1 INFO-VAX	Tue, 31 Oct 2000	Volume 2000 : Issue 608       Contents: Re: Broadcasting and UCX6 Re: Cheap Alpha - What config is good and WHAT PRICE ?6 Re: Cheap Alpha - What config is good and WHAT PRICE ? Re: DECdtm QUESTIONS DECforms and ACMS  Re: DECforms and ACMS $ Re: Decwindows TPU/EVE: Window title$ Re: Decwindows TPU/EVE: Window title Re: eBay (guess what) again. Re: eBay (guess what) again. Re: eBay (guess what) again.- Example of using psuedoterm device from DCL ? 1 Re: Example of using psuedoterm device from DCL ? ! File encryption software for VMS? % Re: File encryption software for VMS? % Re: File encryption software for VMS? % Re: File encryption software for VMS? % Re: File encryption software for VMS? % Re: File encryption software for VMS? ) How to remove trailing FF from a printjob - Re: How to remove trailing FF from a printjob - Re: How to remove trailing FF from a printjob - Re: How to remove trailing FF from a printjob < I know the answer but just can't remember what...(Pagefiles)@ Re: I know the answer but just can't remember what...(Pagefiles)@ Re: I know the answer but just can't remember what...(Pagefiles) Re: MOZILLA M18 crashes ! Re: Nice job - porting VNC to VMS  Number of users exceeded Re: Number of users exceeded Re: Number of users exceeded Re: Number of users exceeded= Re: Oracle/VMS performance problems with Multithreaded Server 4 Questions about SHOW MEM/SLOTS and swapped processes8 Re: Questions about SHOW MEM/SLOTS and swapped processes Re: Rdb on VMS reference?  Re: Sun "uptime" belly laugher Re: Sun "uptime" belly laugher RE: Sun "uptime" belly laugher Re: Sun "uptime" belly laugher Re: Sun "uptime" belly laugher Re: Sun "uptime" belly laugher Re: Sun "uptime" belly laugher Re: Sun "uptime" belly laugher Re: Sun Moon Re: Sun Moon Re: Sun Moon' Sun's haunting ecache problems detailed  TCPIP V5.0A SMTP ACCVIO  Re: TEK 4111 emulator?	 TLZ tapes # Tough times at the Sun homestead... ' Re: Tough times at the Sun homestead... ' Re: Tough times at the Sun homestead... # VAX Mail question - corrupted files ' Re: VAX Mail question - corrupted files ' Re: VAX Mail question - corrupted files ' Re: VAX Mail question - corrupted files # VAX Mail question - corrupted files ' Re: VAX Mail question - corrupted files ' Re: VAX Mail question - corrupted files ' Re: VAX Mail question - corrupted files ' Re: VAX Mail question - corrupted files ' Re: VAX Mail question - corrupted files  VAXstation 4000 Model 60 manual * Why missing [000000] after volume restore?  F ----------------------------------------------------------------------  # Date: Mon, 30 Oct 2000 21:19:12 GMT  From: mindphaser23@my-deja.com! Subject: Re: Broadcasting and UCX ) Message-ID: <8tkok6$7vm$1@nnrp1.deja.com>   E I must be missing something, somewhere. Could be very simple. Problem C is documentation is scarce. I will definitely post my findings if I 7 find a solution. Suggestions are still appreciated. :-)   A > >> Have you tried a $ UCX SET [CONF] INT ZE0/BR=255.255.255.255 3 > >> after running the config procedure initially ?  > > D > >Thanks for the reply. I tried the above command, but I receive an  > >invalid broadcast mask error. > C > And now, everone starts to ask, why such a command is implemented  then, D > if the system decides itself what value to use (from addr + mask). > Silly, isn't it ?  >  > --> > Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651= > Network and OpenVMS system manager  Fax.    +43 1 81111-888 > > <<< KAPSCH AG  Wagenseilgasse 1     E-mail  eplan@kapsch.netA > A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a  realist" >     & Sent via Deja.com http://www.deja.com/ Before you buy.    ------------------------------  % Date: Mon, 30 Oct 2000 18:25:14 -0500 - From: "Island Computers" <sales@islandco.com> ? Subject: Re: Cheap Alpha - What config is good and WHAT PRICE ? / Message-ID: <svs0ehbbu605e9@corp.supernews.com>    Well  C As to processor speeds - can't do anything slower at a better price   I We got these 533Mhz CPU and M/brds for a steal (they're still selling for : over $900 for just the CPU + M/brd on the Reseller market)  ? IDE CD-ROMS will not work on the M/brds when trying to load VMS  Must have SCSI !!!  K As to the saving on a CD-ROM - we sell the Toshiba's included in the system  at $50 - ($100 by themselves)    David         2 "JM" <vmswiz@geonospamcities.com> wrote in messageI news:C1485D99D473E35E.B3429B83D3CB66A9.E77F954C04D419ED@lp.airnews.net...  | > Island Computers wrote:  | > > , | > > Please take a look at www.islandco.com | > > < | > > Click on low cost alpha and let me know what you think | + | Great! That's getting where I want to be.  | I | $849 for my home dream machine (533mhz, 128mb mem, no disk) is not bad. G | That's close to the price I paid for the Alphastation 200 three years > | ago through Onsale.com. Gives me a number to save towards... | J | ... and I can start hinting to the wife that I really NEED to spend that& | much money to be a happier person :) | 6 | Ever notice computer widows have no sense of humour? | C | Any chance of lowering the price by dropping the processor speed?  | B | 266mhz or 333mhz would be ideal starter systems if significantlyJ | cheaper, or are the old chips just unavailable or more expensive? I know? | they probably aren't typical "commodity" items. What can this  | motherboard handle?  | = | I'd also want to make the Cdrom reader optional (I've got a I | lightning-fast Plextor 12plex already). You might want to use IDE cdrom G | readers on this box. The XP-900 at the office seems to work just fine $ | and plenty fast using IDE for CDs. | G | The narrow scsi adapter is useful and relatively necessary in all VMS 
 | situations.  | F | How about a good (333mhz), better (433mhz), best (533mhz) selection? | E | At the low end, even a $50 reduction is a big deal. We want as many D | C-code porting college students buying these things as possible :) | H | Links directly to the vendors web documents on each of the items wouldG | be very useful. Keep your margins up by not having to answer too many  | questions. | 	 | Thanks!  |  | *JM*   ------------------------------  # Date: Tue, 31 Oct 2000 02:57:35 GMT 2 From: "Zane H. Healy" <healyzh@shell1.aracnet.com>? Subject: Re: Cheap Alpha - What config is good and WHAT PRICE ? 7 Message-ID: <zyqL5.7486$Qz2.180298@typhoon.aracnet.com>   & JM <vmswiz@geonospamcities.com> wrote:F > 333 or 433mhz would be ideal. Whatever is cheapest to drive the costG > down. I have several Alphastation 500 266 and 333's at the office and > > they are easily 3-4 times faster than the Alphastation 200.   F Beware, the AlphaStation 500's are one of the *WORST* Hobbyist Alpha'sH around.  They look nice, and are a good machine, BUT the RAM for them is9 ungodly expensive!  As a result of the cost of memory my  L AlphaStation 500/333 is still sitting with the 96MB that it came with, whileG my PWS433au which replaced it came with 64MB was immediatly upgraded to F 320MB.  Of course my AlphaStation 500/333 makes a nice server when youJ don't run DECwindows on it (especially with an extra SCSI controller and a DE500).    			Zane    ------------------------------  % Date: Mon, 30 Oct 2000 20:27:23 -0000 2 From: "Kevin Playford" <kevin.playford@virgin.net> Subject: Re: DECdtm QUESTIONS B Message-ID: <IQkL5.30034$h%4.234558@news2-win.server.ntlworld.com>   Alan,   G I think the original question was about whether DECdtm works over SMCI.   K As Jim has already pointed out if the SCS cluster transport is able to make K use of SMCI for intra-cluster comms then DECdtm will also make use of it by H virtue of its use of IPC and by IPCs use of SCS for intra-cluster comms.   regards  Kevin   8 "Alan Potter" <alan_potter@hotmail.com> wrote in message) news:8ti13v$76p$1@newsg2.svr.pol.co.uk... / > fabio_compaq@ep-bc.petrobras.com.br wrote ... J > > Do the DECdtm works over SCS  in a Cluster Configuration ? What means, if > I  > > begin to use SMCI (Shared J > > Memory Cluster Interconnect) , it will  improve the speed of my DECdtm > > transactions ??? > I > From memory, the amount of CPU time and network traffic taken up by the I > DECdtm part of most database transactions was miniscule compared to the K > amount of work being done by the database applications themselves.  Hence J > the benefit to be gained from speeding up DECdtm traffic, as compared to any C > other parts of the transaction, would be likely to be very small.  > > > Are you sure that your problem lies in the DECdtm subsystem? > 
 > regards, > /alan  >  >    ------------------------------  % Date: Mon, 30 Oct 2000 22:50:47 -0400 % From: Jose Mendez <jmendez@coqui.net>  Subject: DECforms and ACMS, Message-ID: <B623ABC7.F82%jmendez@coqui.net>  I I am a Network Manager, not a System Manager.  However, there is an error H popping up on my user's screen and my network is being blamed for it.  I: don't agree so I ask for someone to help me clear this up.  F We are running a fairly large database application using RdB, ACMS forL transaction management, and DECforms.  When the first hundred or two hundredK user log in they do so with no problem.  Then after users start getting the  following error on screen:  D     No layout in the DECforms form for menu fits this terminal type,     language, and display size.   I The login script then displays a text message, requesting to hit enter at J the end, to then enter into the first form of the application.  As soon asK return is hit, the error message "Initial task failed" is displayed and the  session is disconnected.  K I've noticed that during lunch hour, terminals log in with no problems, but 3 then after 1:00PM things start getting hairy again.   H My question is, who provides DECforms with a description of the terminal type, language and display?     Help will be deeply appreciated.  
 Thank you.   Jose Mendez  Dept. of Transportation  Puerto Rico    ------------------------------  % Date: Mon, 30 Oct 2000 21:11:46 -0600 7 From: "David J. Dachtera" <djesys.nospam@earthlink.net>  Subject: Re: DECforms and ACMS- Message-ID: <39FE3872.3DCB4625@earthlink.net>    Jose Mendez wrote: > K > I am a Network Manager, not a System Manager.  However, there is an error J > popping up on my user's screen and my network is being blamed for it.  I< > don't agree so I ask for someone to help me clear this up. > H > We are running a fairly large database application using RdB, ACMS forN > transaction management, and DECforms.  When the first hundred or two hundredM > user log in they do so with no problem.  Then after users start getting the  > following error on screen: [snip]M > I've noticed that during lunch hour, terminals log in with no problems, but 5 > then after 1:00PM things start getting hairy again.  > J > My question is, who provides DECforms with a description of the terminal > type, language and display?   F Well, there's nowhere near enough info. here to determine the cause of
 the problem.    H However, you may be right if your bs detector is buzzing as loud as mineE is. The systems people need to dig deeper into this. It may well be a F "network" device - a terminal server comes to mind as one possibility.D Until you get some more in-depth data, however, it's all just finger	 pointing.    --   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: Mon, 30 Oct 2000 14:56:37 -0500 - From: "Peter Weaver" <peter.weaver@stelco.ca> - Subject: Re: Decwindows TPU/EVE: Window title / Message-ID: <svrkkp8bpfac3a@corp.supernews.com>   : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:39FCCF91.29E51728@videotron.ca... >...C > But how can I set that value ? I can't seem to find documentation  for @ > GET_DEFAULT on the documenmtation (except for the TPU built-in
 help), andD > there doesn't seem to be any reference to the file EVE.DAT that is
 being read > nor for its format.  > D > I want each window to show the filename being edited as the title. (get buffer name). >...  	 Try this;    tpu set (WIDGET, get_info F (SCREEN,"widget"),"title",get_info(get_info(buffers,"current"),"name") );   ------------------------------  % Date: Mon, 30 Oct 2000 17:29:31 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca> - Subject: Re: Decwindows TPU/EVE: Window title , Message-ID: <39FDE82B.731459DB@videotron.ca>   Peter Weaver wrote:  > tpu set (WIDGET, get_info H > (SCREEN,"widget"),"title",get_info(get_info(buffers,"current"),"name") > );     WOW ! and it works too !    - By the way, you can get the buffer name with  D 	get_info(current_buffer,"name")  a bit simpler than your syntax :-)    L Searching the SET (WIDGET in the sys$examples:*.tpu, I did find the example,M as a matter of fact, it is the offending procedure that removes the copyright  and puts "EVE" instead.   L It is interesting that there is a simple SET(ICON_NAME,"value") command, butB to set the window's title, you have to fiddle with widgets et all.   Thanks for the answer.   ------------------------------  % Date: Mon, 30 Oct 2000 20:31:54 +0100 = From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com> % Subject: Re: eBay (guess what) again. ) Message-ID: <39FDCCAA.BDAB9B7A@gtech.com>   : > > What we can gather from eBay's statements is that they> > > continue to have availability problems with their back-end? > > database servers (which are all Sun/Oracle, I believe), and B > > they are trying yet something else to improve this sorry stateC > > of affairs.  You'd think that Sun would make sure that they are E > > a shining example of availability since the embarrassing failures D > > of the past, but apparently they've been unable to do this, evenB > > after much talk about how Sun engineers were working with eBay > > to improve the situation.   7 > Does the post say this, more FUD based on no evidence  > on your part.   E I do not think it is FUD that eBay have extremely much downtime. I do  not # think it is FUD that eBay uses SUN.    Arne   ------------------------------  % Date: Mon, 30 Oct 2000 22:14:27 -0000 - From: wspencer@ap.nospam.org (Warren Spencer) % Subject: Re: eBay (guess what) again. / Message-ID: <svrsm370gn2773@news.supernews.com>   0 jfmezei.spamnot@videotron.ca (JF Mezei) wrote in" <39FDAD6D.45C99AB9@videotron.ca>:    >andrew harrison wrote: < >> If however you want availability and scalability then for: >> scalability reasons you need to partition your data and7 >> your users which is a much trickier proposition than # >> simply running OPS on two nodes.  >>  : >> Strangely this isn't spin as any Oracle OPS trained DBA >> will tell you >  >   >OK, here is a serious question: > B >Is this where 64 bit memory and VLM make a significant differenceI >because it allows you to scale a single database without partitioning it  >?    K AFAIK, this is precisely the Compaq/Oracle vision for super-scale Internet   server installations.    ws   --  3 << What if there were no hypothetical questions? >>    ------------------------------  # Date: Mon, 30 Oct 2000 23:03:58 GMT + From: Jordan Henderson <jordan@my-deja.com> % Subject: Re: eBay (guess what) again. ) Message-ID: <8tkuoq$df3$1@nnrp1.deja.com>   * In article <39FDB480.8738A244@uk.sun.com>,3   andrew harrison <andrew.nospam@uk.sun.com> wrote:  > Jordan Henderson wrote:  > > . > > In article <39F961BA.9CB2B675@uk.sun.com>,6 > > andrew harrison  <andrew.nospam@uk.sun.com> wrote: > > >Warren Spencer wrote: > > >>7 > > >> wwebb1@email.usps.gov (Webb, William W) wrote in E > > >> <D46FE9B132FB9B44AEC242A96E4AB7502CB284@rlghncst625.usps.gov>:  > > >> > > >> -- snip --  > > >>H > > >> >As part of the upgrade, eBay has begun to divide its system intoH > > >> >different databases and onto different systems. The goal is thatD > > >> >if one feature on the system goes down, the rest of the site will- > > >> >still be available, the company said.0 > > >> > > > >> > > >> -- snip --a > > >>C > > >> It appears they are partitioning the application to increaseh availabilityC > > >> and scalability.  If so, IMHO, it's a clumsy and course hack  that likelyaF > > >> would be unneccessary it the application was hosted on OpenVMS, wheretF > > >> galaxy and clusters provide native partitioning and incremental > > >> scalability.. > > >>E > > >So if you had an Oracle Database running on say a 2 node OpenVMSrE > > >cluster for performance reasons. How would you tune the databaseyD > > >if for example you wanted to add another 2 nodes and run OPS onA > > >4 rather than 2 nodes. How would you ensure that the queriesr. > > >are evenly balanced across all the nodes. > > >rB > > >Of course you don't have to partion the database but rememberD > > >you want the best performance ideally you are aiming for linearB > > >scalability into your single DBMS by adding additional nodes. > > >rA > > >Get back to me when you have though it out and please pleasemD > > >don't come back with OpenVMS clusters just do it transparently, > > >they don't. > > >  > >L> > > This is how Andrew spins here, knocking down one issue and= > > ignoring the real issue at hand.  