads on the VMS machine!    >  Are% >there standard ways of combating it?a  A Sure, either block uploads on the anonymous account completely ors* block access to it from off site machines.   Regards,   David Mathog mathog@seqaxp.bio.caltech.edun? Manager, sequence analysis facility, biology division, Caltech .J **************************************************************************J *                                RIP VMS                                 *J **************************************************************************   ------------------------------  # Date: Mon, 09 Apr 2001 19:57:36 GMT 0 From: sander@vmsbiz.enet.dec.com (Warren Sander)? Subject: Re: Galaxy Network Interfaces/cards and TCP/IP Sockets02 Message-ID: <QuoA6.698$fB6.18341@news.cpqcorp.net>   I was on vacation....R  + I didn't read news groups for a whole week.0   I feel so refreshed..e     --  B ------------------------------------------------------------------6 Warren Sander                        OpenVMS MarketingD Compaq Computer Corporation          Work:  warren.sander@compaq.comE 200 Forest Street MR01-3/J1          Personal: sander@ma.ultranet.comp3 Marlboro, MA 01752                   (508) 467-4875-6    My opinions are my own and I only speak for myself -          Read http://www.openvms.compaq.com/ pB ------------------------------------------------------------------   ------------------------------  $ Date: Mon, 9 Apr 2001 18:09:39 -04002 From: "John Gemignani, Jr." <john@ossc.DELETE.net>7 Subject: Re: Help needed with RMS Character Field Index + Message-ID: <3ad233eb$1@newsfeed.vitts.com>-  D This sounds like a job for a segmented key.  You can pick and chooseK multiple fields within the record and concatenate them logically using thiseL mechanism.  That would allow you to place the surname first without actually  having to move the field around.  I This all assumes that you pretty much want to maintain the current recordg! format.  If not, then never mind..   -Johnr  , <Gabriel_hogan@hotmail.com> wrote in message$ news:9asbl2$2uc$1@news.netmar.com... >c > Hi,rJ > I have a sequential RMS file of fixed record length. I want to index the fileI > on a character field (third field, length 30, contains surnames). To doP thisL > I need to extract the character field, reverse it and insert the new field at; > the beginning of each record before indexing it with FDL.0 >wL > Has anyone had to do this before and if so, is there a VMS tool to do this or > how did you go about it? >  > thx,	 > Gabrielo >iF >  -----  Posted via NewsOne.Net: Free (anonymous) Usenet News via the	 eb  -----oH >   http://newsone.net/ -- Free reading and anonymous posting to 60,000+ groupsK >    NewsOne.Net prohibits users from posting spam.  If this or other posts-< > made through NewsOne.Net violate posting guidelines, email abuse@newsone.netc   ------------------------------  # Date: Tue, 10 Apr 2001 01:10:01 GMTs From: LBohan@dbc.spam_less..comlJ Subject: Re: How to tell if a system is booting immediately after a crash.8 Message-ID: <ssj4dt0b96frv672ra077mu5ilha6616b5@4ax.com>  2 On Mon, 02 Apr 2001 19:55:53 -0500, Lyndon Bartels  <lbartels@pressenter.com> wrote:  G >I want, during a startup, if the system has gone done as a result of af) >crash or power outage or... whatever....e >pH >Why? So if the system crashes and reboots in the middle of the night, I/ >can tell the computer to page me or something.T >eC >I was thinking of having syshtdown.com write out a data file. ThenpC >during subsequent boot, systartup_vms look for the file, if found, G >delete all versions of the file, and peacefully exit. If not found, so  >my alarm procedures.n
 >Thoughts? >Thanks in advance,S >LyndonO   maybe look at :e    M http://www.openvms.compaq.com:8000/72final/6017/6017pro_071.html#index_x_4059-    inre Crash Log Utility Extractor  6 some of the info you're after, is saved automatically / on recent-enough versions of VMS,  and could bei extracted, and searched ,,,   5 the details differ slightly between Alpha and VAX ...    ------------------------------  $ Date: Mon, 9 Apr 2001 20:57:07 +0100, From: "Richard Maher" <maher_rj@hotmail.c*m>+ Subject: Re: More VMS ads spotted in the UK13 Message-ID: <9at45f$9gg$1@plutonium.btinternet.com>t   Hi,   F I've seen the one in Computer Weekly and I'll have a look at Computing2 tomorrow but what was the third trade rag IT Week?  J If anyone knows who's responsible for these adds please pass on heart feltH thanks! I was speaking to a mate today who contracts at a large UK TelcoG that is looking to replace one of its VMS based billing systems.  He is-G going to make sure that everyone sees these adds and knows that the VMSeJ market is in fact growing and that Compaq isn't afraid to push it. Keep up the good work.  H Given recent share price dips and cost cutting across the board anythingE that makes people feel good about staying with their tried and testedr systems has to be good.    Regards Richard Maher.    0 Alan Greig <a.greig@virgin.net> wrote in message2 news:jg03dtg4f0k97r2okgu49p48um8kpcmunv@4ax.com... >v > They are still coming... > F > In the latest issue of Computing dated 5th April the VMS AlphaserverF > ad appears in one of the most prominent possible positions taking upG > the whole of Page 2. Page 1 is a newspaper style front page so Page 2R6 > is the inside front cover. You can't really miss it. >FF > So that's it now appeared in all three of the biggest circulation UKG > professional computing weekly magazines. Now would be a good time for G > Compaq to take advantage by sending out some updates and links to VMStH > future presentations. Although there isn't supposed to be a connectionE > there always is one. Magazines are more likely to regurgitate presseA > releases as news if the company is also buying ad space for the1H > product. Both the upcoming VMS technical update days in London and the> > DII COE project should be worth of inclusion by any of these > publications.  > -- > Alan   ------------------------------  % Date: Mon, 09 Apr 2001 15:50:45 -0400 7 From: David Beatty <David.Beatty@sasSPAMITUPTHEZZZ.com>m5 Subject: Re: NSLOOKUP: *** Can't initialize resolver. 2 Message-ID: <ERHSOlb5n2Gef5URmY=2mq0L4QfB@4ax.com>  7 Try SHOW CONFIG NAME SERVICE to check your non-volatileL= name service settings.  You could also check your routing via  SHOW ROUTE [/PERM].i  & You should be able to enable BIND  via! SET NAME SERVICE /SYSTEM /ENABLE._   David R. Beattyt  # On Mon, 09 Apr 2001 13:25:38 -0300, * fabio_compaq@ep-bc.petrobras.com.br wrote:  L >I never tested the nslookup before in this machine.  The ping works but the >NSLOOKUP is not ..... >aL >Now I STOPPED the NAME SERVICE and I am having troubles reconfiguring it  . >. . > J >The message of Local Domain *Mismatch* is displayed when I check the NAME	 >SERVICE.1 >  >o< >And how do I restart the BIND Resolver after shutdown it !? >a >Regards >  >FCa >  >d >  > 2 >steven.reece@quintiles.com em 09/04/2001 13:08:21 >a- >Favor responder a steven.reece@quintiles.com  >  >  >a >      Info-VAX@Mvb.Saic.Com >e >2 > 6 >Assunto: Re: NSLOOKUP: *** Can't initialize resolver. >  >e >2 > K >Is your DNS still there and still accessible?  Can you telnet/ping it fromo >your VMS system? L >I'd initially suspect (without any better information) that your VMS systemG >can't get through to the DNS to do a lookup.  This could be because ofeD >configuration problems, transport problems, DNS crashed due to bits+ >flipping in cache or cosmic radiation etc.i >h2 >Has it ever worked?  If yes, then what's changed? >Steve.  >k
 >Fabio asked:nL >>>>Any idea why it is appering this message. My NAME SERVICE is Started and >Enable  >  >TCPIP SERVICE V5.0A ><<< >h >  >n >r >e >u   ------------------------------  $ Date: Mon, 9 Apr 2001 18:14:01 -04002 From: "John Gemignani, Jr." <john@ossc.DELETE.net>5 Subject: Re: NSLOOKUP: *** Can't initialize resolver. + Message-ID: <3ad234f2$1@newsfeed.vitts.com>0  I You may have name service turned on -- meaning that you are running a DNSAE server.  Did you setup the DNS client information correctly, with the0I default domain and such?  You may be running a BIND server but unless youc+ point your DNS to it, you won't talk to it.u   -John   6 <fabio_compaq@ep-bc.petrobras.com.br> wrote in messageF news:OF0B05895F.D0D2519D-ON03256A29.005522AD@ep-bc.petrobras.com.br... > People >eJ > Any idea why it is appering this message. My NAME SERVICE is Started and > Enable >t > TCPIP SERVICE V5.0A  >t	 > Regards  >  > FC >e   ------------------------------   Date: 9 Apr 2001 10:56:20 -0500.+ From: young_r@encompasserve.org (Rob Young)r$ Subject: RE: OT: "The Puzzle Palace"3 Message-ID: <RFmcWWCLHPay@eisner.encompasserve.org>y  z In article <3B55D7F383B0D31197D9009027541CBF0D9D1D2E@cmiexch1.cmi.itds.com>, Christopher Smith <csmith@amdocs.com> writes: >> -----Original Message-----/E >> From: young_r@encompasserve.org [mailto:young_r@encompasserve.org]d > C > 	Hmmmm.  Much of the technical abilities of the NSA are shrouded jA > 	in mystery.  Things of national security and whatnot are still_? > 	very crucial to trying to get a leg up on others set against  > 	our country.p > K > Well said, but somewhat incorrect, IMO.  If it were really "our country," N > we'd know what was going on with it, don't you think?  It's "their country,"# > and they're not willing to share.0 >   = 	No.  In security circles , you work with stuff on a "need to < 	know" basis.  That is why Hansen was such a prize as he was; 	in the breech so to speak.  As citizens, we don't have theCG 	"need to know" such potentially very damaging information.  Trust has tG 	to come into play at some point.  This will quickly spiral off into a  ! 	Libertarian slug-fest, whatever.   ? 	Point is, if there were no secrets some folks not so favorableEC 	to our country would use that to wreck havoc in one form or other.,   	Secrets aren't a bad thing.   				RobI   ------------------------------  % Date: Mon, 09 Apr 2001 14:07:16 -0700h! From: Shane.F.Smith@Healthnet.comw$ Subject: RE: OT: "The Puzzle Palace"D Message-ID: <OF55D9C465.703CD2D9-ON88256A29.0073AF08@foundation.com>  G Secrets aren't a bad thing in themselves. It's when you don't trust thecK people holding the secret that the problems start. You only have to look at C the success of programs like The X Files to get a feel for how much , Americans in general trust their government.  I It's not just Americans, either. I have no respect for politicians of any H flavour. To badly paraphrase Douglas Adams, anybody suficiently vain andD arrogant to campaign to be elected, is the last person who should be allowed to govern.   Shaneo          ? young_r@encompasserve.org (Rob Young) on 04/09/2001 08:56:20 AMt  7 Please respond to young_r@encompasserve.org (Rob Young)u   To:   Info-VAX@Mvb.Saic.Com  cc:h  % Subject:  RE: OT: "The Puzzle Palace"     
 In articleA <3B55D7F383B0D31197D9009027541CBF0D9D1D2E@cmiexch1.cmi.itds.com>,s- Christopher Smith <csmith@amdocs.com> writes:e >> -----Original Message-----fE >> From: young_r@encompasserve.org [mailto:young_r@encompasserve.org]n >.D >    Hmmmm.  Much of the technical abilities of the NSA are shroudedC >    in mystery.  Things of national security and whatnot are stillwA >    very crucial to trying to get a leg up on others set againstl >    our country.- >-K > Well said, but somewhat incorrect, IMO.  If it were really "our country,"aD > we'd know what was going on with it, don't you think?  It's "their	 country,"m# > and they're not willing to share.- >u  A      No.  In security circles , you work with stuff on a "need toD@      know" basis.  That is why Hansen was such a prize as he was?      in the breech so to speak.  As citizens, we don't have the J      "need to know" such potentially very damaging information.  Trust hasJ      to come into play at some point.  