1 INFO-VAX	Tue, 13 Jun 2000	Volume 2000 : Issue 329       Contents:. 7.3 SDK1/EFT1 and AlphaServer GS Series: Don't Re: Autogen   BACKUP and tape related question$ Re: BACKUP and tape related question$ Re: BACKUP and tape related question$ Re: BACKUP and tape related question$ Re: BACKUP and tape related question$ Re: BACKUP and tape related question$ Re: BACKUP and tape related question$ Re: BACKUP and tape related question Backup performance Re: Backup performance Re: Backup performance/ Re: C bashing (was Re: VMS File Caching Futures > Re: Canadian Association of Compaq Users Symposium rescheduled" Re: CMQ/Dec with a sense of Humor?" Re: CMQ/Dec with a sense of Humor? Re: CONV$RECLAIM question  Re: DCPS 1.7 and HP 6MP  DECnet help! Re: DECnet help! Fun VMS Facts? Re: Fun VMS Facts?$ Re: Hobbyist vms 7.2 alpha questions) how to stop echo on DS90M telnet listener - Re: how to stop echo on DS90M telnet listener - Re: how to stop echo on DS90M telnet listener - Re: how to stop echo on DS90M telnet listener  Re: Info on VAXstation 400-VLC Re: Internet VMSCluster  Re: Internet VMSCluster  Re: Internet VMSCluster  LATNET Wiring standards  Re: LATNET Wiring standards  Re: Process States0 Re: Qt2.1.0 GUI library now available on OpenVMS1 Question about pbxgk-bb and vrc21-ha on as500/400  RE: Recording CD's on VMS - Re: Serial port programming (TX, LTA, TELNET)  Re: TAR FOR VMS? Re: VAX on Intel?  Re: VAX on Intel?  Re: VAX on Intel?  Re: VAX on Intel?  Re: VMS Security features  Re: VMS Security features  Re: VMS Security features  Re: VMS Security features  Re: VMS Security features  Re: VMS Security features  Re: VMS Security features ; Re: Which cpu does Sun server use: I386, Mip, Ppc or Alpha? ; Re: Which cpu does Sun server use: I386, Mip, Ppc or Alpha?  www.compaq.com [Fwd: SF36 documentation]   F ----------------------------------------------------------------------   Date: 13 Jun 2000 14:46:51 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)7 Subject: 7.3 SDK1/EFT1 and AlphaServer GS Series: Don't 6 Message-ID: <8i5hgr$9ac$3@mailint03.im.hou.compaq.com>  -   OpenVMS V7.3 EFT1/SDK1 System Information:  /     o AlphaServer GS80, GS160, and GS320 series      o Multia Series   J   OpenVMS Engineering does not have all of the pieces for the AlphaServer J   GS80, GS160, and GS320 series systems integrated into the first externalG   field test (field test or SDK) of OpenVMS Alpha V7.3, nor has much in L   the way of testing been performed on these configurations -- please don't #   bother to try this combination...   G   Pending the availability of this support, please continue to use the  L   OpenVMS Alpha V7.2-1H1 hardware release on these AlphaServer GS80, GS160,    and GS320 systems.  J   Systems supported on OpenVMS Alpha V7.2-1 should continue to operate on H   the EFT1/SDK.  (nb: the Multia box is NOT a supported system.  I'd notM   initially assume that the Multia will boot, though I need to confirm this.)   J   We do expect to have support for the AlphaServer GS80, GS160, and GS320 I   series systems integrated into the first external this work integrated  &   into subsequent field test releases.  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Mon, 12 Jun 2000 23:03:00 -0700 9 From: tforsyth <ted.forsythNOteSPAM@amcor.com.au.invalid>  Subject: Re: Autogen9 Message-ID: <03486990.cd0cc9f2@usw-ex0105-034.remarq.com>    Shawn,  ; Yes, savparams to setparams will set the system ready for a > reboot (only specify feedback if it's been up less than 24 hrs1 and you _need_ feedback from the running system).   @ It'll also run a genfiles phase, so be prepared for some disk to; disapear as that'll create new page, swap and dump files if  needed.   @ In which case, please note that SETPARAMS only needs SYSPRV, but GENFILES also needs CMKRNL...    Hope it helps...   Ted.  L * Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *G The fastest and easiest way to search and participate in Usenet - Free!    ------------------------------  # Date: Tue, 13 Jun 2000 06:27:51 GMT 2 From: "Zane H. Healy" <healyzh@shell2.aracnet.com>) Subject: BACKUP and tape related question 6 Message-ID: <Hvk15.518$ru5.159252@typhoon.aracnet.com>  I OK, this most likely qualifies as a stupid question.  Basically I think I , know what's going on, but want to make sure.  G If you have a DCL script that does a BACKUP command, then the tape will L automatically be mounted if it hasn't already been mounted.  Now suppose youD don't dismount it anywhere in the script.  I take it that it will beL unmounted when the script ends?  If this is in fact the case is there anyway( to keep it mounted when the script ends?   				Zane   ------------------------------   Date: 13 Jun 2000 07:12:27 GMT) From: leslie@clio.rice.edu (Jerry Leslie) - Subject: Re: BACKUP and tape related question ' Message-ID: <8i4msr$mes$1@joe.rice.edu>   1 Zane H. Healy (healyzh@shell2.aracnet.com) wrote: K : OK, this most likely qualifies as a stupid question.  Basically I think I . : know what's going on, but want to make sure.  I : If you have a DCL script that does a BACKUP command, then the tape will N : automatically be mounted if it hasn't already been mounted.  Now suppose youJ : you don't dismount it anywhere in the script.  I take it that it will beH : unmounted when the script ends?  If this is in fact the case is there 1 : anyway to keep it mounted when the script ends?  : 
 : 				Zane  / Add the following to the end of the DCL script:      $ dismount/nounload  mka0:  3 substituting your system's tape device for 'mka0:'.   4 --Jerry Leslie     (my opinions are strictly my own)   ------------------------------  % Date: Tue, 13 Jun 2000 10:31:12 +0900 2 From: Mike Rechtman <michael.rechtman@digital.com>- Subject: Re: BACKUP and tape related question + Message-ID: <39460D6F.54174D91@digital.com>    Zane H. Healy wrote: > K > OK, this most likely qualifies as a stupid question.  Basically I think I . > know what's going on, but want to make sure. > I > If you have a DCL script that does a BACKUP command, then the tape will N > automatically be mounted if it hasn't already been mounted.  Now suppose youF > don't dismount it anywhere in the script.  I take it that it will beN > unmounted when the script ends?  If this is in fact the case is there anyway* > to keep it mounted when the script ends? > & >                                 ZaneB DCL script running how? If its SUBMITted as a batch job, the tape 2 will no longer be allocated when the job logs out.H If you run the DCL script from an interactive job that does not log out,- the tape will remain mounted (and allocated.)    ~Mike  --  E --------------------------------------------------------------------- E Usual disclaimer: All opinions are mine alone, perhaps not even that. ? Mike Rechtman                            *rechtman@tzora.co.il* F Kibbutz Tzor'a.                          Voice (home): 972-2-9908337  B   "20% of a job takes 80% of the time, the rest takes another 80%"E ---------------------------------------------------------------------    ------------------------------  % Date: Tue, 13 Jun 2000 11:39:35 +0100   From: steven.reece@quintiles.com- Subject: Re: BACKUP and tape related question > Message-ID: <802568FD.003AA56C.00@qedilc01.qedi.quintiles.com>   Jerry Leslie wrote/quoted : 4 >>>Zane H. Healy (healyzh@shell2.aracnet.com) wrote:K : OK, this most likely qualifies as a stupid question.  Basically I think I . : know what's going on, but want to make sure.  I : If you have a DCL script that does a BACKUP command, then the tape will N : automatically be mounted if it hasn't already been mounted.  Now suppose youJ : you don't dismount it anywhere in the script.  I take it that it will beG : unmounted when the script ends?  If this is in fact the case is there 1 : anyway to keep it mounted when the script ends?  :  :                   Zane  / Add the following to the end of the DCL script:      $ dismount/nounload  mka0:  6 substituting your system's tape device for 'mka0:'.<<<  4 That covers the loading, but not the mounting Jerry.  