Here he focusses on the > > > scaleability aspect that wasn't actually mentioned by eBay= > > in the quoted paragraph, ignored the availability aspect,M> > > which is what was at issue and was given as the reason for > > the change.e > > 6 > Jordan it doesn't make any difference. OPS gives you= > better availability and if you don't care about scalabilityh: > then you don't need to partition your users or you data.  7 Hey, don't argue with me about database availability onl= Sun Hardware.  It's _eBay_ above who are saying that the goalu@ of partitioning their databases is to provide more availability.   >5; > If however you want availability and scalability then fori9 > scalability reasons you need to partition your data andm6 > your users which is a much trickier proposition than" > simply running OPS on two nodes. > 9 > Strangely this isn't spin as any Oracle OPS trained DBA  > will tell you. >e: > > What we can gather from eBay's statements is that they> > > continue to have availability problems with their back-end? > > database servers (which are all Sun/Oracle, I believe), andrB > > they are trying yet something else to improve this sorry stateC > > of affairs.  You'd think that Sun would make sure that they areiE > > a shining example of availability since the embarrassing failures-D > > of the past, but apparently they've been unable to do this, evenB > > after much talk about how Sun engineers were working with eBay > > to improve the situation.v > >n7 > Does the post say this, more FUD based on no evidence  > on your part.p > < > You have consigned yourself to the Rob Young backing choir" > which is a very sad place to be. >e  < I think it's sadder to the pathetic lone Sun voice singing a! capella in the Church of OpenVMS..  > > > Of course, we can't _prove_ that Sun HW/SW has anything to= > > do with this (not surprising with Sun's history of havingv@ > > customers sign NDAs in order to get service), but there sure9 > > are more and more indications that Sun reliability isfC > > questionable.  And, it's not just people in comp.os.vms who sayi@ > > this.  I redirect everyone's attention to the recent Gartner > > statements reported here:w > >n >p" > > http://x63.deja.com/getdoc.xp?2 AN=686121920&CONTEXT=972659765.1119944705&hitnum=0 > > B > > Back to the issue of partitioning for scaleability, since thisA > > is all Andrew wants to talk about.  Sure, it's a great thing.jA > > I wonder if we could get Andrew to admit that Galaxy provides D > > flexibility in this area that Sun cannot?  Naaah, he's just here$ > > to tear down Compaq and OpenVMS. > >  >f9 > Right Jordan so you are happy to try to FUD Java on theo< > basis of what Sun might do but hasn't since Java appeared,5 > you continue to try to FUD Sun over eBay and at the36 > same time you accuse me of trying to tear Compaq and > OpenVMS down.   9 What?  You're NOT trying to tear Compaq and OpenVMS down?i  ; What is it you are doing here then, Andrew?  Just educatingC, us poor lost souls on the error of our ways?  : If you aren't trying to tear down Compaq and OpenVMS down,; you sure seem to spend a lot of time doing it.  I guess youm8 should write a book "How to tear down Compaq and OpenVMS without really trying."e   >t; > As I said in a previous posting get more mirror time look  > long and hard.  7 In any case, what you do is FUD, what _I_ do is counterk: what you say.  FUD is dependent on the audience.  The term9 FUD was invented by Gene Amdahl, in reference to what IBMm8 salesmen would do to his accounts.  IBM would attempt to; scare these Amdahl accounts, indicating that Amdahl systemsn< might not work with newer OS releases, might not support the8 peripherals they need, might not exist at _all_ in a few years.  ; I don't hang out in comp.sys.sun telling Sun customers thatt; they had better look into reliable, truly scaleable OpenVMSi; solutions as Sun doesn't really seem to have a clue.  No, Io8 post in comp.os.vms, on-topic discussions of OpenVMS and? countering people who are obviously trying to panic the OpenVMSe
 community.  9 What you do is question every Compaq move, belittle every 8 OpenVMS advantage, tout every Sun sale and generally try= to disparage all the things that the people in cov hold dear.p That is spreading FUD.   >t: > The real damage isn't being done by me its being done by: > ill informed OpenVMS boosters posting stuff either about< > OpenVMS or about UNIX which is total B***S**T particularly3 > when the masses fall for it hook line and sinker.b > 9 > As for Galaxy, show me a customer reference for a large 8 > database implimentation using Galaxies on Alphas and I+ > will sit up and take a bit more interest.u >t  5 Well, this is changing the subject.  Unable to accepty2 that Galaxy features might enhance your ability to6 partition applications, you quickly change the subject to market penetration.  ; Well, OK.  Cerner, for one, seems to tout the advantages of-6 Galaxyfor partitioning systems for their applications:  : http://www.openvms.compaq.com/openvms/brochures/cerner-gs/  5 > As Mr Heller once said, technology is like fish theg5 > longer you leave it on the shelf the less appealingM > it is. >o	 > Regardsc > Andrew Harrisonr > Enterprise IT Architectt >r   -- -Jordan Henderson  jordan@greenapple.comv    & Sent via Deja.com http://www.deja.com/ Before you buy.y   ------------------------------  % Date: Mon, 30 Oct 2000 14:37:05 -0600h* From: Patrick Spinler <pspinler@yahoo.com>6 Subject: Example of using psuedoterm device from DCL ?) Message-ID: <39FDDBF1.35145CBA@yahoo.com>s   Hi:s  F Is it possible to create and use psuedoterm devices from DCL ?  (FTA*:D devices made via PTD$CREATE() from the I/O users guide)  If so, does? anyone have any pithy examples they might be willing to share ?y   Thanks ! -- Pat   ------------------------------  % Date: Mon, 30 Oct 2000 17:01:23 -0500 5 From: Forrest Kenney <Forrest.Kenney@compaq.com.doom> : Subject: Re: Example of using psuedoterm device from DCL ?/ Message-ID: <39FDEFB3.41D702AC@compaq.com.doom>o  G     There is no support for using pseudo terminals from DCL.  Depending4! upon what is needed you should beoI able to hack up the example I sure have for all kinds of things..The reals$ issue is that the buffer objects andG channels needed for the psuedo terminals do not survive image run down.n  H     I am the person originally responsible for the pseudo terminal code.     Forrest Kenney OpenVMS Development Compaq     Patrick Spinler wrote:   > Hi:  >tH > Is it possible to create and use psuedoterm devices from DCL ?  (FTA*:F > devices made via PTD$CREATE() from the I/O users guide)  If so, doesA > anyone have any pithy examples they might be willing to share ?n > 
 > Thanks ! > -- Pat   ------------------------------   Date: 30 Oct 2000 20:41:03 GMT* From: chd@vms.arizona.edu (Chris De Young)* Subject: File encryption software for VMS?0 Message-ID: <8tkmcv$thk$1@news.ccit.arizona.edu>   Hi,f  M Does anyone know of a utility for encryping files or sets of files using somegM strong algorithm (Blowfish, Twofish, Rijndael, etc.) for VMS?  Something thatnJ will encrypt individual files is ok, or something which makes an encryptedL containter file which is then mounted like another disk device would be fine too.   Thanks,n -Chris   chd@arizona.edu    ------------------------------  % Date: Mon, 30 Oct 2000 17:32:33 -0400,- From: JF Mezei <jfmezei.spamnot@videotron.ca> . Subject: Re: File encryption software for VMS?, Message-ID: <39FDE8E0.39873ACD@videotron.ca>   Chris De Young wrote:r >  > Hi,t > O > Does anyone know of a utility for encryping files or sets of files using somer? > strong algorithm (Blowfish, Twofish, Rijndael, etc.) for VMS?     K Digital had it then. Now sure if Compaq has it now. In V5.5-2's help, thereg: was even a topif for ENCRYPT. But it has disapeared in 7.2   ------------------------------    Date: 30 Oct 2000 17:55:08 -05009 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)n. Subject: Re: File encryption software for VMS?+ Message-ID: <UbGdR1SeVFCI@eisner.decus.org>e  \ In article <39FDE8E0.39873ACD@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes: > Chris De Young wrote:r >>   >> Hi, >> mP >> Does anyone know of a utility for encryping files or sets of files using some@ >> strong algorithm (Blowfish, Twofish, Rijndael, etc.) for VMS?  H I believe Bruce Schneier's latest book recommends Triple-DES for a whileI over AES (which subsequently turned out to be Rijndael).  Since Bruce wasdI the originator of the other two on your list, that would seem to indicate F that triple-DES is adequate for your application, unless there is some other unspecified requirement.  C Of course DEC Encryption for OpenVMS only does DES, but would it be 7 possible to compose triple-DES through that interface ?k  M > Digital had it then. Now sure if Compaq has it now. In V5.5-2's help, theren< > was even a topif for ENCRYPT. But it has disapeared in 7.2  4 Whatever works for you, don't tell any Canadians :-)   ------------------------------   Date: 30 OCT 2000 22:08:32 GMT4 From: karcher@thuria.waisman.wisc.edu (Carl Karcher). Subject: Re: File encryption software for VMS?6 Message-ID: <30OCT00.22083250@thuria.waisman.wisc.edu>  B In a previous article, chd@vms.arizona.edu (Chris De Young) wrote:  O ->Does anyone know of a utility for encryping files or sets of files using some O ->strong algorithm (Blowfish, Twofish, Rijndael, etc.) for VMS?  Something that L ->will encrypt individual files is ok, or something which makes an encryptedN ->containter file which is then mounted like another disk device would be fine ->too.  F OpenSSL can do that with it's ENC utility. Probably more than you needF though. See http://www.openssl.org/ and http://www.free.lp.se/openssl/   ------------------------------  % Date: Mon, 30 Oct 2000 19:55:07 -0500o, From: "Glenn C. Everhart" <Everhart@GCE.com>. Subject: Re: File encryption software for VMS?' Message-ID: <39FDD21B.2CAC70E0@GCE.com>r  ; I wrote an encrypting-disk host file for fddriver ages ago.-: Its crypto is pretty simple and was not designed to resist9 serious cryptographic attacks, but it has 3 subroutines,  < one for key setup, the others to encrypt and decrypt blocks.? Add in a strong algorithm for these (possibly, use the existingp9 selected keys as the key in some combo) and you have whata: you want. The whole thing is in source, and it runs on Vax; or Alpha. Fddriver also has appeared as frdriver, dtdriver,-9 and zrdriver (i.e., pick your favorite name). It allows a : user mode process to do the actual storage work. There are< several of these host processes written. First one just used! an array to make a "memory disk".J  < The resulting cryptodisk acts like a normal VMS disk, though= because it handles blocks one at a time it could be speedier.t< That is an artifact of the host process btw, not the driver.   Chris De Young wrote:  >  > Hi,  > O > Does anyone know of a utility for encryping files or sets of files using someEO > strong algorithm (Blowfish, Twofish, Rijndael, etc.) for VMS?  Something thatwL > will encrypt individual files is ok, or something which makes an encryptedN > containter file which is then mounted like another disk device would be fine > too. > 	 > Thanks,) > -Chris >  > chd@arizona.edu-   ------------------------------  % Date: Mon, 30 Oct 2000 19:29:11 -0500m, From: taterskins@patriot.net (Ramon L. Tate). Subject: Re: File encryption software for VMS?D Message-ID: <taterskins-ya023480003010001929110001@news.patriot.net>  K In article <8tkmcv$thk$1@news.ccit.arizona.edu>, chd@vms.arizona.edu wrote:    > Hi,O > O > Does anyone know of a utility for encryping files or sets of files using some(O > strong algorithm (Blowfish, Twofish, Rijndael, etc.) for VMS?  Something that L > will encrypt individual files is ok, or something which makes an encryptedN > containter file which is then mounted like another disk device would be fine > too. > 	 > Thanks,t > -Chris >  > chd@arizona.eduO  F Once upon a time there was an encryption package available for ZIP andH UNZIP, but I've not been actively involved in using these tools for someG time and don't know where you might find it. Try the InfoZIP web site --& they might be able to point you to it.   -- c
 Ramon L. Tatee	 Casa Maaa= taterskins@patriot.net   "Skin" that "tater" before replying!    ------------------------------  # Date: Mon, 30 Oct 2000 23:14:04 GMTl* From: "Kevin Bergman" <kbergman7@home.com>2 Subject: How to remove trailing FF from a printjob; Message-ID: <0hnL5.12112$25.2838097@news1.rdc1.ne.home.com>s  L I need to print labels/forms on an impact printer, but every print job seems6 to be including a FF at the end, causing a page-eject.  G I need to send 6 lines to the printer, and have the printer print the 6d lines, and NOT receive a FF.  ' Any qualifiers/parameters I'm missing ?t   ------------------------------   Date: 30 Oct 2000 23:48:18 GMT& From: brown@taiga.gmcl.com (Rob Brown)6 Subject: Re: How to remove trailing FF from a printjob. Message-ID: <8tl1c2$l9m$1@dagger.ab.videon.ca>  ) Kevin Bergman (kbergman7@home.com) wrote:-N : I need to print labels/forms on an impact printer, but every print job seems8 : to be including a FF at the end, causing a page-eject.  I : I need to send 6 lines to the printer, and have the printer print the 6o : lines, and NOT receive a FF.  D It seems to me that what you need to do is set your page length to 7I lines, or whatever is appropriate to get from the top of one label to thea top of the next.  C There are a few different ways to do this, depending on your setup.i  J If your printer *always* has the labels, it might be appropriate to changeG the printer setup (ie, in the printer) to have the correct form-length.sF Alternately, you might to a SET TERMINAL (or SET PRINTER?) to tell VMSE that your printer doesn't know how to do form-feeds, so VMS will send G multiple line-feeds instead.  You would use the same command to specifyc how many lines per page.  E If sometimes you use these labels, and sometimes something else, thenoH perhaps you want to set up a special form for the labels, specifying how5 long the label is.  See DEFINE/FORM.  Then you could:o     $ PRINT/FORM=LABEL file-list  G Later on (or before), you could put the label stock in the printer and:a  #   $ SET QUEUE/FORM=LABEL queue-name   . OTOH, does the printer even need to be queued?   - Robd       -- --  / Rob Brown                        brown@gmcl.coms6 G. Michaels Consulting Ltd.      (780)438-2101 (voice)4 Edmonton                         (780)437-3367 (FAX)1                                  http://gmcl.com/    ------------------------------  # Date: Tue, 31 Oct 2000 01:48:07 GMT * From: "Kevin Bergman" <kbergman7@home.com>6 Subject: Re: How to remove trailing FF from a printjob; Message-ID: <rxpL5.12120$25.2872195@news1.rdc1.ne.home.com>   . Thanx - I will look into this tonight/tomorrow  I The printer doesnt allow a form-length as small as the label, and I didnto$ think of the device characteristics.  ! I'lllet you know how I make out !r  3 "Rob Brown" <brown@taiga.gmcl.com> wrote in messagev( news:8tl1c2$l9m$1@dagger.ab.videon.ca...+ > Kevin Bergman (kbergman7@home.com) wrote:lJ > : I need to print labels/forms on an impact printer, but every print job seemsr: > : to be including a FF at the end, causing a page-eject. >iK > : I need to send 6 lines to the printer, and have the printer print the 6s  > : lines, and NOT receive a FF. > F > It seems to me that what you need to do is set your page length to 7K > lines, or whatever is appropriate to get from the top of one label to thet > top of the next. >oE > There are a few different ways to do this, depending on your setup.h >oL > If your printer *always* has the labels, it might be appropriate to changeI > the printer setup (ie, in the printer) to have the correct form-length. H > Alternately, you might to a SET TERMINAL (or SET PRINTER?) to tell VMSG > that your printer doesn't know how to do form-feeds, so VMS will senddI > multiple line-feeds instead.  You would use the same command to specifyo > how many lines per page. > G > If sometimes you use these labels, and sometimes something else, then J > perhaps you want to set up a special form for the labels, specifying how7 > long the label is.  See DEFINE/FORM.  Then you could:, >n  >   $ PRINT/FORM=LABEL file-list >hI > Later on (or before), you could put the label stock in the printer and:  > % >   $ SET QUEUE/FORM=LABEL queue-namec >O0 > OTOH, does the printer even need to be queued? >c > - Rob  >l >l >h > -- > -- >n1 > Rob Brown                        brown@gmcl.com,8 > G. Michaels Consulting Ltd.      (780)438-2101 (voice)6 > Edmonton                         (780)437-3367 (FAX)3 >                                  http://gmcl.com/e   ------------------------------  % Date: Mon, 30 Oct 2000 22:27:50 -0400e- From: JF Mezei <jfmezei.spamnot@videotron.ca>t6 Subject: Re: How to remove trailing FF from a printjob, Message-ID: <39FE2E22.525EF7D3@videotron.ca>   Kevin Bergman wrote: > N > I need to print labels/forms on an impact printer, but every print job seems8 > to be including a FF at the end, causing a page-eject.  7 What software is generating the file that gets queued ?a    L you might want to do a SHOW QUEUE queue_name/FULL and look at the meaning of each option.   (HELP SET QUEUE /option)  G Also, you can SHOW QUEUE/FORM form_name /FULL will give you info on the ; "form_name" that is set as the default form for that queue.i  D I print labels to an LA75 on a queue and have no problems with this.   ------------------------------  # Date: Mon, 30 Oct 2000 23:31:44 GMT + From: "Andy Proctor" <aproctor@hotmail.com>tE Subject: I know the answer but just can't remember what...(Pagefiles) 4 Message-ID: <972948541.481695@ananke.eclipse.net.uk>  : VMS gurus, please may I ask a really simple/daft question?L I am currently working on an Alpha3000/300LX VMS 6.2-1h3 and all I am trying+ to do is increase the size of the pagefile.tI Its currently 29000 blocks (too small especially with Pathworks running),tH and I want to increase it to 130000 blocks. There are over 300000 blocksI free on the disk. (only a 1GB system disk, user areas are on other disks)sL When running swapfiles.com I get the error: "File Header full" when it triesL to create the new pagefile, and of course it does not do it. The swapfile is4 10600 blocks, dumpfile currently set to 1000 blocks.  L Now, I've had a long day :( and I know I know the solution but I just cannot remember where to go now!yJ I have performed an Analyze/disk/repair and all reports OK, and purged theF old versions of the pagefile that were on the disk. (It's a customer's system)x  J So, would you experts be able just to point out the causes of this sort ofK thing and things to do. It will I'm certain jog my memory and ensure that I  never forget this again!   Thanks on advance as always.   Andy   ------------------------------  # Date: Tue, 31 Oct 2000 00:28:40 GMTt From: i.burgess@UQ.edu.au I Subject: Re: I know the answer but just can't remember what...(Pagefiles)m) Message-ID: <8tl3nh$hie$1@nnrp1.deja.com>e  4 In article <972948541.481695@ananke.eclipse.net.uk>,.   "Andy Proctor" <aproctor@hotmail.com> wrote:< > VMS gurus, please may I ask a really simple/daft question?G > I am currently working on an Alpha3000/300LX VMS 6.2-1h3 and all I amt trying- > to do is increase the size of the pagefile.eA > Its currently 29000 blocks (too small especially with Pathworks 	 running),dC > and I want to increase it to 130000 blocks. There are over 300000h blocksD > free on the disk. (only a 1GB system disk, user areas are on other disks)H > When running swapfiles.com I get the error: "File Header full" when it trieseB > to create the new pagefile, and of course it does not do it. The swapfile iso6 > 10600 blocks, dumpfile currently set to 1000 blocks. >lG > Now, I've had a long day :( and I know I know the solution but I justt cannot > remember where to go now!sH > I have performed an Analyze/disk/repair and all reports OK, and purged the H > old versions of the pagefile that were on the disk. (It's a customer's	 > system)c ...n Help/message HEADERFULL says...l  HEADERFULL,  file header is full  '   Facility:     SYSTEM, System Servicess  G Explanation:  The file header map area on the volume is full and cannot E be extended. This error occurs only when the file header extension is > inhibited, for example, when the index file is being extended.  A User Action:  Compress the volume by copying it with BACKUP. Thisb@ condition can be avoided by increasing the value for the /HEADERF qualifier of the INITIALIZE command. For ODS-2 disks, the value of theB /HEADER qualifier should equal the approximate number of files you expect to store on the disk.  ? In this case, the pagefile (not the index file) has file headermD extension inhibited.  In other words, to extend the pagefile a largeC number of pointers to contiguous blocks on the disk have been added = to the file header, but it still is not as big as you wanted.t, (It may be big enough...?  See DIR/SIZE=ALL)  G The real answer is backup and restore the disk to make files, includingn9 the pagefile, contiguous and you'll be able to extend it.r  B Otherwise clean up the disk as much as possible and see if you canH create a new version in the free space (doubtful).  In any case you will@ have to reboot to use the extra space in the pagefile or the new	 pagefile.   F A 1GB system disk is small these days.  Maybe you can install a second pagefile on another disk.-  F You probably don't need that much swapfile (or any), since if you ever swap the page file will do.>  E In a later OpenVMS version you would be able to put the dump file offFA the system disk, but failing that you can dump to the pagefile tou reclaim that 1000 blocks.D  ; Cheers,  Ian        (Ian Burgess, University of Queensland)g    & Sent via Deja.com http://www.deja.com/ Before you buy.    ------------------------------   Date: 31 Oct 2000 00:37:46 GMT& From: brown@taiga.gmcl.com (Rob Brown)I Subject: Re: I know the answer but just can't remember what...(Pagefiles):. Message-ID: <8tl48q$2ls$1@dagger.ab.videon.ca>  * Andy Proctor (aproctor@hotmail.com) wrote:  N : When running swapfiles.com I get the error: "File Header full" when it triesN : to create the new pagefile, and of course it does not do it. The swapfile is6 : 10600 blocks, dumpfile currently set to 1000 blocks.  D Your disk or your pagefile is too fragmented to extend your pagefile. without filling up the pagefile's file header.  D At this point, I would defragment the disk (I always do a backup and, restore, but perhaps you have better tools).  C If this task is too onerous, you could dump the file header of your F current page file and see how fragmented it is.  If it is already veryI fragmented, you might be able to make some gains by copying the file.  IftF the copy is less fragmented, you could reboot, delete the original and? then try extending.  However, I doubt that the results would bes satisfying.u  
 Good luck.   - Robs     -- --  / Rob Brown                        brown@gmcl.comd6 G. Michaels Consulting Ltd.      (780)438-2101 (voice)4 Edmonton                         (780)437-3367 (FAX)1                                  http://gmcl.com/t   ------------------------------   Date: 30 Oct 2000 23:03:45 GMT* From: chd@vms.arizona.edu (Chris De Young)  Subject: Re: MOZILLA M18 crashes0 Message-ID: <8tkuoh$223$1@news.ccit.arizona.edu>  W In article <39ECC73E.FB8BBA67@theblakes.com>, Colin Blake <colin@theblakes.com> writes: 0 >The new kit is available from the Mozilla site. >cL >It has been submitted to the OpenVMS site, but I don't think its there yet.' >It should be by morning at the latest.  >t  # Got it from the Mozilla.org site...o  	 $ mozillai Starting mozilla-bin...g3 %DCL-W-ACTIMAGE, error activating image CMA$TIS_SHReL -CLI-E-IMGNAME, image file VMS$DKA0:[SYS0.SYSCOMMON.][SYSLIB]CMA$TIS_SHR.EXE: -SYSTEM-F-SHRIDMISMAT, ident mismatch with shareable image  N System is V7.1-1H2, and I do have the ALPLIBR patch.  The date on the PCSI kitL I got from mozilla.org is 16-OCT-2000 20:31:14.74 - is this the current one?   Thanks,  -Chris   chd@arizona.edu?   ------------------------------  % Date: Mon, 30 Oct 2000 20:30:44 -060037 From: "David J. Dachtera" <djesys.nospam@earthlink.net>>* Subject: Re: Nice job - porting VNC to VMS- Message-ID: <39FE2ED4.7F4E0068@earthlink.net>r   Larry Kilgallen wrote: > j > In article <39FD48A6.1139514@gtech.com>, Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com> writes:
 > > [snip]7 > > VNC is a cross-platform "PC Anywhere"-like product.s > I > It is important when describing things in comp.os.vms to avoid assumingtI > that people have your own background in other operating systems.  Thereu? > is no "PC Anywhere" product that comes with VMS or Macintosh.   H Just try explaining a console prompt, DCL or RMS to someone who has onlyC ever seen Intel boxes, mainframes, UN*X, W/3x, W/9x or MacOS not to ) mention AIX, OS/400, MVS, OS3x0, etc. ...a   -- w David J. Dachtera  dba DJE Systemsn 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: Mon, 30 Oct 2000 20:35:40 GMTi From: pasmith@ppg.comi! Subject: Number of users exceededi) Message-ID: <8tkm2r$5cs$1@nnrp1.deja.com>t  D Our users have recently started getting the message Number of UsersF Exceeded when they try to log in.  To get around this, we simply giveE their accounts oper privileges (the users connect to a command filetF with no way to get out of it except to log out).  Is there a parameterC in my system or the individual user account that can be manipulated G such that they do not get this message?  The only thing we have changedhE recently was to increase the number of telnet sessions allowed and toD@ increase the drop timer for the telnet sessions.  We average 267G processes and 13 different users. (Many users log into the same accountiE based upon the occupation they are doing on the manufacturing floor.), Any advice will be appreciated.       & Sent via Deja.com http://www.deja.com/ Before you buy.h   ------------------------------  % Date: Mon, 30 Oct 2000 21:04:14 +0000n From: SysAdmin <djesys@fsi.net>h% Subject: Re: Number of users exceededV' Message-ID: <39FDE24E.3FC1A023@fsi.net>a   pasmith@ppg.com wrote: > F > Our users have recently started getting the message ?Number of UsersH > Exceeded? when they try to log in.  To get around this, we simply giveG > their accounts ?oper? privileges (the users connect to a command fileeH > with no way to get out of it except to log out).  Is there a parameterE > in my system or the individual user account that can be manipulatedeI > such that they do not get this message?  The only thing we have changed G > recently was to increase the number of telnet sessions allowed and toiB > increase the drop timer for the telnet sessions.  We average 267I > processes and 13 different users. (Many users log into the same account G > based upon the occupation they are doing on the manufacturing floor.)o! > Any advice will be appreciated.o    See HELP SET LOGINS /INTERACTIVE  A To increase the value SET at system startup time, do this in yourc SYSTARTUP_VMS.COM procedure:  # $ STARTUP$INTERACTIVE_LOGINS == xxx    David J. Dachterar   ------------------------------  + Date: Mon, 30 Oct 2000 21:33:38 +0000 (GMT)., From: Bill McLaughlin <mcbill20@hotmail.com>% Subject: Re: Number of users exceededw5 Message-ID: <F78Bf88Vtu2xqLAUtMu000055f8@hotmail.com>o   >rF >Our users have recently started getting the message =91Number of Use= rsF >Exceeded=92 when they try to log in.  To get around this, we simply = giveF >their accounts =91oper=92 privileges (the users connect to a command=  file F >with no way to get out of it except to log out).  Is there a paramet= erD >in my system or the individual user account that can be manipulatedF >such that they do not get this message?  The only thing we have chan= gediF >recently was to increase the number of telnet sessions allowed and t= oTA >increase the drop timer for the telnet sessions.  We average 267 F >processes and 13 different users. (Many users log into the same acco= unteF >based upon the occupation they are doing on the manufacturing floor.= )   >Any advice will be appreciated. >m  F It sounds like you exceeded the default interactive user limit. To ch=
 eck the=20 limit, type: $ set login@   You should see something like:  F %SET-I-INTSET, login interactive limit =3D 64, current interactive va=
 lue =3D 25  $ In order to increate the limit, use:   $ SET LOGIN/INTERACTIVE=3Dnnn   F where nnn is the maximum number of interactive sessions you want to a= llow.u   Also see $ help set login   >  >-' >Sent via Deja.com http://www.deja.com/O >Before you buy. >  >m  F _____________________________________________________________________= ____F Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.= com.  D Share information about yourself, create your own public profile at= =20e http://profiles.msn.com.   ------------------------------  % Date: Mon, 30 Oct 2000 21:48:42 +0000 . From: mpatt644 <mpatt644@netscapeonline.co.uk>% Subject: Re: Number of users exceedede4 Message-ID: <39FDECBA.6900AEE8@netscapeonline.co.uk>   try    $set login/inter=x  2 where x >= number of users you want on the system.   pasmith@ppg.com wrote: > F > Our users have recently started getting the message Number of UsersH > Exceeded when they try to log in.  To get around this, we simply giveG > their accounts oper privileges (the users connect to a command file-H > with no way to get out of it except to log out).  Is there a parameterE > in my system or the individual user account that can be manipulatedlI > such that they do not get this message?  The only thing we have changedcG > recently was to increase the number of telnet sessions allowed and tosB > increase the drop timer for the telnet sessions.  We average 267I > processes and 13 different users. (Many users log into the same account G > based upon the occupation they are doing on the manufacturing floor.))! > Any advice will be appreciated.t > ( > Sent via Deja.com http://www.deja.com/ > Before you buy.    ------------------------------  % Date: Tue, 31 Oct 2000 07:41:40 +0100w. From: "I.A. Saez" <"i.a.saez.<secret>"@tue.nl>F Subject: Re: Oracle/VMS performance problems with Multithreaded Server& Message-ID: <39FE69A4.2DED1146@tue.nl>  * OpenVMS fully supported by Oracle? No way!6 Here is my list of complaints about Oracle on OpenVMS:  - - Oracle Apps 11i is not supported on OpenVMShF - did you ever try to get Oracle I. Agent working on a OpenVMS system?H - MTS doesn't work properly on OpenVMS; even 8.1.6. has the same problemF - porting of Oracle products is far behind Unix (some time more than a year)l - etce  ? That Oracle is better supported on Unix than OpenVMS is a fact.eD Statements of direction say nothing to me. They sound like "the veryG soon now" frases of Microsoft. The daily pratice is what matters to me. H I hope that Oracle will support OpenVMS better than in the past, but forG us it will be too late and I'm sorry because I find OpenVMS a very goodn OS.u     Syltrem wrote: > 	 > Oh! Oh!t > N > When I read this, it sounded an alarm in my head! How could 7336 be the lastK > "real" version of Oracle on OpenVMS, given the current agreements betweeneL > Oracle and Compaq? And how about the fact that the customer said they willI > move to Unix? Hopefully not only because they think Oracle will be moreo > supported on that platform...  > M > In short, OpenVMS is FULLY supported by Oracle, in details, see the comment I > on that from Alan Belancik, OpenVMS Relationship Manager with Oracle at 	 > Compaq., > J >  As the OpenVMS Relationship Manager with Oracle, I am in a position toM > tell you that our two groups are working very closely together. Our goal istN > to offer customers high performance solutions that are based on AlphaServersN > running the Oracle database. These have been put to use in many of the worldA > 's most mission critical  applications requiring extremely highh > availability.bE > Your question dealt with the nature of the Oracle database and itsfK > applicability to the OpenVMS world. I can assure you that there is a veryiK > astute team of engineers and managers in place who work on the porting ofpL > the Oracle database to OpenVMS. These people are very well acquainted withM > the OpenVMS operating system and work to ensure that the products available K > from Oracle take advantage of OpenVMS's unique attributes. Our pioneeringAN > work in clustering has been reflected in Oracle's Parallel Server offerings.L > Engineers at Oracle have been working on the interplay between new OpenVMSD > technologies, such as the Galaxy Software architecture for dynamicN > partitioning, and the new "WildFire" series of AlphaServers, the GS80, GS160 > and GS320.M > True, the porting of the Oracle database begins with a set of common code,nN > however once it is released to the various "platform groups" with in Oracle,K > they begin to customize the final product to reflect the features of thatF, > operating system and hardware combination.L > Unfortunately, there is always some confusion in the field. We at DigitalN > Equipment, and now Compaq, have unfortunately contributed to that confusion.J > One item of confusion would be the perceived differences between VMS andM > OpenVMS. Some people have erroneously concluded that VMS is for VAX systemsrL > and OpenVMS is for AlphaServer systems. Actually, OpenVMS was the new nameL > applied to ALL VMS in order to better reflect its adherence to many of theN > world's standards for openness. I mention this because you reference versionM > 7.3.3 6 as being the last real VMS version. It is true that Oracle database F > products were not ported to support VAX systems once Oracle made theN > transition from their version 7.x products to their 8.x version products. ItN > is very close to being correct that version 7.3.3.6 was the last version forH > VAX systems. However, I do not see how this can be said of AlphaServer > systems running OpenVMS.J > Nonetheless, I will confer with my Oracle colleagues to see if they canK > share any insight into this contention. Naturally, it is in any company's K > best interest to try to standardize the core of their product in order towI > reduce operating expenses. Still, they need to be able to customize theKG > final product to best suit the platform it is intended for as well as'K > optimizing its performance on that platform. We have been working closelyhJ > with the engineers in the porting group and in the performance groups to > ensure that this is the case. L > We will continue to work with Oracle as our premier database partner. OurM > user surveys show that approximately 88% of our OpenVMS customers who use aaL > commercial database use one from Oracle. We will continue to invest in ourN > relationship. One benefit that we expect from this relationship is a productE > from Oracle that is best suited for our AlphaServer systems running  > OpenVMS..eK > Please let me know if more information comes to light or how I can be ofAK > further assistance. Again, thank you for bringing this information to oureI > attention. We share your belief that OpenVMS continues to be one of the - > world's best operating environments - EVER!e > Thanks and merci milles fois,v > Alan > Alan Belancike > Oracle Relationship Managerm > OpenVMS System Software Groupi > 603-884-0363 > fax: 603-884-2006g > ; > "I.A. Saez @tue.nl>" <"i.a.saez.