This will quickly spiral off into a%      Libertarian slug-fest, whatever.   C      Point is, if there were no secrets some folks not so favorable G      to our country would use that to wreck havoc in one form or other.          Secrets aren't a bad thing.                       Robe   ------------------------------  % Date: Mon, 09 Apr 2001 16:19:34 -0500T+ From: Christopher Smith <csmith@amdocs.com>h$ Subject: RE: OT: "The Puzzle Palace"L Message-ID: <3B55D7F383B0D31197D9009027541CBF0D9D1D37@cmiexch1.cmi.itds.com>   > -----Original Message-----D > From: young_r@encompasserve.org [mailto:young_r@encompasserve.org]  ? > 	No.  In security circles , you work with stuff on a "need toi> > 	know" basis.  That is why Hansen was such a prize as he was= > 	in the breech so to speak.  As citizens, we don't have thec1 > 	"need to know" such potentially very damaging s > information.  Trust has 7 > 	to come into play at some point.  This will quickly v > spiral off into a # > 	Libertarian slug-fest, whatever.   I No offense (really), but this whole paragraph seems to just re-enforce mycG point of view, and doesn't make much sense in your context.  (Maybe I'me reading it wrong.)  A > 	Point is, if there were no secrets some folks not so favorable 7 > 	to our country would use that to wreck havoc in one S > form or other.  % Point taken, but see my next comment.    > 	Secrets aren't a bad thing.  A I agree -- but you know what's said about "too much of anything?"    Regards,   Chrisl  ! Christopher Smith, Perl Developerg Amdocs - Champaign, IL   /usr/bin/perl -e '? print((~"\x95\xc4\xe3"^"Just Another Perl Hacker.")."\x08!\n");1 'i   ------------------------------   Date: 9 Apr 2001 22:52:56 -0500m+ From: young_r@encompasserve.org (Rob Young)l$ Subject: RE: OT: "The Puzzle Palace"3 Message-ID: <dsT6JPZ2BKqg@eisner.encompasserve.org>p  z In article <3B55D7F383B0D31197D9009027541CBF0D9D1D37@cmiexch1.cmi.itds.com>, Christopher Smith <csmith@amdocs.com> writes: >> -----Original Message-----fE >> From: young_r@encompasserve.org [mailto:young_r@encompasserve.org]s > @ >> 	No.  In security circles , you work with stuff on a "need to? >> 	know" basis.  That is why Hansen was such a prize as he wase> >> 	in the breech so to speak.  As citizens, we don't have the2 >> 	"need to know" such potentially very damaging  >> information.  Trust has  8 >> 	to come into play at some point.  This will quickly  >> spiral off into a P$ >> 	Libertarian slug-fest, whatever. > K > No offense (really), but this whole paragraph seems to just re-enforce myaI > point of view, and doesn't make much sense in your context.  (Maybe I'ml > reading it wrong.) >   A 	Perhaps not.  But what I was answering "No" to has been trimmed.r
 	You said:   > K > Well said, but somewhat incorrect, IMO.  If it were really "our country,"eN > we'd know what was going on with it, don't you think?  It's "their country,"# > and they're not willing to share.c >   B   	I disagree with that principle totally.  It is our country, but? 	we have no need to know what was going on with it to the level : 	of knowing what our secret agencies are doing.  Yes it isC 	important to know what our Congress is up to.  President.  Senate. G 	Police and other public officials.  Our secret agencies can't functionr? 	in the light of day and I contend a reasonable person would beH: 	right there with me.  However, the Libertarians and other, 	pseudo-anarchists believe things like this:  ( http://www.libertarian.org/libphilo.html  G What is the proper role of government in a free society? To answer this.A question, we must first understand what is meant by "government."o  P "Government is the use of force. To govern means to control. The use of force isI implicit in the definition of control. Otherwise, it would be "influence" M rather than control. Even the good things that governments do involve the useeH of force somewhere, somehow. Sometimes government uses force directly toK control behavior. Other times, government uses money taken by force to fundo@ activities which would otherwise not involve the use of force. "  > 	If you read further, you see Government is equated to thieves
 	and robbers.o  C 	But this view of Government runs contrary to what is in Scripture:C  F Titus 3:1    Remind the people to be subject to rulers and authoritiesE Romans 13:1  The authorities that exist have been established by God.a  A 	But what about the Revolution?  Much fear and trepidation there,cH 	sure.  Wisdom wasn't thrown out and new order was established.  But to G 	go far to the Libertarian side and suggest that Government is bad and lF 	therefore harboring secrets is bad isn't a proper view in my opinion.   				Rob*   ------------------------------  % Date: Tue, 10 Apr 2001 00:35:05 -0400S' From: "Bill Todd" <billtodd@foo.mv.com>*$ Subject: Re: OT: "The Puzzle Palace"( Message-ID: <9au2hc$brn$1@pyrite.mv.net>  8 "Rob Young" <young_r@encompasserve.org> wrote in message- news:dsT6JPZ2BKqg@eisner.encompasserve.org...n > In articleA <3B55D7F383B0D31197D9009027541CBF0D9D1D37@cmiexch1.cmi.itds.com>,D- Christopher Smith <csmith@amdocs.com> writes:a   ...c  C > > Well said, but somewhat incorrect, IMO.  If it were really "our 	 country,"eF > > we'd know what was going on with it, don't you think?  It's "their	 country,"r% > > and they're not willing to share.S > >e >gC >   I disagree with that principle totally.  It is our country, butm@ > we have no need to know what was going on with it to the level0 > of knowing what our secret agencies are doing.  G Well, that really depends on exactly what they're doing, doesn't it?  I D mean, back in the Iran-Contra days, one could suggest that they wereJ actively subverting our Constitutionally-defined process of government, inL active collusion with one of its major branches - and that therefore knowing/ what they were up to was very important indeed.c  H Or take explicitly-prohibited domestic activities by the CIA, as anotherD example from a bit farther back.  