 For Zane :  L Loading is the physical status of whether the tape is actually available forN mounting.  In the case of a 1/2" tape or a DLT or a TK50/TK70, this is whetherI or not the tape leader has been grabbed by the take up reel and is at the M beginning of the tape.  For an Exabyte or DAT it's whether the tape is in the K drive (the /UNLOAD qualifier on the mount command will spit the tape out in K these cases since both the supply and take up reels are in the cartridge of  course).  O Mounting is the VMS part of the process by which the tape is made available for ! reading and/or writing by the OS.   O Mounted tapes will be dismounted at or before process or job termination.  I.E. O if you mount the tape and then logout, the tape is dismounted when you log out. P If you mount the tape you can issue a dismount command against it and it will beA dismounted when all files are closed and processing has finished.   O The /NOUNLOAD means that the tape is still in the drive and can be remounted by , another process without manual intervention.   ------------------------------  % Date: Tue, 13 Jun 2000 11:00:02 -0500 ! From: "News" <michaela@hkusa.com> - Subject: Re: BACKUP and tape related question * Message-ID: <39465881@news.advancenet.net>  L If you run backup from a .COM file, the tape will NOT dismount automaticallyJ after it's done.  However, if you send the .COM file to a batch queue, theI tape will be dismount and ejected after the job finished.  I am trying to H find a way to have the tape stay in the drive after the batch job ended.H Our situation is that on weekends, we send two batch jobs (one on FridayI night and one on Saturday night) to run BACKUP.  Since the tape is always K ejected after the first job, somebody has to come to the office to reinsert 1 the tape so that the second job can run properly.   2 Does anyone know how to fix the problem?   Thanks!  
 Michael An Human Kinetics, Publishers      ; Zane H. Healy <healyzh@shell2.aracnet.com> wrote in message 0 news:Hvk15.518$ru5.159252@typhoon.aracnet.com...K > OK, this most likely qualifies as a stupid question.  Basically I think I . > know what's going on, but want to make sure. > I > If you have a DCL script that does a BACKUP command, then the tape will J > automatically be mounted if it hasn't already been mounted.  Now suppose you F > don't dismount it anywhere in the script.  I take it that it will beG > unmounted when the script ends?  If this is in fact the case is there  anyway* > to keep it mounted when the script ends? >  > Zane   ------------------------------  # Date: Tue, 13 Jun 2000 17:22:13 GMT 3 From: "Gord Coulman" <nospam_gcoulman@ccinet.ab.ca> - Subject: Re: BACKUP and tape related question 9 Message-ID: <95u15.8277$vT6.891621@news1.telusplanet.net>    $DISMOUNT/NOUNLOAD  * News <michaela@hkusa.com> wrote in message$ news:39465881@news.advancenet.net...@ > If you run backup from a .COM file, the tape will NOT dismount
 automatically L > after it's done.  However, if you send the .COM file to a batch queue, theK > tape will be dismount and ejected after the job finished.  I am trying to J > find a way to have the tape stay in the drive after the batch job ended.J > Our situation is that on weekends, we send two batch jobs (one on FridayK > night and one on Saturday night) to run BACKUP.  Since the tape is always D > ejected after the first job, somebody has to come to the office to reinsert3 > the tape so that the second job can run properly.  > 4 > Does anyone know how to fix the problem?   Thanks! >  > Michael An > Human Kinetics, Publishers   ------------------------------  # Date: Tue, 13 Jun 2000 17:32:50 GMT 2 From: "Zane H. Healy" <healyzh@shell1.aracnet.com>- Subject: Re: BACKUP and tape related question 6 Message-ID: <6fu15.551$ru5.164831@typhoon.aracnet.com>  ! steven.reece@quintiles.com wrote: N > Loading is the physical status of whether the tape is actually available forP > mounting.  In the case of a 1/2" tape or a DLT or a TK50/TK70, this is whetherK > or not the tape leader has been grabbed by the take up reel and is at the O > beginning of the tape.  For an Exabyte or DAT it's whether the tape is in the M > drive (the /UNLOAD qualifier on the mount command will spit the tape out in M > these cases since both the supply and take up reels are in the cartridge of 