<secret> wrote in message " > news:39F91C4B.ECB2555E@tue.nl... > > Malcolm, > >pL > > We have the same problem with Oracle 7.3.4.4 on OpenVMS 7.1. We have now	 > > a mixoI > > of MTS and dedicated connections. Dedicated for long running apps andeJ > > MTS for short connections. Oracle did investigate the problem and toldH > > me it still exists in 8.1.6 for OpenVMS. Oracle 7.3.3.6 was the last9 > > real VMS version (probably made by VMS especialists).L > >s > >  > >  See (some) of ther tar: > >  > >i > >  27-SEP-00 12:35:39S > >e > >  > >n > >  27-SEP-00 13:24:41a > >a< > >  To reproduce the slow MTS performance (for development)0 > >  - The following patches are installed here:I > >  686524-(Base BUG#544858) ORA-602 RUNNING PRO*C WITH TO_DATE AND BIND  > >  VARIABLES.mF > >  1055554-(Base BUG#553138) ORA-12612 ON W95 CLIENTS AFTER APPLYING > >  520734(430972./I > >  1218450-(Base BUG#845454) ORA-1403 REFERRING TO CERTAIN PL/SQL TABLEt > >  ELEMENTS.3 > >  1326425- ORA-1013 when sqlnet.expire_time > 0.u) > >  1309610-(Base BUG#1021962) MTS SLOW.a8 > >  1326725- Listener/Dispatchers consumes Buffered IO.F > >  - Used the following MTS parameters in V7.3.4.4 INIT.ORA (OTS14): > >  > >e > >  27-SEP-00 14:19:34n > >d > >  mts_listener_address =f6 > > "(ADDRESS=(PROTOCOL=tcp)(HOST=nlvms2)(PORT=1521))"< > >  MTS_DISPATCHERS="(ADDRESS=(PARTIAL=TRUE)(PROTOCOL=TCP)\. > >  (HOST=nlvms2)(PORT=5000))(DISPATCHERS=1)"< > >  MTS_DISPATCHERS="(ADDRESS=(PARTIAL=TRUE)(PROTOCOL=TCP)\. > >  (HOST=nlvms2)(PORT=5001))(DISPATCHERS=1)" > >  mts_servers = 2 > >  mts_service = ots14D > >  - Restart the instance --> will register the dispatchers at the > > listener+ > >  - Now within the database using SCOTT:iG > >  - Create a BIGEMP table from EMP with a large number of rows in iteA > >  (I used BIGEMP with 458752 rows) ---> BIGEMP.DMP export filecL > >  - Create two connect aliases, one for MTS (OTS14) and one for DEDICATED
 > > server > >  (OTS14D) :d > >  ots14.world = > >  (DESCRIPTION =y > >  (ADDRESS_LIST = > >  (ADDRESS =p > >  (PROTOCOL = TCP)e > >  (Host = nlvms2) > >  (Port = 1521) > >  ) > >  ) > >  (CONNECT_DATA = > >  (SID = ots14)  > >  (GLOBAL_NAME = ots14.world) > >  ) > >  ) > >  ots14d.world =- > >  (DESCRIPTION =  > >  (ADDRESS_LIST = > >  (ADDRESS =a > >  (PROTOCOL = TCP)) > >  (Host = nlvms2) > >  (Port = 1521) > >  ) > >  ) > >  (CONNECT_DATA = > >  (SID = ots14)  > >  (GLOBAL_NAME = ots14.world) > >  (SERVER = DEDICATED)m > >  ) > >  )0 > >  - Create the following test script TEST.SQL5 > >  select to_char(sysdate, 'hh24:mi:ss') from dual;  > >  set termout off;o- > >  select * from bigemp where ename='KING';i > >  set termout on;5 > >  select to_char(sysdate, 'hh24:mi:ss') from dual;t > >  spool a.a! > >  select count(*) from bigemp;p > >  spool off;a5 > >  select to_char(sysdate, 'hh24:mi:ss') from dual;l	 > >  exita, > >  - and reproduce the problem as follows:6 > >  $ sqlplus scott/tiger@ots14d --> dedicated server > >  SQL> @testf > >  14:08:16 --> start time0 > >  14:08:55 --> end of query with where clause > >  count(*) --> 458752 > >  14:09:11 --> end timeK > >  Total elapsed time : 39 seconds for the first part, 16 seconds for theh
 > > second > >  Now with MTS:2 > >  $ sqlplus scott/tiger@ots14 --> shared server > >  SQL> @testd > >  14:12:15 --> start time0 > >  14:18:12 --> end of query with where clause > >  count(*) --> 458752 > >  14:18:29 --> end timeI > >  Total elapsed time : 357 (!!) seconds for the first part, 17 seconds  > > for the > > >  second part, a huge difference with DEDICATED connection. > >2 > >3 > >  27-SEP-00 15:47:41. > >n > >  .E > >  Tested with V8.1.6.0 on OpenVMS (same configuration, same data):2> > >  The following patches are installed on this installation:E > >  patch#1335503 - ORA-1013 on queries when sqlnet.expire_time > 0. 8 > >  $ sqlplus scott/tiger@ots16d ----> dedicated server > >  SQL> @test  > >  15:39:41 --> start time0 > >  15:40:16 --> end of query with where clause > >  count(*) --> 458752 > >  15:40:29 --> end timeK > >  Total elapsed time : 35 seconds for the first part, 13 seconds for theI > >  second part > >  Now with MTS:4 > >  $ sqlplus scott/tiger@ots16 ----> shared server > >  SQL> @testi > >  15:45:26 --> start time0 > >  15:47:34 --> end of query with where clause > >  count(*) --> 458752 > >  15:47:47 --> end timeL > >  Total elapsed time : 128 seconds for the first part, 13 seconds for the > >  second partG > >  Conclusion: MTS with V816 is better performing than with V7344 but  > > elapsed : > >  time differs a lot from dedicated server connections.2 > >  Above needs to be transferred to development. > >. > >0 > >  27-SEP-00 17:06:13r > >gG > >  Advised customer to not use MTS where elapsed time is impacted and8 > > databaseL > >  links are accessed. Customer will setup prespawned server (although notJ > >  error free) this evening. AGreed on assisting in tuning such prespawn > >  environment when needed...  > >t > >l > >  27-SEP-00 17:33:19d > >t > >p > >n > >  04-OCT-00 12:47:50n > > F > >  Awaiting the results of the case as distributed to development... > >  > >l > >  12-OCT-00 13:02:28o > > D > >  Problem is reproduced by development and under investigation... > >n > >  > >  13-OCT-00 15:06:49S > >x$ > >  Discussed status with customer:L > >  - Problem is under control as a mix is made with both MTS and dedicated3 > >  (for the long running jobs) server connectionsmJ > >  - Explained that I do not expect a solution on short term based on myH > >  contacts with development. As the problem also reproduces with V816J > >  I expect the problem to be fixed in a future release (and if possible > >  wil be backported)gL > >  - Customer stated that a move will be made to Unix within the next half > > year > >  === Agreed ===.J > >  We decided to close the problem for now as a 'working' environment is	 > > setup G > >  now. Will update this TAR with the 'internal' developments on thise > > matter andA > >  therefore leave status at 14 'Partner (=development) Action' > >  > >l > >( > >s > > Kind regards,m > >f
 > > Ivan Saeza > > i.a.saez.scheihing@tue.nlg > >. > > Malcolm Dunnett wrote: > >t > >  > >n > > >l@ > > >   I'm running Oracle 8.1.6 on a VMS 7.2 system. I've triedB > > > setting up the MultiThreaded server on this system. It works@ > > > but the performance with my programs is terrible ( queriesJ > > > that take 1 second with a dedicated server process might take 5 or 6@ > > > with the Multithreaded server, for example ) [ ie all I doD > > > to get the performance difference is change the client connectB > > > to include a "SERVER=DEDICATED" parameter]. The programs are@ > > > written using Oracle7 OCI calls, they fetch single rows of@ > > > data at a time from the database - I realize they could beD > > > more efficient using array fetch, but it's never been an issue& > > > with dedicated server processes. > > > F > > >   I realize this is a pretty broad question without more detail,E > > > but does anyone have similar experiences ( or alternately, goodcF > > > experiences with the MTS on a VMS system ). Any ideas if there's- > > > anything I could do to make MTS useful?t > > >p > > >rN > ============================================================================ > = @ > > > Malcolm Dunnett      Malaspina University-College   Email: > dunnett@mala.bc.caN > > > Information Systems  Nanaimo, B.C. CANADA V9R 5S5     Tel: (250)755-8738   ------------------------------  # Date: Mon, 30 Oct 2000 18:59:40 GMT * From: Alan E. Feldman <alan48@my-deja.com>= Subject: Questions about SHOW MEM/SLOTS and swapped processeso( Message-ID: <8tkgen$3g$1@nnrp1.deja.com>   Hello,  E We have a test system with very little memory (8Mb). It is a MicroVAXiB 3100-80 running VMS V6.1. Here is a typical SHOW MEM/SLOTS output:    C Slot Usage (slots):       Total        Free    Resident     Swapped-C   Process Entry Slots       100          75          16           9uC   Balance Set Slots          90          74          14           2     
 Questions:  D 1. What is the difference between "Process Entry Slots" and "Balance Set Slots"?g  F 2. I have never noticed the Resident column having anything other thanC two more process entry slots than balance entry slots (16-14=2). Isi( this always the case and why or why not?  G 3. In the Swapped column there are 9 process entry slots with 2 of theme@ also being balance set slots. If I do SHOW SYSTEM, I would see 9E processes swapped out. What is different about the 2 of them that are  also balance set slot holders?  D 4. AUTOGEN seems to always provide 10% fewer total balance set slots than process entry slots. Why?   TIA    -- Disclaimer: JMHO Alan E. Feldmanr alan48  &-)a dellnet.comi    & Sent via Deja.com http://www.deja.com/ Before you buy.    ------------------------------  # Date: Mon, 30 Oct 2000 22:42:32 GMT = From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)lA Subject: Re: Questions about SHOW MEM/SLOTS and swapped processess0 Message-ID: <009F2621.3E8FE8FA@SendSpamHere.ORG>  U In article <8tkgen$3g$1@nnrp1.deja.com>, Alan E. Feldman <alan48@my-deja.com> writes:  >Hello,  > F >We have a test system with very little memory (8Mb). It is a MicroVAXC >3100-80 running VMS V6.1. Here is a typical SHOW MEM/SLOTS output:  >- >SD >Slot Usage (slots):       Total        Free    Resident     SwappedD >  Process Entry Slots       100          75          16           9D >  Balance Set Slots          90          74          14           2 >l >  >Questions:  >oE >1. What is the difference between "Process Entry Slots" and "Balance/ >Set Slots"?   In simple terms:  H "Process Entry Slots" refers to the maximum number of processes you have! configured your machine to allow.u  F "Balance Set Slots" refers to memory used to house the process header.  oG On the VAX, the size of the memory needed for a Balance slot depends onyI some fixed amount of space every process header must have and then a var-nF iable about based upon the VIRTUALPAGECNT of the system.  This is the I maximum size of the process address space and the process header is where I the page table for this address space are housed.  There is a computationoH in the back paged of the OpenVMS Internals and Data Structures manual to$ compute the memory that is consumed.  G >2. I have never noticed the Resident column having anything other thanhD >two more process entry slots than balance entry slots (16-14=2). Is) >this always the case and why or why not?h   No.  It's not always the case.  G The 2 comes from the SWAPPER and NULL processes which are processes but 2 do not have headers carved from the balance slots.    H >3. In the Swapped column there are 9 process entry slots with 2 of themA >also being balance set slots. If I do SHOW SYSTEM, I would see 9-F >processes swapped out. What is different about the 2 of them that are >also balance set slot holders?u  9 A process can be swapped out but its header may not be.       E >4. AUTOGEN seems to always provide 10% fewer total balance set slotsy >than process entry slots. Why?    Because it's AUTOGEN.e   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMe            gO city, n., 1. a place where trees are cut down and streets are named after them.1   ------------------------------  % Date: Mon, 30 Oct 2000 20:34:53 +0100s= From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com>u" Subject: Re: Rdb on VMS reference?) Message-ID: <39FDCD5D.664C63D6@gtech.com>o   Tim Llewellyn wrote: > Arne Vajhj wrote: > > Tim Llewellyn wrote:N > > > Hmmm, just why, if you had a pucker Rdb for VMS, would you want to run a > > > free,p > > > unsupported one on NT? > >e > > Training/playing-around. > L > I guess  so as as long as the NT version is feature (and bug!) compatible.# > I don't know if this is the case.   : Feature-wise I think it is ahead of VMS (it is version 8).  D There are probably many more bugs, but for many training purposes it does not mean much.t   Arne   ------------------------------    Date: 30 Oct 2000 13:49:19 -0500* From: young_r@eisner.decus.org (Rob Young)' Subject: Re: Sun "uptime" belly laughers+ Message-ID: <Om7JQxlvxdjQ@eisner.decus.org>h   In article <rdeininger-3010001223300001@user-2ive7bp.dialup.mindspring.com>, rdeininger@mindspring.com (Robert Deininger) writes:i} > In article <910612C07BCAD1119AF40000F86AF0D805284949@kaoexc3.kao.cpqcorp.net>, "Main, Kerry" <Kerry.Main@compaq.com> wrote:  > J >> Oops, maybe not. Let me explain a little slower and in more detail. :-) >>I >> When normal users go home at night or finish a shift or disconnect fr=  >> om C >> their modem, they log out. This means their current sessions arei< >> disconnected from whatever systems they are connected to. >>   >> Do we agree on this?  >  > K > No!  No matter how ofter or how slowly you say it, it isn't the case thateO > "normal" users always log out.  All those windows on our screens have lots ofeN > context information, and re-creating everything is a waste of time.  I closeQ > a window when the work it's supporting is done.  If something gets interrupted,p# > it might last weeks on my screen.r > N > If VMS really wants to brag about "no user impact", you need to dust off the > virtual terminals, make them applicable to DECterm windows, SET HOST, telnet, etc, and then make them so they can be saved indefinitely when a system shutdownK > occurs.  There would still be a small cost to the user, who would have to2G > log everything in again, but at least the old context would be there.o > L > A lot of the DECwindows apps have been abandoned, and one feature they areQ > missing is the "New Desktop" ability to remember their context across sessions.F > M > You seem to think in terms of transactions or web pages or whatever.  Thinko > of an interactive session as a "transaction" that lasts several weeks or more, and you will see that even VMS clusters have an availability gap when systems* > need to be rebooted for whatever reason. >     = 	I think in the context you describe you are talking about a r  	cluster of workstations, right?  ? 	What Kerry and Andrew are describing is large back-end serversi7 	supporting client server applications or transactions.   ? 	Andrew can say they have similar functionaility using Oracle.	e> 	You can DENY new connections and timeout existing connections/ 	using CLIENT_TIMEOUT, then shut the node down.t  < 	Kerry of course is pointing out the general availability of= 	this feature and the strength of a shared disk architecture. = 	But Andrew would point out that the node serving the storagea; 	can also be failed over in a Sun cluster.  However, Andrew < 	always has Oracle lurking behind the scenes whereas the VMSA 	answer will only get better with Galaxy locks , Galaxy APIs, andn= 	allow one to roll their own so to speak without forking over  	a boat load of cash to Oracle.o  > 	But Andrew would respond they could serve the application and@ 	fail it over to another node in about a minute.  Maybe a minute3 	is good enough for most people.  And so it goes.  n  8 	But here is how it works.  SET LOGIN/INTERACTIVE=0, theA 	cluster service is not longer being served from that node.  Nexth> 	day you find 1 user on that node (they didn't timeout as thatF 	isn't turned on for them) and you whack them by hand.  Every one else< 	is in other nodes.  And we can say this breaks down rarely.> 	Use generic batch queues and submit to the generic queue and 3 	close the execution queues.  Perhaps you have thato< 	rare report that runs and you have to whack it.  But if you; 	have good stuff , it simply gets restarted on another nodeu? 	and picks up where it left off.  Bottom line:  no one effectedoA 	by that single node being shut down.  