Or the surveillance activities andL dirt-collecting by the FBI on anyone who didn't meet with Hoover's approval.K The main problem with secret agencies (and even supposedly more accountable G agencies when they *act* secretly, as in the FBI's case) is the lack ofaK normal checks on their activities, and the success record of the substituteGD checks has not been good enough to make people entirely comfortable.     Yes it is D > important to know what our Congress is up to.  President.  Senate.H > Police and other public officials.  Our secret agencies can't function > in the light of dayh  I That's only true *outside* the U.S.  Within the U.S., there's arguably noeL necessity for secrecy beyond that associated with a developing investigationK of some sort, and possibly *never* any necessity for secrecy after the fact G save to protect individuals from possible retribution (i.e., it's nevervL appropriate just to keep dubious official activities from the light of day -J the light of day being the best assurance that official activities *won't* become dubious).  H Secret government functioning *within* a free society is damn close to aL contradiction in terms.  And while secrecy in the international arena may beH a regrettable necessity, there should always be the knowledge that afterD some decent interval those activities will be opened up to encourage+ appropriate restraint in their prosecution.f  +  and I contend a reasonable person would beA; > right there with me.  However, the Libertarians and otherc- > pseudo-anarchists believe things like this:  >t* > http://www.libertarian.org/libphilo.html >oI > What is the proper role of government in a free society? To answer thisoC > question, we must first understand what is meant by "government."r >mI > "Government is the use of force. To govern means to control. The use of. force isK > implicit in the definition of control. Otherwise, it would be "influence"lK > rather than control. Even the good things that governments do involve thec usesJ > of force somewhere, somehow. Sometimes government uses force directly toH > control behavior. Other times, government uses money taken by force to fundB > activities which would otherwise not involve the use of force. " >e? > If you read further, you see Government is equated to thievesi > and robbers. >aD > But this view of Government runs contrary to what is in Scripture: >-H > Titus 3:1    Remind the people to be subject to rulers and authoritiesG > Romans 13:1  The authorities that exist have been established by God.e  K Personally, I've got about as little use for the Libertarian line as I have'J for the biblical one.  But if it comes to a choice between 'being subject'K to a government or considering it evil, I'll choose the latter - especiallyiI if anyone tries to throw any 'established by God' crap into the equation.a  H The strength - and legitimacy - of our society is that the government isH subject to the people, not the reverse.  Forget that at your peril (and, unfortunately, mine too).t   - bill   >aB > But what about the Revolution?  Much fear and trepidation there,H > sure.  Wisdom wasn't thrown out and new order was established.  But toG > go far to the Libertarian side and suggest that Government is bad andeG > therefore harboring secrets is bad isn't a proper view in my opinion.  >  > Robe ><   ------------------------------  $ Date: Mon, 9 Apr 2001 17:27:34 -04008 From: "Scott LePage" <Scott.Lepage@ihatespam.compaq.com>+ Subject: Re: Problems with CSWS and CGI-BINt2 Message-ID: <IOpA6.710$fB6.18563@news.cpqcorp.net>  H I have the same example from the CGI package working fine.  As a test, I copiedK your script to my server and that worked as well.  So the good news is, youi don't H have any typos in the script.  My guess is that there is something wrong	 with youri) configuration of the directory "cgi-bin".b  I Out of the box, cgi-bin is setup to do DCL scripts and executables.  Perld	 won't runM, from there.  I assume you have changed that.  5 Here's my configuration for my perl script directory:g  %     Alias /perl/ "/apache$root/perl/"s       <Location /perl>         SetHandler perl-script$         PerlHandler Apache::Registry         Options ExecCGIt         Allow from all         PerlSendHeader Onb     </Location>i   Hope this helps, Scottc    H <j.t.horn@usa.net> wrote in message news:9aiafj$ug1$1@news.netmar.com... >1G > I posted this before but never heard anything back so I am reposting.m >uF > When it try to execute the following cgi code, it displays the form,	 allows memK > to fill out the form, but it never does the displaying out the results tol thea > form. Any one have any clues:l >c > Apache Information: G > Server Version: Apache/1.3.12 (OpenVMS) mod_perl/1.21 Mod_JServ/1.1.1S > mod_ssl/2.6.2 OpenSSL/0.9.5a$ > Server Built: Dec 12 2000 19:25:20 >:	 > Script:  >      use CGI qw/:standard/;m >       print header,t- >             start_html('A Simple Example'), % >             h1('A Simple Example'),. >             start_form,N7 >             "What's your name? ",textfield('name'),p, + >             "What's the combination?", p,t, >             checkbox_group(-name=>'words',G >                            -values=>['eenie','meenie','minie','moe'],?> >                            -defaults=>['eenie','minie']), p,- >             "What's your favorite color? ",d( >             popup_menu(-name=>'color',I >                        -values=>['red','green','blue','chartreuse']),p,a >             submit,. >             end_form,r >             hr;d >i >        if (param()) {e6 >            print "Your name is",em(param('name')),p,H >                  "The keywords are: ",em(join(", ",param('words'))),p,@ >                  "Your favorite color is ",em(param('color')), >                  hr; > }t >o