In case you are wondering,  	this is a db type application.b  < 	And yes, eBay is a very big mess and folks are taking notes 	all over the place:   http://www.auctionguild.com/         ebaY Outages 10-12 Oct 00R     ebaY Outages 13-15 Oct 00o     ebaY Outages 17 -18 Oct 00     ebaY Outages 19 Oct 00     ebaY Outages 22 Oct 00     ebaY Outages 23-25 Oct 00:     ebaY Outages 26, 27 Oct 00  2 	My favorite recent outage is Sunday October 15th:  K ebaY crashed at 17:13:07 PT in an outage that lasted until approx 21:05:20.mG When the ebaY Announcement Board came back up there was a post that wasaN date/time stamped 17:00:49 PT saying that they would be unavailable from 17:10; PT to 17:20 PT while they perform preventative maintenance.   N Unfortunately for ebaY, TAGs editor was reading/refreshing the AB when it wentN down, andcno such post was posted at the time it wentcdown. So ebaY posted theJ 17:10 PT post AFTER THE FACT . Several other posters on two of the OAI/OTI0 boards verified this obfuscation on ebaYs part.   D At 17:22:17 PT ebaY posted that the site was undergoing preventativeO maintenance until 17:30 PT. At 17:37:37 PT they posted the site was still down.DM ebaY posted at 18:06:17 PT that ebaY engineers took the site down at 17:10 PTTM to perform some system tuning on one of ebaY's database servers. The downtime N was expected to last 10 minutes but problems led to them deciding to fail overN to their so called hot backup system. ebaY claims the site was up at 17:50 PT,& but users continued to post problems.   A At 18:18:11 PT search was not working and users got Network error O Unable to request URL from host cgi3.ebay.netscape.com:80: Operation timed out,fL posts. At 18:33 my ebaY went down. At 18:34 Network error Unable to read URLI from host cgi6.ebay.com: Connection reset by peer, messages showed up. AtSK 20:37:07 PT users could not bid, or open item pages. At 21:05:29PT they got L Archived Items Host is Currently unavailable, messages. The site appeared toM stabilize at that point. TAG's take on this is that if ebaY really has a backtN up system, and they had to take the site down on short notice, why not defaultJ to the back up site first rather than letting hundreds of users lose largeK amounts of money while they mucked up the site during the highest use prime K time part of the week. To then get caught in a blatant lie, adds nothing to H their credibility. To add insult to injury, there were also no offers of refunds or auction extensions.    : 	Gee... I wonder if Scotty has eBay in mind when he chants 	uptime uptime uptime.     				Rob)   ------------------------------  % Date: Mon, 30 Oct 2000 21:51:23 -0500   From: John Santos <JOHN@egh.com>' Subject: Re: Sun "uptime" belly laugher=6 Message-ID: <1001030211839.13148A-100000@Ives.egh.com>  , On Mon, 30 Oct 2000, Robert Deininger wrote:  } > In article <910612C07BCAD1119AF40000F86AF0D805284949@kaoexc3.kao.cpqcorp.net>, "Main, Kerry" <Kerry.Main@compaq.com> wrote:L > K > > Oops, maybe not. Let me explain a little slower and in more detail. :-)  > >sJ > > When normal users go home at night or finish a shift or disconnect fr= > > omD > > their modem, they log out. This means their current sessions are= > > disconnected from whatever systems they are connected to., > >  > > Do we agree on this? >  > K > No!  No matter how ofter or how slowly you say it, it isn't the case that4O > "normal" users always log out.  All those windows on our screens have lots of8N > context information, and re-creating everything is a waste of time.  I closeD > a window when the work it's supporting is done.  If something gets interrupted,# > it might last weeks on my screen.:   Define "normal user".g  N > If VMS really wants to brag about "no user impact", you need to dust off theO > virtual terminals, make them applicable to DECterm windows, SET HOST, telnet,2I  etc, and then make them so they can be saved indefinitely when a system r shutdownK > occurs.  There would still be a small cost to the user, who would have totG > log everything in again, but at least the old context would be there.h  J VIRTUAL TERMINALS DO WORK WITH TELNET!  I'm not sure about DECterm or SET  HOST.e  B But this isn't relevant.  Programmers may be logged in on terminalD sessions continuously, but USERS are not.  There may be USERS on forD 24 hours, 7 days a week, but each individual user goes home when theH workday is done, and when they do, they log out.  This is essential for > accountability.  Otherwise you can't really tell who did what.I (The system model here is a multi-user app running on a "single" machine.tF To the users, the cluster looks like a single machine, even if part ofF it is on one side of the San Andreas fault and part of it is a hundred miles away on the other side.)  D The other system model is a networked server.  No one is logged intoG it (except maybe a system manager or maintenance programmer.)  Instead,sC it is listening for network connections, when it gets a connection,p@ it processes a request (or set of requests), and then closes theC connection.  In this model, the scheduled downtime of a single node2B as Kerry is discussing is trivial, since ALL the connections go toC other nodes in the cluster as soon as you disable the node you want.E to reboot.  The average delay until shutdown is just the average lifetD of a connection, which is probably measured in seconds for HTTP, butF may be minutes to hours, or even until 5PM local time for the clients,D where the client may be on another continent.  The point is that theB time arrives, not to distant in the future, and predictably if youC understand the application, when you can shut down the node withoutn impacting any users.  L > A lot of the DECwindows apps have been abandoned, and one feature they areH > missing is the "New Desktop" ability to remember their context across 	 sessions.h  B This is a completely different problem, and if you think about it,? almost insoluble.  You want to remember precise process contexth@ including pending I/O, etc. across a system shutdown?  You could? roll everything out to disk (checkpoint) and then restore after ; reboot, but you would have to make sure the entire externalw> universe didn't give up on you and go on without you while youC are off-line.  This defeats the entire point of clustering, because A for example, the other systems would have to wait on any database = records you had locked until you came back and released them.e= This could be hours.  (If the other systems didn't honor yourl2 locks, then your app would get blown away anyway.)  > Individual programs can be written to record context (e.g. RMS= journaling or TPU/EVE keystroke journalling, but the OS can'to0 do it because it doesn't know what is important.  M > You seem to think in terms of transactions or web pages or whatever.  ThinkmH > of an interactive session as a "transaction" that lasts several weeks I or more, and you will see that even VMS clusters have an availability gapr when systems* > need to be rebooted for whatever reason.  G That's what production servers do is process transactions.  With a tiny E bit of warning and non-idiot users, clusters let users of interactiveeE sessions swap over to another node trivially, by just logging out andiD back in again (you do know about "recall/output" and "recall/input",G don't you?) but of course Kerry is talking about transaction processing  systems.   > J > But VMS reboots are rare, and can usually be planned in advance.  UnlikeN > that SSlithering SSun SSystem called SSolariSS, which is rapidly making ebayK > into a joke.  It's been weeks since everything at ebay worked at the samef time. I > I'll gladly take an occasional few minutes of window-shuffling instead.= >  > -- = > Robert Deininger > rdeininger@mindspring.comU >  >    -- g John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Mon, 30 Oct 2000 22:04:28 -0500-  From: John Santos <JOHN@egh.com>' Subject: RE: Sun "uptime" belly laugher@6 Message-ID: <1001030215309.13148B-100000@Ives.egh.com>  & On 30 Oct 2000, Bill Gunshannon wrote:  P > In article <910612C07BCAD1119AF40000F86AF0D805284949@kaoexc3.kao.cpqcorp.net>,0 >  "Main, Kerry" <Kerry.Main@compaq.com> writes: > |>L > |> Oops, maybe not. Let me explain a little slower and in more detail. :-) > |> RL > |> When normal users go home at night or finish a shift or disconnect fromE > |> their modem, they log out. This means their current sessions are > > |> disconnected from whatever systems they are connected to. > |> s > |> Do we agree on this?  > D > No, that is not always the case.  And wether or not it is would beB > irrelevant tot he discussion as it would apply equally to either  > OpenVMS, Unix or any other OS. >  > |> oQ > |> While there are exceptions where some users staying logged into a system fore9 > |> days, most shops would say this is a security risk. u > M > We have only your claim of this.  The fact is, you can only make this claimlK > for your shop.  I consider people leaving sessions logged on no more of atO > security risk than the user walking down to the coffe-machine without logging K > off.  That kind of security is not overly important in mine and I am surepJ > many other shops environments.  I have users who stay logged on from one > system boot to the next.  G All my customers ("Real World") feel the same way.  But they have userscG who are not programmers and who don't know the SYSTEM password, and who D work at desks or in cubicals, not in the machine room or in lockableF offices.  Sessions left logged in overnight are both a security and anH accountability risk.  Leaving a terminal or workstation un-attended even! while going to lunch is an issue.t   > O > |>                                                      Especially since most N > |> peoples screensaver password is less than 3-4 characters and almost never > |> (ok, never) changes.  > O > Irrelevant.  screensaver passwords on any of my systems are the same as loginhJ > passwords.  I am sure the same is true for most anything beyond Windows.  + That's true on every system I've ever used.    > Q > |>                       So, this is why inactive connections after x hours are O > |> either killed by the system (Hitman stuff) or the application itself timese, > |> out by design and drops the connection. > K > I do not run an idle process killer, nor am I aware of any applications InN > support that would time the user out.  The real world may be a bit different > than your perception of it.c  F You don't use apps that I have written for commercial customers, then.H They all have built-in timeouts, generally 20 minutes to 2 hours (though  I think 2 hours is excessive...)   >  > |> aH > |> So, lets be clear - this is normal practice, not something special. > I > N, this is normal practice where you are.  It is not necessarily normaln > practice everywhere.  F Actually, it is probably *NOT* normal practice where Kerry is (Compaq,D in a software development environment?), but it probably *IS* normal! practice at his customer's sites.e   >  > |> fP > |> Assuming availability is a concern and OpenVMS Clusters are in place - once > |> again:  > |> aP > |> - system planned to be shutdown in the next 8-16 hours (whatever - could be? > |> longer if long batch job running) has its logins disabled.aP > |>- Current user sessions are not impacted. When done, they logout and go home) > |> (or whatever) assuming normal users. M > |> - Since the system to be shutdown has had its logins disabled, the TCPIP N > |> load broker updates the 8.1.2 BIND compliant DNS - (even a Sun box) to noQ > |> longer include that system in the list of available systems participating in K > |> the DNS TCPIP cluster alias. (I think this also works with TCPware and # > |> Multinet, but need to confirm)m; > |> - all new sessions go to other systems in the cluster.uP > |> - when all sessions have moved to the other systems and there are no longerL > |> any active sessions on the node to be shutdown, then that system can be1 > |> shutdown. No user notices, no fuss, no pain.yP > |> - when that system is rebooted, logins are re-enabled. The load broker seesP > |> the logins re-enabled, updates the DNS with a "this system is again part ofP > |> the TCPIP Cluster alias" record update message and that system again starts: > |> to participate in the dynamic DNS load sharing again. > O > So, tell me again, if we were to assume that users must log off at the end of-O > the day and idle processes are killed (this can be done on systems other thandO > VMS as well) how does this differ from say turning off logins on a Solaris orQP > BSD box, waiting for everyone to log off, making the upgrade and rebooting the5 > machine, thus re-allowing users on that box again??e  I Transparency.  The users never know it happened.  This is because all the9D data is accessible from any node and any node can be removed without affecting the accessibility.   >  > |> aP > |> Some Customers today tend to do things like disabling logins on the node toN > |> be shutdown before they go home at night so in the morning, they can shut > |> that system down. > @ > And how does this differ from doing the exact same thing on a  non-OpenVMS system?? >  > |>   > |> No user impact. > |> w7 > |> No active users sessions are disconnected. Period.  > |> a > |> No middleware.  > 6 > And not unique to OpenVMS.  Now I'm really confused. > M > I mean, far be it from me to defend Solaris, but nothing you have describedwN > is unique to OpenVMS.  If the same social rules are forced on the users of a- > UNIX cluster, then the same result applies.l > I > But then, your original premise was wrong and therefore the result willm* > probably not be what you claimed either. >  > bill >  > -- tL > Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesF > bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. > University of Scranton   |C > Scranton, Pennsylvania   |         #include <std.disclaimer.h>   e >  >    -- s John Santosa Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Mon, 30 Oct 2000 22:31:04 -0500   From: John Santos <JOHN@egh.com>' Subject: Re: Sun "uptime" belly laugher 6 Message-ID: <1001030220949.13148C-100000@Ives.egh.com>  + On Mon, 30 Oct 2000, andrew harrison wrote:e   > "Main, Kerry" wrote: > >  > > Andrew,s > > J > > >>> Quite so Kerrys point about people being concerned about scheduledM > > downtime is not solved by OpenVMS by itself because upgrading a node willWJ > > require the people connected to that node either being disconnected or > > logging out themselves. >>>n > > , > > Hey, I think its starting to sink in :-) > > 7 > Kerry do you have a clue what you are talking about ?i > M > > >>> This has always been my point, Kerrys assertion that nodes in OpenVMSr7 > > clusters can be rebooted with no impact ignores the / > > sessions connected to the rebooted node.>>>r > > K > > Oops, maybe not. Let me explain a little slower and in more detail. :-)a > > ; > Try not to be patronising when it is clear that you don't # > know what you are talking about ?e > K > > When normal users go home at night or finish a shift or disconnect fromaD > > their modem, they log out. This means their current sessions are= > > disconnected from whatever systems they are connected to.d > >  > > Do we agree on this? > >  > Yes assuming that:D > A	They are connected users in the terms that you use, many systemsB > 	don't have connected users they have applications that connect , > 	and remain connected to a backend system.  D It is trivial to make the front-end apps disconnect and reconnect toA the server, either on receipt of a signal or periodically.  It iseB especially easy to signal things like this on a VMS cluster, usingC the lock manager.  Total time to disconnect and reconnect (probablye' to a different server) is milliseconds.e  @ > B	You arn't running a system thats hosting users from multiple? > 	time zones. Many of the systems I have worked on have users  < > 	in Tokyo, London and New York and for these systems there2 > 	is no quiet time, there is no night and no day.  A Irrelevent to Kerry's argument.  The system may run 24 hours, butbC the users in Tokyo will eventually go home, and when they come backlB to work the next day, they will connect to another node.  After 24A hours (or 16 hours, if you disable logins at the right time, just1A before a large group of users goes offline for the day), everyone0B will be off the node.  So it is 16-24 hours instead of 12-24 hoursC between disabling logins and the time it is okay to shut the system  down.  So what?     P > > While there are exceptions where some users staying logged into a system forM > > days, most shops would say this is a security risk. Especially since mostwM > > peoples screensaver password is less than 3-4 characters and almost nevermO > > (ok, never) changes. So, this is why inactive connections after x hours are N > > either killed by the system (Hitman stuff) or the application itself times+ > > out by design and drops the connection.S > >  > G > You seem to think that users connect directly to backend systems, oddY > thats C > what used to happen but the majority of big backend systems that eB > I work on don't have any users (except the DBA's and the admins)B > the users connect to mid tier servers. Perhaps this is where you > are getting confused.T  B What makes you think that?  It doesn't matter whether there is oneA tier, two tiers, or three tiers, or whether the tiers are runningmC on the same cluster or different clusters or some tiers are runningaC on an entirely different architecture.  Kerry's "transparent reboots? and upgrade" scheme works fine on any level implemented on VMS.   ? > I won't bother responding to indevidual points in the rest ofO= > you posting except to say that it is all scheduled downtime3? > and the you can also, particularly if you are allowed to kicke@ > users/connected apps off a node do this with any cluster that  > I can think of.n  A Kerry *DOESN'T* kick anyone off any node!  He just waits for themy. to go away naturally.  That's the whole point!  A > The fact that you don't seem to realise that this is scheduled aC > downtime and the fact that you seem to think that the capabilites,F > that you talk about are only available for OpenVMS illustrates that 3 > you really should not be posting on this subject.k  A You persist in confusing node downtime with application downtime.T  = > Let me ask you a question, why do you think that people areb> > asking for 99.9XXXX% uptime ? Could it not have something to? > do with the fact that many systems have active users 24 hourse> > a day. For this kind of system where is your window when youE > can do maintenance. There isn't one hence the 99.9XXX% requirement.0  D Which is precisely where the virtues of VMS clusters shine!  This is@ a splendid arguement *FOR* implementing all such systems on VMS.  8 [snip all Kerry's points that Andrew studiously ignored]   > -- 2 > Andrew Harrison8 > Enterprise IT Architectc     -- t John Santosr Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Mon, 30 Oct 2000 22:55:18 -0500 2 From: rdeininger@mindspring.com (Robert Deininger)' Subject: Re: Sun "uptime" belly laughereL Message-ID: <rdeininger-3010002255190001@user-2ivebbg.dialup.mindspring.com>  W In article <Om7JQxlvxdjQ@eisner.decus.org>, young_r@eisner.decus.org (Rob Young) wrote:     F >         I think in the context you describe you are talking about a ) >         cluster of workstations, right?   M Pretty much the case.  In this environment, the workstations ARE the servers.s Or is it the other way 'round?  E >         And yes, eBay is a very big mess and folks are taking notesp >         all over the place:m >  > http://www.auctionguild.com/ >  >  >     ebaY Outages 10-12 Oct 00n >     ebaY Outages 13-15 Oct 00o  >     ebaY Outages 17 -18 Oct 00 >     ebaY Outages 19 Oct 00 >     ebaY Outages 22 Oct 00 >     ebaY Outages 23-25 Oct 00b  >     ebaY Outages 26, 27 Oct 00 > ; >         My favorite recent outage is Sunday October 15th:   I Interesting.  I have certainly noticed on my own that ebay has been lyingtK about their down time.  Most often, they don't make an announcement at all.a   -- C Robert Deininger rdeininger@mindspring.comn   ------------------------------  % Date: Mon, 30 Oct 2000 23:10:22 -0500E2 From: rdeininger@mindspring.com (Robert Deininger)' Subject: Re: Sun "uptime" belly laughersL Message-ID: <rdeininger-3010002310230001@user-2ivebbg.dialup.mindspring.com>  X In article <1001030211839.13148A-100000@Ives.egh.com>, John Santos <JOHN@egh.com> wrote:    L > VIRTUAL TERMINALS DO WORK WITH TELNET!  I'm not sure about DECterm or SET  > HOST.   I Hmm.  I'll have to look into this again.  I'm sure it was a no-no when I f
 last checked.f   D > But this isn't relevant.  Programmers may be logged in on terminalF > sessions continuously, but USERS are not.  There may be USERS on forF > 24 hours, 7 days a week, but each individual user goes home when the4 > workday is done, and when they do, they log out.    F Piffle!  I'm a programmer.  I'm a user.  I DO go home occasionally.  I never log out.   > This is essential for @ > accountability.  Otherwise you can't really tell who did what.  J Ummm.  I don't follow you.  Unless you only allow 1 user at a time, how doJ enforced logouts improve accountability?  Is this kind of like the network) security goons who just unplug the wires?e  K > (The system model here is a multi-user app running on a "single" machine.m   Not my system model...  D > This is a completely different problem, and if you think about it,A > almost insoluble.  You want to remember precise process contexta9 > including pending I/O, etc. across a system shutdown?  n  J It seems you can do quite a bit without remembering nearly as much.  If anM app is "quiet" at the time of the shutdown, it could certainly remember a loteH of it's context.  Let's take a crusty old example just for discussion --L Bookreader.  If I'm in the middle of something, I might have 2 dozen windowsopen in 5 or 6 different books.  Maybe a search or two in progress.  If the system has to be shut down, it would be a lot less work for the computer to remember all this, instead of making me recreate it.  Bookreader is perhaps an easy example, since it's mostly read-only.r    @ > Individual programs can be written to record context (e.g. RMS? > journaling or TPU/EVE keystroke journalling, but the OS can't 2 > do it because it doesn't know what is important.  K Right.  The environment has to provide enough hooks.  The application knowskN what's important and saves it when it gets a shutdown message.  There's a hintA of this in the New Desktop, but the implementation is ... sparse.o  A > ... but of course Kerry is talking about transaction processingm
 > systems.   Maybe Kerry has the old Digital sickness.  He's assuming he knows the customer's needs, even when they tell him something quite different.  I don't fit into his: favorite mold, and no amount of scolding will make me fit.   -- h Robert Deininger rdeininger@mindspring.comr   ------------------------------  % Date: Mon, 30 Oct 2000 23:40:53 -0500h  From: John Santos <JOHN@egh.com>' Subject: Re: Sun "uptime" belly laughert6 Message-ID: <1001030231950.13148A-100000@Ives.egh.com>  , On Mon, 30 Oct 2000, Robert Deininger wrote:  Z > In article <1001030211839.13148A-100000@Ives.egh.com>, John Santos <JOHN@egh.com> wrote: >  > N > > VIRTUAL TERMINALS DO WORK WITH TELNET!  I'm not sure about DECterm or SET 	 > > HOST.d > K > Hmm.  I'll have to look into this again.  I'm sure it was a no-no when I y > last checked.n >  jF > > But this isn't relevant.  Programmers may be logged in on terminalH > > sessions continuously, but USERS are not.  There may be USERS on forH > > 24 hours, 7 days a week, but each individual user goes home when the6 > > workday is done, and when they do, they log out.   > H > Piffle!  I'm a programmer.  I'm a user.  I DO go home occasionally.  I > never log out.   Aha!  I suspected as much...  E This transparent cluster node shutdown business that Kerry and Andrew1G have been arguing about pertains to production servers, not developmentnE systems.  Presumably, programmers know how to save their work and hownF to log in and out at will...  (I never log out, either, though my workC station, which I use mostly as a glorified multi-session VT with an F oversized screen, does lock itself if idle for more than a half hour.)  A A business system being used by non-programmers running a captives. application is an entirely different matter.     >  > > This is essential for B > > accountability.  Otherwise you can't really tell who did what. > L > Ummm.  I don't follow you.  Unless you only allow 1 user at a time, how doL > enforced logouts improve accountability?  Is this kind of like the network+ > security goons who just unplug the wires?i  F It prevents the night shift or the janitor or someone else with accessE to the building from, e.g. ordering 50 sheets of plywood and charginguC it to the company and making it look like some random user who wentnG home at 5PM did it.  (Or changing billing records or engaging in random D data destruction.)  The general rule is each user's transactions canB be traced back to that user, and no one else.  Each user must have? their own account and password (no sharing), and can't leave ant= un-attended, logged-in terminal or workstation sitting aroundeB available for someone else to use.  Idle process timeouts are part of enforcing this.  D This is probably completely irrelevent to a programming environment,; except when you have untrusted programmers (i.e. students).   M > > (The system model here is a multi-user app running on a "single" machine.o >  > Not my system model...  ( So I should be forced to use your model?   > F > > This is a completely different problem, and if you think about it,C > > almost insoluble.  You want to remember precise process contextr; > > including pending I/O, etc. across a system shutdown?  p > L > It seems you can do quite a bit without remembering nearly as much.  If anO > app is "quiet" at the time of the shutdown, it could certainly remember a loteJ > of it's context.  Let's take a crusty old example just for discussion --N > Bookreader.  If I'm in the middle of something, I might have 2 dozen windows> open in 5 or 6 different books.  Maybe a search or two in progress.  If the system has to be shut down, it would be a lot less work for the computer to remember all this, instead of making me recreate it.  Bookreader is perhaps an easy example, since it's mostly read-only.o > : Yes, but it is the app, not the OS, that needs to do this.  B > > Individual programs can be written to record context (e.g. RMSA > > journaling or TPU/EVE keystroke journalling, but the OS can't 4 > > do it because it doesn't know what is important. > M > Right.  The environment has to provide enough hooks.  The application knowsnP > what's important and saves it when it gets a shutdown message.  There's a hintC > of this in the New Desktop, but the implementation is ... sparse.b > C > > ... but of course Kerry is talking about transaction processingA > > systems. > H > Maybe Kerry has the old Digital sickness.  He's assuming he knows the D customer's needs, even when they tell him something quite different.  A This may not match your need, but it does match the needs of manyn
 customers.     I don't fit into his< > favorite mold, and no amount of scolding will make me fit.  A Nobody is asking you to fit.  I don't have any need for a FORTRAN A compiler in my use of VMS, but that doesn't mean Compaq shouldn'tu bother providing one.o   --   John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Tue, 31 Oct 2000 01:44:17 -0400a- From: JF Mezei <jfmezei.spamnot@videotron.ca> ' Subject: Re: Sun "uptime" belly laugher , Message-ID: <39FE5C2F.FC579C52@videotron.ca>   John Santos wrote:G > This transparent cluster node shutdown business that Kerry and Andrew I > have been arguing about pertains to production servers, not developmenteG > systems.  Presumably, programmers know how to save their work and howsH > to log in and out at will...  (I never log out, either, though my workE > station, which I use mostly as a glorified multi-session VT with an H > oversized screen, does lock itself if idle for more than a half hour.)  N If you plan to shutdown a node "tonight", then you can send a broadcast to allL users connected to that one node telling them about the shutdown, and askingF them to log out at their convenience and log back in , or reboot theirE workstation to be served by a node that won't be shutdown that night.S   ------------------------------  % Date: Mon, 30 Oct 2000 19:00:52 +0000a From: SysAdmin <djesys@fsi.net>  Subject: Re: Sun Moona' Message-ID: <39FDC564.A6F54834@fsi.net>s   Warren Spencer wrote:  > [snip]? > Is there significance to the choice of the name "Full Moon" ?    To me, yes - perhaps.t  H Once upon many moons ago, circa. 1986, I encountered an occasion where aD DIBOL app. had been ported from VMS to UN*X using Software Ireland'sC SIBOL (DIBOL emulator). In experimenting with some UN*X commands tomF locate text strings within an otherwise binary file, I discovered thatE the string "milky white buttocks" appeared not once, but twice within , the SIBOL run-time (pseudocode interpreter).  @ So, if you take a common context within which the term "moon" is sometimes used, ...c  * Well, it *WAS* a UN*X system, after all...   ...or is that reaching too far?    David J. Dachterau   ------------------------------  % Date: Mon, 30 Oct 2000 19:43:00 -0000X. From: shawbrow@delilah.cisco.com (Shawn Brown) Subject: Re: Sun Moon - Message-ID: <972935074.633293@sj-nntpcache-3>m  / In article <svrga6jhcecd14@news.supernews.com>, . Warren Spencer <wspencer@ap.nospam.org> wrote:  m> >Is there significance to the choice of the name "Full Moon" ?   Yes.   It makes a wolf pack howl.   Shawne   ------------------------------  % Date: Mon, 30 Oct 2000 15:02:55 -0500l0 From: Jim Jennis <jjennis@discovery.fuentez.com> Subject: Re: Sun Moon D Message-ID: <3.0.5.32.20001030150255.0090dd00@discovery.fuentez.com>  ( At 06:43 PM 10/30/2000 -0000, you wrote: >  >n >-- > >Is there significance to the choice of the name "Full Moon" ?  C Possible reference to a "highly available clustering" of bare butts0I sometimes displayed as the result of sopho-moronic behavior at fraternity K parties in this country???....hopefully this is not the image Sun is tryingd) to present to its customers ...or is it??i   :-)e   Regards,   Jimo8 --------------------------------------------------------7 FSC - Building Better Information Technology Solutions->7       from the Production Floor to the Customer's Door. 8 --------------------------------------------------------5 Jim Jennis, Technical Director for Commercial Systemss Fuentez Systems Concepts, Inc. 1 Discovery Place, Suite 2 Martinsburg, WV. 25401 USA   # Phone: +001 (304) 263-0163 ext. 235n Fax:   +001 (304) 263-0702% Email: jjennis@discovery.fuentez.com          jhjennis@shentel.net & WEB: http://www.discovery.fuentez.com/   ------------------------------    Date: 30 Oct 2000 23:18:36 -0500* From: young_r@eisner.decus.org (Rob Young)0 Subject: Sun's haunting ecache problems detailed+ Message-ID: <6whLsdl5xMmb@eisner.decus.org>a  ; 	The chickens are coming home to roost in a mighty big way.o  1 	We got client names adding credence to all this.i 	tD 	Seems Andrew has been spinning us but won't spin this one.  Bet me." 	"What a wicked web we weave", eh?  6 http://www.forbes.com/forbes/2000/1113/6613068a_3.html  G "Sun says it has finally figured out what's wrong. It is an odd problemnN involving stray cosmic rays and memory chips in the flagship Enterprise serverO line, whose models are priced at $50,000 to more than $1 million. Yet Sun won't N fix all of the servers it has sold; instead it will make repairs when it deems them necessary."    B 	Note to FEs... the angle here is "repair when necessary."  Please> 	check field manual special bulletin number 79 which states in 	part:  8 		1)  "If client claims to have the problem , tell them / 		     that their datacenter feels a bit warm."a 	 7 		2)  "If they are still whining after the fourth crashh6 		     in two weeks, probe to see if they have had the( 		     datacenter power checked lately."  5 		3)  "If they are still a pissin' and a moanin' tell A 		     them they are next right after other important customers."e    M "Last November Verisign Global Registry Services, a domain name registry, was O down for two hours after a crucial Sun box crashed. Verisign complained but gotmL no explanation. Months later an executive at Verisign ran across the Gartner
 bulletin."  M "I said to Sun, 'My God, you knew about this problem, and you didn't tell me?iL That's unconscionable,' " he says.Verisign still uses Sun for some tasks but3 has moved important systems onto IBM Unix servers.	o  9 	Get over it pal... Scotty Uptime doesn't care about your  	problems ya whiner!!!  P "Sun servers lack ECC protection. "Frankly, we just missed it. It's something weL regret at this point," Shoemaker says. Its next high-end servers, based on aN new processor called the UltraSparc III, will have ECC protection; they are to debut in mid-2001."   C 	There you have it folks.  After a number of months the crux of thet> 	problem.  They left out ECC.  Don't you just hate overlooking= 	a design consideration like that?  "It's something we regretk5 	at this point."   A-yup... yup surely do, surely do.w  O "A major telecommunications company endured repeated crashes on Sun servers. ItdN recently committed to buy its next batch of Unix servers from HP. While no oneO has reported a consequential data loss from a crash, one Sun customer claims tomJ have spent $3 million trying to diagnose and fix the problem, according to Gartner Group."     F 	Three mill?  Chump change.  You guys love Sun so much , what's a few  	more million amongst friends?    L "Shoemaker says Sun has spent tens of millions of dollars on the problem andJ put in countless hours individually briefing customers all over the world.L Every Monday, Wednesday and Friday Shoemaker reviews the latest "red alerts"L with Sun President Edward Zander and Chief Executive Scott McNealy to assess the damage."  @ 	Ah, I wouldn't worry Mr. Zander, Mr. Shoemaker and Mr. McNealy.< 	One of your finer employees, Mr. Andrew Harrison assured us< 	in December 1999 that these problems were behind you.  You : 	may wish to drop him an email so he can reassure you.  He< 	can be reached at Andrew.Harrison@uk.sun.com.  How can I be/ 	so confident that this is "old news" and thesew= 	problems are behind you?  Well... Andrew said so in December = 	1999 and labeled me a FUDster once again (like he did today,a 	10/30/00):n  > http://www.deja.com/[ST_rn=ps]/getdoc.xp?AN=560445620&fmt=text  O "This thread looks more and more like a very sorry attempt to throw as much FUDO9 as possible in the vague hope that something will stick."'  A 	Well sorry Andrew.  In actuality, it wasn't FUD in December 19998D 	and it isn't FUD in October 2000.  Sun has serious ecache problems.; 	So it seems your attempts to deflect this are as lame then- 	as they are today.3  O "But Sun's simmering image problem could heat up if a really high-profile crashj  or two occurs in coming months."  B 	Oh... don't worry about that.  Sun's crashing problems are behind> 	them.  I am sure Andrew will be along in a bit to tell us all