 > Results:* > Just goes back to the start of the form.% http://nell.shsu.edu/cgi-bin/test.cgiw > to see what happens. > 7 > Thank you for any kind of response anyone can supply.i >sF >  -----  Posted via NewsOne.Net: Free (anonymous) Usenet News via the	 eb  -----EH >   http://newsone.net/ -- Free reading and anonymous posting to 60,000+ groupsK >    NewsOne.Net prohibits users from posting spam.  If this or other postse< > made through NewsOne.Net violate posting guidelines, email abuse@newsone.net$   ------------------------------  % Date: Mon, 09 Apr 2001 20:34:48 +0100a+ From: "antonio.carlini" <arcarlini@iee.org>t< Subject: Re: Q: Converting a VAX 6000-420 to a VAX 6000-620?' Message-ID: <3AD20ED8.2AD9E3B9@iee.org>    Paul Repacholi wrote:o > / > "antonio.carlini" <arcarlini@iee.org> writes:  > H > > So I think the short answer is: T2019 and that widget you plug in atG > > the back (sorry, no part number ... I can hit the IPBs if you thinkd( > > you might have it but are not sure). > 3 > Two blue power buses that go to the bottom slots?5 > C > 17-002751-01 abd another cumming hidden under the main power bus.a  3 I cannot find a number for the powerbus cable and Ie2 the manual does not say anything about playng with4 any other cables. (I'll read more carefully if Brian% actually gets all the bits together!)7  7 I stumbled across a step-by-step description of exactlyS? how to apply the upgrade in the VAX 6000-600 cpu install guide. : The part number for the complete kit is H9657-CX (H9657-CU4 is the full power upgrade kit that includes swapping1 out PSUs). I seriously doubt that either of thesen+ are still orderable ... but you never know._  ( If you have a H7242 regulator it already* supplies 3.3V so no upgrade is needed. You0 may need to diddle with the +3.3V-inhibit cable.# (You never know, you may be lucky).k   Antonios   -- o   ---------------u- Antonio Carlini             arcarlini@iee.orgo   ------------------------------  # Date: Tue, 10 Apr 2001 00:00:34 GMTf( From: Terry Kennedy <terry@gate.tmk.com>< Subject: Re: Q: Converting a VAX 6000-420 to a VAX 6000-620?' Message-ID: <GBJuoy.J1L@spcuna.spc.edu>e  9 In comp.os.vms antonio.carlini <arcarlini@iee.org> wrote:t9 > I stumbled across a step-by-step description of exactlyyA > how to apply the upgrade in the VAX 6000-600 cpu install guide.i< > The part number for the complete kit is H9657-CX (H9657-CU6 > is the full power upgrade kit that includes swapping3 > out PSUs). I seriously doubt that either of these - > are still orderable ... but you never know.0  H   H9657-CU was cancelled on 8-Nov-1999. But you wouldn't want to pay theF $15,345.00 list price plus $4,329.00 mandatory installation anyway 8-( FWIW, the part description is:  @ "VAX 6000 Power and Packaging Upgrade to complete H9657-CA level  o XMI-2 Card Cage  o 3.3 Volt Power Regulator.  o XBIA+ Module   o XTC Module"  * > If you have a H7242 regulator it already, > supplies 3.3V so no upgrade is needed. You2 > may need to diddle with the +3.3V-inhibit cable.% > (You never know, you may be lucky).n  4   H7242 is still available, $413.00, 1 day lead time#   The XBIA+ (DWMBB-AB) is cancelledpI   I think the XTC is 20-29176-01, claims to be available, $150.00, but no ?   lead time (which usually means it's been "stealth cancelled")2  4         Terry Kennedy             http://www.tmk.com5         terry@tmk.com             Jersey City, NJ USAc   ------------------------------  $ Date: Mon, 9 Apr 2001 16:44:17 +0100* From: "Richard Brodie" <R.Brodie@rl.ac.uk>3 Subject: Re: Running a Diagnostic on a Network Cardt+ Message-ID: <9asld3$po6@newton.cc.rl.ac.uk>:  N <sfm1115@bjc.org> wrote in message news:3ad1c86e.512320277@news.starnet.net...5 > I am running an AlphaServer 2100 with OpenVMS 7.2-1  > F > Is there an inherited tool inside of the OpenVMS software that I canE > use to diagnose the netword card in this server.  I am not seeing acH > device errors on the system but I need to make sure.  The server keeps > dropping users on and off.  B Typically network errors are not logged on the device but there isA comprehensive built in diagnostics. LANCP is probably easiest fori8 quick, low-level diagnostics. MC LANCP SHOW DEVICE/COUNTC would be a good thing to check first. Most higher level layers willh- also have per protocol diagnostics somewhere.    ------------------------------  # Date: Mon, 09 Apr 2001 19:33:35 GMTs2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)< Subject: Re: Solaris tmpfs (was Re: VMS-Related: Affordable)2 Message-ID: <j8oA6.696$fB6.18336@news.cpqcorp.net>  ] In article <3AD1F484.9C8E4598@uk.sun.com>, andrew harrison <andrew.nospam@uk.sun.com> writes:   ; :Sort of. Solaris has an in-memory filesystem called tmpfs.  :-; :/tmp is normally mounted as a tmpfs volume but some people39 :also use it for things like Sybase TMPDB. It never gets j4 :flushed to disk unless you run out of memory and it :gets paged.  F   A good idea, but certainly not the implementation that I would have J   chosen.  I'd have gone the route of a host-based RAM disk with pageable G   memory, and the same file system everything else uses.  (Why?  A diskbG   device is a whole lot less involved than a file systems.  I'd prefer  0   to avoid maintaining yet another file system.)  N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  $ Date: Mon, 9 Apr 2001 18:04:01 -0400' From: "Bill Todd" <billtodd@foo.mv.com> < Subject: Re: Solaris tmpfs (was Re: VMS-Related: Affordable)( Message-ID: <9atbk9$o05$1@pyrite.mv.net>  = Hoff Hoffman <hoffman@xdelta.zko.dec.nospam> wrote in messageu, news:j8oA6.696$fB6.18336@news.cpqcorp.net...< > In article <3AD1F484.9C8E4598@uk.sun.com>, andrew harrison" <andrew.nospam@uk.sun.com> writes: >r= > :Sort of. Solaris has an in-memory filesystem called tmpfs.0 > :)= > :/tmp is normally mounted as a tmpfs volume but some people:: > :also use it for things like Sybase TMPDB. It never gets6 > :flushed to disk unless you run out of memory and it > :gets paged. >mG >   A good idea, but certainly not the implementation that I would haveiK >   chosen.  I'd have gone the route of a host-based RAM disk with pageablerI >   memory, and the same file system everything else uses.  (Why?  A diskcH >   device is a whole lot less involved than a file systems.  I'd prefer2 >   to avoid maintaining yet another file system.)  G While my acquaintance with the situation is less than detailed, I think L tmpfs may *be* just another instance of an otherwise normal file system thatK happens to exist on pageable memory formated like a disk - which if true is L pretty much the same as the approach you describe.  But remember that in any@ event Unix is used to dealing with a handful or more of somewhatJ strange-looking 'filesystems' (devfs, procfs, probably others), so even ifJ tmpfs is somewhat different that would be less noticeable than it would be in a VMS environment.u   - bill   > ( >  ---------------------------- #include' <rtfaq.h> -----------------------------kL >       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com, >  --------------------------- pure personal# opinion ---------------------------t1 >    Hoff (Stephen) Hoffman   OpenVMS Engineering  hoffman#xdelta.zko.dec.com >e   ------------------------------  % Date: Mon, 09 Apr 2001 18:38:33 +0100t0 From: andrew harrison <andrew.nospam@uk.sun.com>% Subject: Re: Spice for OpenVMS Alpha?a* Message-ID: <3AD1F399.2E84375F@uk.sun.com>   David Mathog wrote:I > _ > In article <3AC996E4.57EC594D@uk.sun.com>, andrew harrison <andrew.nospam@uk.sun.com> writes:p > > > >2.     The Cert advisories prove beyond doubt that security> > >       information released by Compaq relating to OpenVMS's@ > >       vunerability to common exploits is either non existant< > >       when OpenVMS was vunerable, incorrect when OpenVMSB > >       was vunerable, or unhelpfull when OpenVMS was vunerable. > E > You'd better cite letter and verse when you make a claim like that. G > On the face of it what you say is totally ridiculous, if for no otherrM > reason than that most "common exploits" depend upon buffer overruns in UnixaJ > utilities that are wholly absent from VMS.  Conversely, there seem to beI > quite a few security holes in Solaris specific software.  For instance:h >    Well here is four.   1.	Incorrect and later< The OpenVMS response for POD was origionally wrong. It cited= no vunerability. This was incorrect and was replaced 4 monthse> after the origional violations with a reference to the patches9 for UCX that need to be applied to avoid being vunerable.   * 2.	Non existant when OpenVMS is vunerable.: There is no OpenVMS/UCX response for Land/Teardrop in the 6 Cert advisory, however the Wizard says that OpenVMS is2 vunerable and lists the  patches required for UCX.  
 3.	Unhelpfull 5 The OpenVMS response to Smurf suggests hiding OpenVMS-5 boxes behind a firewall and taking internal measures 53 against internal violations. Other vendors offered  2 configuration changes to their IP stack which did 0 not require you to go out and buy a Firewall or - have to attempt to maintain security through J social engineering.a   4.	Late 1 There is no response to the Naptha violation for 02 OpenVMS, Compaq is investigating the vunerability 1 of all its OS's to this exploit. So far they havei0 posted a Tru64 fix but nothing for OpenVMS. The 6 advisory was origionally posted in November 30th 2000.   > >n< > >       This may be for any one of a number of reasons but= > >       which would you prefer ? Incompetence, arrogance ort@ > >       a conspiracy, how do you explain the fact that Compaqs: > >       responses for OpenVMS leave a lot to be desired. > G > You have not made your point.  To assert that they "leave a lot to beoL > desired" without citing specific cases, and a lot of them at that, is just5 > slinging mud and hoping that some of it will stick.d >    Ok. I will let you decide.  ; All the other major systems vendors offer either patches ore
 configurationh; changes to their IP stacks to prevent a violation. (Smurf).l  B Compaq suggest that you stick your OpenVMS boxes behind a FirewallC presumably not running on OpenVMS, they also suggest that internal l violations should be punished.  @ How would you describe Compaqs response in the light of everyoneC elses. Describing Compaqs response as "Leaving a lot to be desired"c is letting them off lightly.  )M > If you think that OpenVMS is so insecure I'm sure that we can arrange for atJ > regular user account for you on a VMS system so that you can demonstrateK > for us just how easy it is to break in.  We'll even give you a head starteM > by allowing a telnet and providing a valid login.  The worst you'll be able'F > to do is to become a PITA by running something interactively at fullM > priority - until the system manager drops your default priorities, and theno$ > you won't even be able to do that. >   D This is getting a bit repetitive. I don't think OpenVMS is insecure,@ I do however think that it is no where near as secure as OpenVMS) advocates would like everyone to beleive.s   Regardsc Andrew Harrison  Enterprise IT Architect    ------------------------------  % Date: Mon, 09 Apr 2001 19:11:27 +0100b0 From: andrew harrison <andrew.nospam@uk.sun.com>% Subject: Re: Spice for OpenVMS Alpha?a* Message-ID: <3AD1FB4F.77DE3231@uk.sun.com>   Bob Koehler wrote: > _ > In article <3ACAFF52.7F5A96C1@uk.sun.com>, andrew harrison <andrew.nospam@uk.sun.com> writes:t > > Mark Berryman wrote: > >u > >>  I had responsiblity J > >> for literally hundreds of VMS systems running IP and everyone of themK > >> tested immune and did so at the time the attack became known, not somea > >> undefineable time later.h > >> > > 2 > > I am sorry but this simply cannot be tha case. > I > The facts don't match the preconceived notions so the messenger must ben > at fault?s >   3 Exactly though I think you meant Marks facts don't m0 match my preconcieved notions. When in fact its + the other way around. The OpenVMS received e/ wisdom on security unfortunately does not matchu