 	about it.   	Spin this Andrew!!!   			HA!   				Robh   ------------------------------  + Date: Mon, 30 Oct 2000 23:55:57 -0600 (CST)r From: sms@antinode.org  Subject: TCPIP V5.0A SMTP ACCVIO) Message-ID: <00103023555686@antinode.org>   D    Back around 8-NOV-1999 18:40:12.28, I reported a problem with theG TCPIP V5.0 SMTP (AlphaStation 200 4/233, VMS V7.2) symbiont dying whiletE trying to deliver certain malformed (typically junk) e-mail messages:    ALP $ start TCPIP$SMTP_ALP_01   8 %%%%%%%%%%%  OPCOM   8-NOV-1999 15:49:34.24  %%%%%%%%%%% Message from user SYSTEM on ALP ; -SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual : address=0000000000000000, PC=00000000000712D8, PS=0000001B  D    I'm now up to VMS V7.2-1 and TCPIP V5.0A, and it appears still to@ suffer from the same problem.  Whenever a suitably nasty messageF arrives, the server queue stops, causing messages to accumulate in the> queue, undelivered.  It makes a nice denial-of-service attack.  F    Is there a patch to cure this, or is it known to be better in TCPIP V5.1?:  G    If any responsible person needs a set of *.TCPIP_ALP[_TEXT] files tou> reproduce the failure, I'm building up a fair collection.  The5 work-around is to remove the queue-stopper files from G SYS$SPECIFIC:[TCPIP$SMTP], delete the queue-stopper job from the queue,DF and then restart the queue.  I'm getting good at it, but it's tedious.  H ------------------------------------------------------------------------  C    Steven M. Schweda               (+1) 651-699-9818  (voice, home)bC    382 South Warwick Street        (+1) 763-781-0308  (voice, work)eG    Saint Paul  MN  55105-2547      (+1) 763-781-0309  (facsimile, work)e9    sms@antinode.org                sms@provis.com  (work)'   ------------------------------  % Date: Mon, 30 Oct 2000 17:06:31 -0600t/ From: Chris Scheers <chris@applied-synergy.com>r Subject: Re: TEK 4111 emulator? O Message-ID: <EB30233BB3FC3063.1D545623B6834702.00AC04BF0A9A3B3E@lp.airnews.net>g   Jim Agnew wrote: >  > I think msdos kermit does a tek emulation, and it's is possible to run it under win95... if you want a free solution, however, it'si< > not the best solution... but didn't see kermit mentioned.. >  > Chris Scheers wrote: > >mG > > I am looking for a TEK 4111 emulator for VAX/VMS V7.1 and eXcursionr	 > > V3.0.s > >tL > > The TEK display needs to appear on the PC screen.  Ideally, it should beJ > > addressable as a device from the VAX (although I have some flexibility > > there).i > > % > > Does anyone have any suggestions?e  @ Most of the suggestions I have received are for 4014 emulation. H Unfortunately, I need 4111.  (I think the difference is that the 4111 is color and supports surfaces.)r  G The best lead so far is Teem-X.  I have sent Pericom a request for moret= information, but they don't list VMS as a supported platform.   D Is anyone using Teem-X on VMS?  Can you point me to some information	 about it?    Thanx!  G -----------------------------------------------------------------------e$ Chris Scheers, Applied Synergy, Inc.  C Voice: 817-237-3360            Internet: chris@applied-synergy.com d   Fax: 817-237-3074m   ------------------------------  % Date: Tue, 31 Oct 2000 16:56:02 +1200c. From: Nivlesh Chandra <NChandra001@itc.gov.fj> Subject: TLZ tapesO Message-ID: <791C2856E8FDD211BAFB0008C759919501A629B6@exchange01.govnet.gov.fj>.  K We backup our data on TLZ tapes. But sometimes when I try and recover data,mL I cannot read the label of the tape and thus cannot see what is on the tapesL and do not have any other resort but either reinitialise the tape and use itG again or throw it (if it is of no use anymore) .... either way I end upmI losing data ..is there any way whereby I will be able to read the data on-J the tape even though its label cannot be read... or maybe just rewrite the label of the tape ...??c    $ Your help will be really appreciated   Niv1   ------------------------------  % Date: Mon, 30 Oct 2000 20:20:14 -0500n2 From: norm lastovica <norman.lastovica@oracle.com>, Subject: Tough times at the Sun homestead...* Message-ID: <39FE1E4E.4FA2FB2F@oracle.com>  3 Interesting read...  Here's a portion of an article 4 from Forbes discussing at least one problem with the2 Enterprise server line.  visit the forbes web site for the complete article.t  4 http://www.forbes.com/forbes/2000/1113/6613068a.html  9 THE MYSTERIOUS GLITCH has been popping up since late lastR. year. At a new Web company in San Francisco, a9 telecommunications company in the Midwest, a Baby Bell ind: Atlanta, anInternet domain registry on the East Coast--for0 no apparent reason, high-end servers made by Sun Microsystems suddenly crashed.  8 It happens not to all of them, and not all the time, but< enough to cause problems for America Online, Ebay and dozens9 of other major corporate accounts, baffling Sun engineersm0 who spent months trying to identify the problem.  : Adding to the mystery is Sun's own reticence. It has never: issued a warning to its customers or disclosed the flaw to: new buyers. For months Sun told customers seeking a repair: that they must sign a legal agreement promising to keep it< secret. Many still don't dare speak out. Even now Sun hasn't9 published on its main Web site an official explanation ofa the bug.  . 	<the rest of it is on on the forbes web page> -- d< norm lastovica / oracle rdb engineering / usa / 610.696.4685% reply to: norman.lastovica@oracle.comd   ------------------------------  % Date: Mon, 30 Oct 2000 21:01:33 -0600i7 From: "David J. Dachtera" <djesys.nospam@earthlink.net>i0 Subject: Re: Tough times at the Sun homestead..., Message-ID: <39FE360D.68A4161@earthlink.net>   norm lastovica wrote:n > [snip]7 >         <the rest of it is on on the forbes web page>w   The article goes on to say:c  G "Verisign still uses Sun for some tasks but has moved important systemse onto IBM Unix servers."r  H Micro$hit says, "If you want to sell our software, you can't sell anyone else's".  H Sun says, "If you buy our stuff, you can't tell anyone when it fails for any reason".  G If these examples are indicative of the quality of our colleagues, thendH maybe I *SHOULD* go get that job on the garbage truck. At least there isF no expectation of integrity of one's colleagues in such an occupation.  G At least the bulk of my fellow posters here are respectable people. Thet exceptions know who they are...m   -- o David J. Dachterae dba DJE SystemsI 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.c   ------------------------------  % Date: Mon, 30 Oct 2000 22:38:06 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca>I0 Subject: Re: Tough times at the Sun homestead...+ Message-ID: <39FE308A.A9F01E4@videotron.ca>s   "David J. Dachtera" wrote:I > "Verisign still uses Sun for some tasks but has moved important systems  > onto IBM Unix servers."l  N Who now owns Verisign, and does that owning entity have a tighter relationship with IBM than with Sun ?  J With all the changed relationships in the .DOM arena with companies buyingG others, merging , sleeping, generating offspring and generally having agO promiscuous life, I am not surprised if some companies change allegances often.t  K Look at Amazon which made a lot of noises about going with HP. I think thatoN changed allegiances are common. The trick is for the VMS folks to capture more than they lose.h  J > Sun says, "If you buy our stuff, you can't tell anyone when it fails for > any reason".  N You don't find companies bragging about systems failures or security failures,K except perhaps Microsoft bragging about its software being so sophisticatedt0 that it allows hackers to come n and steal code.   ------------------------------    Date: 30 Oct 2000 10:54:35 -0800 From: old_timer@user.vms, Subject: VAX Mail question - corrupted files( Message-ID: <8tkg5b0hc4@drn.newsguy.com>   Dear VMS Gurus,g  L I work for a large international company which has "outsourced" its computer3 support to yet another large international company.i  M I recently had occasion to experience the level of quality of knowledge which9! results from such an arrangement.   O I am a 20-year end-user of VMS, and have picked up a little knowledge along the.M way.  My question has to do with the answer I finally got (after many initialvO silly and obviously wrong suggestions) from the outsourced "help desk".  Let men start by stating the problem.a  L I have recently encountered trouble sending large text files to a remote VAXL node via VAX mail.  When using VAX mail to send a large text file (either asN in-line text embedded in the message or using the /foreign qualifier to send aK saveset) to a remote node, I occasionally get transmission errors (or, moresP precisely stated, the file arrives corrupted).  This does not happen 100% of the time.h  < Until recently, I had never before encountered this problem.  N The bottom line is, the help desk is telling me that it's a "network bandwidthP problem".  What I would like to know is, is this a reasonable explanation, or am I being taken?  L Here's my reasoning.  I know the network bandwidth going out of and into ourJ site is less than it should be, but wouldn't you think that running out ofM bandwidth would cause very slow transmissions, or even aborted transmissions,s; rather than random errors interspersed throughout the file?   P Besides, the answer "it's a network bandwidth problem" just seems too convenientL to me.  Call me a cynic perhaps.  The company to whom we outsourced the helpN desk has certain performance metrics they are required to meet.  How better toL handle a question they can't answer than by claiming that the physical layerL needs expensive upgrading?  And guess who gets the contract for the upgrade?  = If any of you can shed light on this it would be appreciated.e   ------------------------------  % Date: Mon, 30 Oct 2000 12:18:00 -0800a! From: Shane.F.Smith@Healthnet.com 0 Subject: Re: VAX Mail question - corrupted filesD Message-ID: <OF6909885C.0D040F92-ON88256988.006E531D@foundation.com>  K I've used and supported VMS systems for 15 years or so, both as an in-houseeK support guy and a paid support guy working on multiple customer machines. IwG have used VMS mail on single nodes, clusters, WANs, using DECnet, X.25,IK several TCP/IP implementations, and between everything from machines in thekF same room to machines literally on opposite sides of the world. I haveE NEVER seen VMS mail corrupt anything due to network bandwidth issues..  K It sounds more like a software problem to me. It's almost certainly not VMSdF mail itself, do you have a custom/third party mail transport mechanismJ installed? If so, it may be a bug in that, or maybe the network stack. CanI we have some more details on what you're using please? Version(s) of VMS,oI what sort of network (TCP/IP, DECnet etc), whose implementation (Compaq /a Process software etc).   Shaneu          , old_timer@user.vms on 10/30/2000 10:54:35 AM   To:   Info-VAX@Mvb.Saic.Com  cc:   - Subject:  VAX Mail question - corrupted files-       Dear VMS Gurus,-  C I work for a large international company which has "outsourced" itsm computer3 support to yet another large international company.-  G I recently had occasion to experience the level of quality of knowledger whichc! results from such an arrangement.p  K I am a 20-year end-user of VMS, and have picked up a little knowledge along0 the4E way.  My question has to do with the answer I finally got (after manye initialzG silly and obviously wrong suggestions) from the outsourced "help desk".o Let me start by stating the problem.   H I have recently encountered trouble sending large text files to a remote VAXjI node via VAX mail.  When using VAX mail to send a large text file (eithern asG in-line text embedded in the message or using the /foreign qualifier too send aK saveset) to a remote node, I occasionally get transmission errors (or, moreiI precisely stated, the file arrives corrupted).  This does not happen 100%e of the time.w  < Until recently, I had never before encountered this problem.  D The bottom line is, the help desk is telling me that it's a "network	 bandwidthyJ problem".  What I would like to know is, is this a reasonable explanation, or am  I being taken?  H Here's my reasoning.  I know the network bandwidth going out of and into ourtJ site is less than it should be, but wouldn't you think that running out of> bandwidth would cause very slow transmissions, or even aborted transmissions,; rather than random errors interspersed throughout the file?a  E Besides, the answer "it's a network bandwidth problem" just seems tooa
 convenientG to me.  Call me a cynic perhaps.  The company to whom we outsourced the  helpK desk has certain performance metrics they are required to meet.  How bettery toF handle a question they can't answer than by claiming that the physical layernC needs expensive upgrading?  And guess who gets the contract for thes upgrade?  = If any of you can shed light on this it would be appreciated.u   ------------------------------  % Date: Mon, 30 Oct 2000 16:46:08 -0400a- From: JF Mezei <jfmezei.spamnot@videotron.ca> 0 Subject: Re: VAX Mail question - corrupted files, Message-ID: <39FDDE02.B79B274C@videotron.ca>   old_timer@user.vms wrote:nN > node via VAX mail.  When using VAX mail to send a large text file (either asP > in-line text embedded in the message or using the /foreign qualifier to send aM > saveset) to a remote node, I occasionally get transmission errors (or, more R > precisely stated, the file arrives corrupted).  This does not happen 100% of the > time.x     do a MAIL> SHOW ALL   B See if there is a default transport defined (towards the bottom). K do a SET NOTRANSPORT to remove it. It is possible that they are routing themB mail to the other nodes with something different that DECnet-MAIL.  M I agree with others, a slow link should NOT corrupt the messages or any data.e  K If you are JDoe on nodeA and the other on NodeB, you can try the following:l  < SEND/foreign mybigfile.txt/subject="test" nodeB::NodeA::JDoe  F This should route the message to node B and then back to Node A and be delivered to yourself.  F When you send a big file (or when you specify /foreign) the message isN actually stored as a separate file in your mail directory. Doing a DIR/FULL onJ that newly received file in the mail directory and the original file might reveal something.l    M I think you need to find out exactly how the mail is being sent and over what L it is being routed. If the network allows corrupted mail, god only know what) else is being corrupted during transport.n   ------------------------------    Date: 30 Oct 2000 13:08:56 -0800 From: old_timer@user.vms0 Subject: Re: VAX Mail question - corrupted files) Message-ID: <8tko1801fue@drn.newsguy.com>a   Shane,  K Thanks for the response.  I've tried to provide some additional informatione below.      " Shane.F.Smith@Healthnet.com wrote:   >  >yL >I've used and supported VMS systems for 15 years or so, both as an in-houseL >support guy and a paid support guy working on multiple customer machines. IH >have used VMS mail on single nodes, clusters, WANs, using DECnet, X.25,L >several TCP/IP implementations, and between everything from machines in theG >same room to machines literally on opposite sides of the world. I haveiF >NEVER seen VMS mail corrupt anything due to network bandwidth issues. >OL >It sounds more like a software problem to me. It's almost certainly not VMSG >mail itself, do you have a custom/third party mail transport mechanismn >installed?.  N Like I said, I'm an end-user, not a sys-admin.  But if you tell me how to findH out whether or not we have a custom/third party mail transport mechanismN installed, I've been around here long enough to have made some contacts withinM the organization that would be willing to get the answers for me (despite our ! help desk outsourcing situation).   @ >If so, it may be a bug in that, or maybe the network stack. CanJ >we have some more details on what you're using please? Version(s) of VMS,  J It's VMS 5.5-2 (ancient, I know; but it's never been updated due to legacy software issues).5  + >what sort of network (TCP/IP, DECnet etc),   P I don't know the details of the underlying network architecture.  But I can findK out if you tell me how.  To the end-user, it just looks like DECnet like itl4 always did.  That is,  to send VAX mail I just type:  ( $ mail/foreign myfile.sav node::username  6 >whose implementation (Compaq / Process software etc).  K I don't know.  But I will try to find out.  Is there a command I can issue?0O Our support comes from Compaq (filtered through the third-party outsourced help  desk).   >  >Shane >l >o >u >r >o- >old_timer@user.vms on 10/30/2000 10:54:35 AMm >a >To:   Info-VAX@Mvb.Saic.Com >cc: >u. >Subject:  VAX Mail question - corrupted files >r >  >  >Dear VMS Gurus, >fD >I work for a large international company which has "outsourced" its	 >computer 4 >support to yet another large international company. >sH >I recently had occasion to experience the level of quality of knowledge >which" >results from such an arrangement. >oL >I am a 20-year end-user of VMS, and have picked up a little knowledge along >theF >way.  My question has to do with the answer I finally got (after many >initialH >silly and obviously wrong suggestions) from the outsourced "help desk". >Let me- >start by stating the problem. > I >I have recently encountered trouble sending large text files to a remote  >VAXJ >node via VAX mail.  When using VAX mail to send a large text file (either >as9H >in-line text embedded in the message or using the /foreign qualifier to >send a L >saveset) to a remote node, I occasionally get transmission errors (or, moreJ >precisely stated, the file arrives corrupted).  This does not happen 100% >of theg >time. >2= >Until recently, I had never before encountered this problem.  >iE >The bottom line is, the help desk is telling me that it's a "network'