 the facts.   Regardso Andrew Harrisono Enterprise IT Architects   ------------------------------  % Date: Mon, 09 Apr 2001 18:42:28 +0100 0 From: andrew harrison <andrew.nospam@uk.sun.com>$ Subject: Re: VMS-Related: Affordable* Message-ID: <3AD1F484.9C8E4598@uk.sun.com>   Bill Todd wrote: > 9 > Paul Repacholi <prep@prep.synonet.com> wrote in messagen) > news:87puep7wur.fsf@prep.synonet.com...a- > > "Bill Todd" <billtodd@foo.mv.com> writes:o >  > ...  > J > > > Incorrect.  It might be the *next-to-last* thing you want to do, butD > > > the *last* thing you want to do is create a need for more diskI > > > accesses, because the relative speed of disk access has lagged evenn@ > > > farther behind than the relative speed of memory accesses. > >tF > > If we restrict it to sequential read/write, then there is no extra > > disk access. > I > Even that's not necessarily true.  For example, one of the most visibleiK > performance wins a Unix development environment has is that its temporary M > files may never even touch the disk, since they're deleted before the cachen > needs to flush their data. >   : Sort of. Solaris has an in-memory filesystem called tmpfs.  : /tmp is normally mounted as a tmpfs volume but some people8 also use it for things like Sybase TMPDB. It never gets 3 flushed to disk unless you run out of memory and it  gets paged.r   Regardsi Andrew Harrisonw Enterprise IT Architecto   ------------------------------  $ Date: Mon, 9 Apr 2001 18:09:31 -0400' From: "Bill Todd" <billtodd@foo.mv.com>e$ Subject: Re: VMS-Related: Affordable( Message-ID: <9atbuj$p3o$1@pyrite.mv.net>  ; andrew harrison <andrew.nospam@uk.sun.com> wrote in messagem$ news:3AD1F484.9C8E4598@uk.sun.com... > Bill Todd wrote: > >o; > > Paul Repacholi <prep@prep.synonet.com> wrote in messagen+ > > news:87puep7wur.fsf@prep.synonet.com...c/ > > > "Bill Todd" <billtodd@foo.mv.com> writes:  > >y > > ...f > >tL > > > > Incorrect.  It might be the *next-to-last* thing you want to do, butF > > > > the *last* thing you want to do is create a need for more diskK > > > > accesses, because the relative speed of disk access has lagged even B > > > > farther behind than the relative speed of memory accesses. > > >eH > > > If we restrict it to sequential read/write, then there is no extra > > > disk access. > > K > > Even that's not necessarily true.  For example, one of the most visiblecC > > performance wins a Unix development environment has is that itsn	 temporary.I > > files may never even touch the disk, since they're deleted before theg cache  > > needs to flush their data. > >g >i< > Sort of. Solaris has an in-memory filesystem called tmpfs. > < > /tmp is normally mounted as a tmpfs volume but some people9 > also use it for things like Sybase TMPDB. It never gets 5 > flushed to disk unless you run out of memory and it 
 > gets paged.S  F Actually, I was thinking more along the lines of SGI's XFS, which IIRCH provides the kind of optimization I was describing as part of its normalG activity, without requiring use of something like tmpfs.  I think BSD'saD soft-update mechanism was designed to do the same kind of thing, butJ couldn't say whether it made it into any existing product-quality release.I And from some of David Mathog's posts, I would have guessed that whateveruL system he's been measuring to highlight the deficiencies of VMS in this area supports this.   - bill   > 	 > Regardsa > Andrew Harrisono > Enterprise IT Architectu   ------------------------------  $ Date: Mon, 9 Apr 2001 18:11:14 -0400' From: "Bill Todd" <billtodd@foo.mv.com>u$ Subject: Re: VMS-Related: Affordable( Message-ID: <9atc1s$p3t$1@pyrite.mv.net>  7 Paul Repacholi <prep@prep.synonet.com> wrote in message ' news:87bsq6a1iq.fsf@prep.synonet.com...mK > Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de> writes:e >aB > > Does any system offer double-mapping file blocks into both theH > > process's buffer and the system filesytem cache? Accounting would be- > > complicated, but it should be possible...  > < > Some unix systems go this way I believe for mmap'ed files.  G I was told just today that Linux does this, though I can't vouch for itn myself.t   - bill   >nD > One of the side effects of the unix kernel buffers, is that the IOA > system has only ever had to deal with kernel addresses. On someNH > machines, a context switch nukes the user section of the address spaceB > (sound like a machine or two we know :) ), so the file is doubleA > mapped to some set of kernel addresses and the IO done to thoserF > addresses, and the 'bottom' of the kernel wories about fixing up the > user mode mappings.o >w > --> > Paul Repacholi                               1 Crescent Rd.,9 > +61 (08) 9257-1001                           Kalamunda.lB >                                              West Australia 60760 > Raw, Cooked or Well-done, it's all half baked.   ------------------------------  $ Date: Mon, 9 Apr 2001 18:21:32 -0400' From: "Bill Todd" <billtodd@foo.mv.com>e$ Subject: Re: VMS-Related: Affordable( Message-ID: <9atcl3$pcr$1@pyrite.mv.net>  J Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de> wrote inJ message news:y4hezy7ble.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de...+ > "Bill Todd" <billtodd@foo.mv.com> writes:c >SL > > No matter what XFC does, there's likely to be some minor performance hit@ > > compared with the Unix approach simply because in making its enhancementsK > > transparent (which I agree is generally a Good Thing) it will result inr twolK > > data copies (once to/from a process's internal RMS buffer, and a seconda time@ > > to/from the XFC's lower-level cache) rather than Unix's one. > C > I'm not sure I understand this completely. _Somebody_ must do thesD > blocking/deblocking from the application's byte stream to the disk > blocks  F Unless, of course, the application can handle the bytes in situ in theG blocks - which is certainly what I'd hope something like high-bandwidth6H streaming video would try to do:  copying such data even once can have a& noticeable impact on system bandwidth.  <  - I guess that somebody isn't the file system, so it is RMSF > or the C RTL. Then these blocks must go to disk - traditionally, VMSE > does direct I/O without a system buffer, instead of Unix's approach L > to first copy the data to a system buffer and then, if required/requested,C > to disk. So I'd say the usual VMS approach saves one memory copy,e- > at the cost of possibly unwanted disk I/Os.9  L No, since the traditional Unix approach (if I understand it correctly) movesJ application bytes directly to/from a system buffer, which is moved to/fromL disk - still only one copy operation.  This would truly suck if the overheadJ of a system call were what it used to be when a full-fledged address-spaceF context-switch was required, but in today's systems it's no worse (andB possibly better) than the contortions RMS goes through just to get2 application bytes into the (P1 space?) RMS buffer.   - bill   >1J > Does any system offer double-mapping file blocks into both the process'sI > buffer and the system filesytem cache? Accounting would be complicated,u > but it should be possible... >l > Jan    ------------------------------  $ Date: Sat, 7 Apr 2001 12:49:37 -0400# From: "Ram Rajadhyaksha" <dev@null> ' Subject: [Q] Adding Memory to a MV 3100  Message-ID: <3acf4532$1@news>   I I have added 64 Mb to our MV 3100/95. The SHOW MEMORY command at the boot K prompt shows (correctly) 128 Mb total. However, OpenVMS 6.2 still sees onlyb 64 Mb.  F It appears that I need to run the AUTOGEN command to update the systemK configuration for the new memory. This seems scary, especially if you don'ti" know what you are doing. (I don't)  I Is there a simpler way to (ie, changing some conf files) reset the memory G size and swap file sizes? I am concerned that AUTOGEN will change other  things as well when it runs.   Thanks,    -- Ram Rajadhyaksha DLZ Corporationc www.dlzcorp.como   ------------------------------  % Date: Tue, 10 Apr 2001 00:05:50 -0400w- From: JF Mezei <jfmezei.spamnot@videotron.ca> + Subject: Re: [Q] Adding Memory to a MV 3100g+ Message-ID: <3AD28697.D512FD7@videotron.ca>l   Ram Rajadhyaksha wrote:  > K > I have added 64 Mb to our MV 3100/95. The SHOW MEMORY command at the bootfM > prompt shows (correctly) 128 Mb total. However, OpenVMS 6.2 still sees onlyM > 64 Mb.   try:
 $MC SYSGEN SYSGEN> SHOW PHYSICALPAGES  H This will tell you how many physical pages VMS really thinks there are.  you could try  USE CURRENT  SET PHYSICALPAGES = new number
 WRITE CURRENTf  4 and then reboot to see if VMS sees your new memory.   M But eventually, you will want to run AUTOGEN to ensure that it adjusts plenty 9 of other parameters  to take advantage of the new memory.   N You can run AUTOGEN in such a way that it doesn't modify any system parametersO and just plays with its own stuff. (@SYSUPDATE:AUTOGEN HELP will give you info)g   ------------------------------  % Date: Tue, 10 Apr 2001 00:49:18 -0400d2 From: rdeininger@mindspring.com (Robert Deininger)+ Subject: Re: [Q] Adding Memory to a MV 3100iL Message-ID: <rdeininger-1004010049190001@user-2iveat9.dialup.mindspring.com>  B In article <3acf4532$1@news>, "Ram Rajadhyaksha" <dev@null> wrote:  K > I have added 64 Mb to our MV 3100/95. The SHOW MEMORY command at the boot M > prompt shows (correctly) 128 Mb total. However, OpenVMS 6.2 still sees onlyM > 64 Mb. > H > It appears that I need to run the AUTOGEN command to update the systemM > configuration for the new memory. This seems scary, especially if you don't:$ > know what you are doing. (I don't)  J The right thing to do is autogen.  See the "VMS System Mangement UtilitiesF Reference Manual", which has a chapter on autogen.  (That's approx theE right name around VMS 7.1 or so; I skipped 6.2 so the titles may be a  little different.)  J The first parameter to autogen is the "starting phase", and the second theH "end phase".   If you start with the SAVPARAMS or GETDATA phase, and endH with the GENPARAMS phase, autogen won't actually change your parameters,G but you can see what it would have changed if you'd gone all the way to I SETPARAMS.  (If this is unclear, either I'm too tired to write clearly or & you need to read the manual, or both.)  J In my most recent experience on a VAX (2 hours ago, after many years of noH memory upgrades), autogen computes PHYSICALPAGES, but it doesn't show upC in the autogen report.  You'll be more interested in verifying thatrC autogen didn't mess anything else up too badly.  The danger is thataH someone else may have changed parameters "by hand" (see below), and your@ MODPARAMS.DAT is not in sync with what your system really needs.  I My memory told me that, with no PHYSICALPAGES in MODPARAMS.DAT, VMS woulduG set the parameter to a value (0?) that means "use all the memory that'saB present".  But that isn't the way it worked out tonight.  What youF actually get if you don't specify PHYSICALPAGES is "use all the memory, that was present the last time autogen ran".  K > Is there a simpler way to (ie, changing some conf files) reset the memoryrI > size and swap file sizes? I am concerned that AUTOGEN will change othert > things as well when it runs.  I You can use SYSMAN (at least on newer versions) and it's PARAM command to-A modify parameters directly in the disk file without going through J autogen.  But if you don't change MODPARAMS.DAT as well, your changes willJ probably be undone the next time autogen is used.  Also, autogen has a lotG of rules that keep it from picking really bad parameter values.  If you7I don't understand what you're doing, by-hand changes may leave your system@ unbootable.a   -- a Robert Deininger rdeininger@mindspring.com    ------------------------------   End of INFO-VAX 2001.200 ************************