 >bandwidthK >problem".  What I would like to know is, is this a reasonable explanation,  >or am >I being taken?d >'I >Here's my reasoning.  I know the network bandwidth going out of and intog >ourK >site is less than it should be, but wouldn't you think that running out of ? >bandwidth would cause very slow transmissions, or even aborted  >transmissions,l< >rather than random errors interspersed throughout the file? > F >Besides, the answer "it's a network bandwidth problem" just seems too >convenientaH >to me.  Call me a cynic perhaps.  The company to whom we outsourced the >helpiL >desk has certain performance metrics they are required to meet.  How better >torG >handle a question they can't answer than by claiming that the physicale >layerD >needs expensive upgrading?  And guess who gets the contract for the	 >upgrade?n >e> >If any of you can shed light on this it would be appreciated. >- >- >- >- >- >- >- >-   ------------------------------  % Date: Mon, 30 Oct 2000 21:41:02 -050022 From: "Richard B. Gilbert" <DRAGON@compuserve.com>, Subject: VAX Mail question - corrupted files6 Message-ID: <200010302141_MC2-B904-A9C@compuserve.com>  B         Clearly you are a man who knows bullshit when he hears it!  F         The "network bandwidth" excuse is just that: pure bullshit.  IF would hestitate to speculate on the causes or motives for your getting bullshit instead of help.a  E         MAIL is not usually the best tool for sending files.  I wouldn9 prefer to use DECnet copy; e.g. COPY FILE.DAT NODENM"usernJ password"::DISK:[DIR.SUB].   The username and password are not required i= fuG you have a network login proxy on the remote node.  Copy is supposed toaJ guarantee that you don't get corruption without knowing it; if the file i= se( corrupt you should get an error message.  @         The actual problem might be hardware or software.  As anJ unprivileged user, you may not have the necessary access to really diagno= seH the problem.  Try "back door" access.  Let someone in your systems groupJ know of the problem, with details; what did you send, from which node  to=  D which node, what commands did you use, what time did you do it, etc.  3 Message text written by INTERNET:old_timer@user.vmsoD >I work for a large international company which has "outsourced" its computer3 support to yet another large international company.R  G I recently had occasion to experience the level of quality of knowledgey which ! results from such an arrangement.a  J I am a 20-year end-user of VMS, and have picked up a little knowledge alo= ng themE way.  My question has to do with the answer I finally got (after manyt initialnI silly and obviously wrong suggestions) from the outsourced "help desk". =    Let me start by stating the problem.,  H I have recently encountered trouble sending large text files to a remote VAXiJ node via VAX mail.  When using VAX mail to send a large text file (either=   asG in-line text embedded in the message or using the /foreign qualifier tor send aJ saveset) to a remote node, I occasionally get transmission errors (or, mo= reJ precisely stated, the file arrives corrupted).  This does not happen 100%=   of the time.i  < Until recently, I had never before encountered this problem.  D The bottom line is, the help desk is telling me that it's a "network	 bandwidth J problem".  What I would like to know is, is this a reasonable explanation= ,e or am3 I being taken?  H Here's my reasoning.  I know the network bandwidth going out of and into our:J site is less than it should be, but wouldn't you think that running out o= fe> bandwidth would cause very slow transmissions, or even aborted transmissions,; rather than random errors interspersed throughout the file?   E Besides, the answer "it's a network bandwidth problem" just seems tooU
 convenientG to me.  Call me a cynic perhaps.  The company to whom we outsourced then helpJ desk has certain performance metrics they are required to meet.  How bett= er toF handle a question they can't answer than by claiming that the physical layergC needs expensive upgrading?  And guess who gets the contract for the  upgrade?  = If any of you can shed light on this it would be appreciated.o     <o   ------------------------------  % Date: Mon, 30 Oct 2000 23:23:20 -0500h2 From: rdeininger@mindspring.com (Robert Deininger)0 Subject: Re: VAX Mail question - corrupted filesL Message-ID: <rdeininger-3010002323210001@user-2ivebbg.dialup.mindspring.com>  B In article <8tkg5b0hc4@drn.newsguy.com>, old_timer@user.vms wrote:    M > I am a 20-year end-user of VMS, and have picked up a little knowledge along  > the O > way.  My question has to do with the answer I finally got (after many initialoJ > silly and obviously wrong suggestions) from the outsourced "help desk".   K There's a special place in hell for people who foist outsourced help desks o on the rest of us.  P > The bottom line is, the help desk is telling me that it's a "network bandwidthc > problem".  What I would like to know is, is this a reasonable explanation, or > am I being taken?m  K I would say your help desk is full of beans.  The various network protocolsnK used by mail are supposed to be "reliable".  They watch for errors and keepaL trying until the transmission is error free.  Or they give up.  They should  corrupt your data.  F For extra fun, call up the supervisor of the help desk and complain...   -- c Robert Deininger rdeininger@mindspring.com    ------------------------------    Date: 30 Oct 2000 20:54:07 -0800 From: old_timer@user.vms0 Subject: Re: VAX Mail question - corrupted files) Message-ID: <8tlj9f01jfk@drn.newsguy.com>   6 In article <200010302141_MC2-B904-A9C@compuserve.com>:  = >        MAIL is not usually the best tool for sending files.a  M Yes, I know.   I won't go into the frustrating details, but suffice it to sayeN that MAIL is the only access permitted due to decisions (now cast in concrete)2 that were made long before I arrived on the scene.  $ >I would prefer to use DECnet copy; = >e.g. COPY FILE.DAT NODENM"user password"::DISK:[DIR.SUB].   5. >The username and password are not required if4 >you have a network login proxy on the remote node.   N Thanks for the suggestion; I was already familiar with that... unfortunately IO can't even try it to get additional evidence because I don't have an account ont the remote node.  I >As an unprivileged user, you may not have the necessary access to reallyw >>diagnose the problem.y  O Access to the remote node is a real problem, but privileged access to the localbP node is not.  I know the guy who has the system password and he'll do anything IP ask.  That's the good news.  The bad news is, he has less VMS experience than meP (the experienced VMS guy left the company when they outsourced).  If anybody canO tell me what what diagnostic commands to issue or what config files to browse IrO can post the details.  Again, any help is very much appreciated.  If I can nail,K this thing down, it will be sweet to debunk the "network bandwidth" answer.1   ------------------------------    Date: 30 Oct 2000 21:07:13 -0800 From: old_timer@user.vms0 Subject: Re: VAX Mail question - corrupted files) Message-ID: <8tlk2101mha@drn.newsguy.com>   , In article <39FDDE02.B79B274C@videotron.ca>:   >do a MAIL> SHOW ALL > C >See if there is a default transport defined (towards the bottom). nL >do a SET NOTRANSPORT to remove it. It is possible that they are routing theC >mail to the other nodes with something different that DECnet-MAIL.t  * I will try this tomorrow and post results.  L >If you are JDoe on nodeA and the other on NodeB, you can try the following: >b= >SEND/foreign mybigfile.txt/subject="test" nodeB::NodeA::JDoe.   ...actually MAIL/foreign...r   >hG >This should route the message to node B and then back to Node A and beH >delivered to yourself.e  L Yes; I have done that many times.  The round-trip e-mail has random embeddedN errors (regardless of whether it's in-line message text or an attached savesetN with /foreign qualifier), even though I get no operating system error message.  G >When you send a big file (or when you specify /foreign) the message is.; >actually stored as a separate file in your mail directory.g   Yes, thanks, I knew that.8   >Doing a DIR/FULL onK >that newly received file in the mail directory and the original file mightn >reveal something.  L Using "extract" to get the attachment (when using the /foreign qualifier) orK looking directly at the MAIL$xxxxxxxxxx file in the mail directory makes noP0 difference; they are both corrupted identically.  N >I think you need to find out exactly how the mail is being sent and over what >it is being routed.  M I think you are right, but I don't know how to get this information.  Is thistL something I can query from the local node? (I can get system access for thisO purpose if it will help). If so, what are the commands?  If not, assuming I cannN talk my way into the glass house, what do I look for?  The problem is, I don'tJ have any formal VMS network training.  All I know comes from many years of end-user experience.    9 >If the network allows corrupted mail, god only know whati* >else is being corrupted during transport.  M I agree.  I wish I had access to an account on the remote node so I could run-O additional tests, but for irrational reasons which I won't go into here, that'su not currently possible.    ------------------------------    Date: 30 Oct 2000 21:16:37 -0800 From: old_timer@user.vms0 Subject: Re: VAX Mail question - corrupted files) Message-ID: <8tlkjl01og5@drn.newsguy.com>r  !  rdeininger@mindspring.com wrote:a  G >For extra fun, call up the supervisor of the help desk and complain...   H Been there; done that.   There's no leverage.   I have no input into theJ performance appraisal process for the help desk personnel or supervisors. M They're all outsourced and their main concern is their contract metrics.  ThejP guy from our company who oversees the contract is so high up in the organizationL that he's inaccessible to me.  Besides, he's probably some pencil-pusher whoD wouldn't have a clue if I tried to explain how we're being had here.  M In the bad old days of the good ole boy network, it mattered what the project1M people thought of the support personnel, and they acted accordingly.  Not anyb more.    ------------------------------  % Date: Tue, 31 Oct 2000 01:52:39 -0400i- From: JF Mezei <jfmezei.spamnot@videotron.ca>i0 Subject: Re: VAX Mail question - corrupted files, Message-ID: <39FE5E25.94856491@videotron.ca>   old_timer@user.vms wrote:nN > Using "extract" to get the attachment (when using the /foreign qualifier) orM > looking directly at the MAIL$xxxxxxxxxx file in the mail directory makes nom2 > difference; they are both corrupted identically.  L If you send file1.dat from nodea:jdoe to nodeb::nodea::jdoe, and you get theK message back, you need to compare the extracted file with the original fileo> which you sent to see if it has maintained its file structure.  J If, for instance, your network converts VMSmail to SMTP/mime format, it isM possible that the file will lose its vms attributes and come back as a simple  binary file.  O > I think you are right, but I don't know how to get this information.  Is thise- > something I can query from the local node? -   do SHOW NETWORKcR also, try SET HOST NODEB   to see if your system can find a decnet route to nodeb.  N Also, can you provide a description of exactly what type of corruption happens
 on the file ?sJ if you could do a DIR/FULL of the original file as well as a DIR/FULL of a7 received file that is corrupt it might give some clues.e  O You can also use the DIFF file1 file2  command to see what the differences are.a   ------------------------------  % Date: Mon, 30 Oct 2000 21:45:13 -0600s% From: Dan O'Reilly <dano@process.com> ( Subject: VAXstation 4000 Model 60 manualD Message-ID: <5.0.0.25.2.20001030214443.00a9c848@pop.clsp.uswest.net>  G Is there one on the web, or does somebody have one they no longer need?h   ------I +-------------------------------+---------------------------------------+hI | Dan O'Reilly                  |                                       |uI | Principal Engineer            |  "Those are my principles. If you     |EI | Process Software              |   don't like them I have others."     |oI | http://www.process.com        |                    -- Groucho Marx    |sI +-------------------------------+---------------------------------------+    ------------------------------   Date: 31 Oct 2000 03:31:35 GMT* From: bleau@umtof.umd.edu (Lawrence Bleau)3 Subject: Why missing [000000] after volume restore?l) Message-ID: <8tleen$fje$1@hecate.umd.edu>o  H Hello, I'm having a problem with a volume set on one of my systems.  I'mO running OpenVMS V7.1-2 AXP.  The volume set contains 3 members.  The 2nd memberuL went bad on me.  I have a spare disk, so I decided to use it in place of theM bad member.  I have an image backup of the volume set as of Oct 1.  I mountedfM the backup tape, mounted the target (spare) disk as /FOREIGN, and entered thes command:  6 BACKUP/IMAGE/VOLUME=2 4MM:saveset/SAVE targetdisk:/LOG  J I know that the files on this volume set were added in such a way that theO first member was nearly filled, then the second member added and nearly filled,rH then the third member added.  The files are added in lexicographic orderN (there's a date field encoded in the filename).  So, I know that the files forL the time frame 19990126 through 20000506 are on volume two, and no others.    E When BACKUP reported it had restored data set 20000506 I did a ^Y andyM dismounted the target disk.  I then mounted the volume set, with the new diskbI in place of the old one, in the normal manner.  This failed with an errormL message that SECURITY.SYS did not exist; it did not say from which volume it was missing.  L I next entered the MOUNT command again, but using /OVERRIDE=SECURITY .  This0 mounted the volume set okay and did a rebuild.    P I next ran ANAL/DISK/REPAIR on the volume set, which gave me a strange message. J I didn't preserve the exact message, but it was to the effect that certainN files were not in any directory.  What worried me was that the files were .SYSN files!  After ANAL/DISK finished, I looked in the [SYSLOST] directory that had- been created on the 1st volume.  It contains:i   Directory DKA600:[SYSLOST]   BADBLK.SYS;1         (3,3,2) BADLOG.SYS;1         (9,9,2) BITMAP.SYS;1         (2,2,2) INDEXF.SYS;1         (1,1,2)   Total of 4 files.   I Note that these files' FileIDs say they are located on the 2nd volume!  IDO checked the second volume explicitly by doing a DIR volume:[000000] to see whatr2 was (or wasn't) there.  I got the following error:   $ DIR DKB400:[000000]t= %DIRECT-E-OPENIN, error opening DKB400:[000000]*.*;* as inpute -RMS-E-DNF, directory not foundu" -SYSTEM-W-NOSUCHFILE, no such file  N It seems that there is no [000000] on volume 2!  (There is such a directory onI volume 1 and 3, btw.)  I know enough about VMS to know that I cannot justcM CRE/DIR/VOL=2 DKB400:[000000] and have it work properly; the file id would bea wrong.  J What has happened?  Is my only recourse to reload the volume 2 from backupM again?  Why did it not work?  Did my ^Y of BACKUP hose the restore operation?n Thanks.c   Lawrence Bleau University of Maryland" Physics Dept., Space Physics Group 301-405-6223 bleau@umtof.umd.edu0   ------------------------------   End of INFO-VAX 2000.608 ************************