 > course).  J It gets more interesting in my case.  Not only am I using a Exabyte drive,I but it's in a rather antiquated 'Dumb' Exabyte Jukebox.  When the tape is L dismounted the jukebox automatically grabs the tape in the next slot.  SinceI this is on my Home Cluster I'm not generating 5GB of data a day and would C just as soon write as many of the weeks Incrementals to one tape as H possible.  So it looks like the DISMOU/NOUNLOAD is what I'm looking for.  H For the people that asked, yes this is a Batch job and it runs while I'mL asleep.  That and the fact the system lives out in the garage is why I was aJ bit confused by how my script is working.  I'm in the middle of a complete< rewrite though, and needed to be sure as to what's going on.   				Zane   ------------------------------  % Date: Tue, 13 Jun 2000 12:12:57 -0400 * From: David A Froble <davef@tsoft-inc.com>- Subject: Re: BACKUP and tape related question - Message-ID: <39465D89.A2B86B51@tsoft-inc.com>    News wrote:  > N > If you run backup from a .COM file, the tape will NOT dismount automaticallyL > after it's done.  However, if you send the .COM file to a batch queue, theK > tape will be dismount and ejected after the job finished.  I am trying to J > find a way to have the tape stay in the drive after the batch job ended.J > Our situation is that on weekends, we send two batch jobs (one on FridayK > night and one on Saturday night) to run BACKUP.  Since the tape is always M > ejected after the first job, somebody has to come to the office to reinsert 3 > the tape so that the second job can run properly.  > 4 > Does anyone know how to fix the problem?   Thanks!  M This has nothing to do with the DCL script.  It has to do with process exit.  G When a process that has a tape mounted exits, the tape is automatically D 'logically' dismounted, which normally causes a 'physical' dismount.  I The command, as others have pointed out, DISMOUNT/NOUNLOAD will logically K dismount the tape, and leave it physically mounted for later use by another N process.  Once you logically dismount the drive, it no longer 'belongs' to the: process, and process exit has no effect on the tape drive.   --  4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com6 T-Soft, Inc.  170 Grimplin Road  Vanderbilt, PA  15486   ------------------------------  % Date: Tue, 13 Jun 2000 11:49:30 +0100 2 From: Chris Sharman <Chris.Sharman@CCAgroup.co.uk> Subject: Backup performance . Message-ID: <39461FCA.6C42CAF1@CCAgroup.co.uk>  > We have a full-ish RAID 230 box, and some older disks (due forG replacement), all backed up with a TZ87. Advice from here some time ago ' has got the backup running pretty well.   G We do backup & defrag every night, for most of the night, which used to D be OK. Increasingly now we're running 24 hours, though, and could do without the performance impact.   F I wondered if there was some better way - perhaps to have the RAID boxE do the work, rather than VMS ? (Having seen how blisteringly fast the A RAID box bring a new shadow set member online, without noticeably A degrading performance, it looks like a better hardware solution). E Alternatively perhaps to fail out a member & back that up, although I < can't see how to access a member after I've swxcr failed it.  H All advice & suggestions appreciated - the pointy-haired ones might even* be willing to part with money for this ...   ------------------------------  # Date: Tue, 13 Jun 2000 14:27:51 GMT 3 From: "Gord Coulman" <nospam_gcoulman@ccinet.ab.ca>  Subject: Re: Backup performance 9 Message-ID: <Hxr15.8247$vT6.883243@news1.telusplanet.net>    A couple of options:  I 1.  Get another big disk or several good sized ones and do a disk-to-disk H backup.  This will be fast and fairly cheap. Then do the backup to tape, taking as long as you want.p  K 2.  Get a faster tape drive (DLT).  I think the TZ87 is a DAT if I remembere right, very slow.   C 3.  Add more than one DLT drives and configure them as a tape RAID.   Assuming VMS supports tape RAID?   Gord.   = Chris Sharman <Chris.Sharman@CCAgroup.co.uk> wrote in messageu( news:39461FCA.6C42CAF1@CCAgroup.co.uk...@ > We have a full-ish RAID 230 box, and some older disks (due forI > replacement), all backed up with a TZ87. Advice from here some time ago7) > has got the backup running pretty well.! >uI > We do backup & defrag every night, for most of the night, which used to F > be OK. Increasingly now we're running 24 hours, though, and could do! > without the performance impact.o > H > I wondered if there was some better way - perhaps to have the RAID boxG > do the work, rather than VMS ? (Having seen how blisteringly fast thelC > RAID box bring a new shadow set member online, without noticeablyTC > degrading performance, it looks like a better hardware solution).vG > Alternatively perhaps to fail out a member & back that up, although IE> > can't see how to access a member after I've swxcr failed it. >NJ > All advice & suggestions appreciated - the pointy-haired ones might even, > be willing to part with money for this ...   ------------------------------  % Date: Wed, 14 Jun 2000 01:32:41 +1000e1 From: "Geoffrey Sinclair" <grsinclair@one.net.au>M Subject: Re: Backup performanceu& Message-ID: <3946517e@pink.one.net.au>  D Chris Sharman wrote in message <39461FCA.6C42CAF1@CCAgroup.co.uk>...? >We have a full-ish RAID 230 box, and some older disks (due foriH >replacement), all backed up with a TZ87. Advice from here some time ago( >has got the backup running pretty well. > H >We do backup & defrag every night, for most of the night, which used toE >be OK. Increasingly now we're running 24 hours, though, and could dou  >without the performance impact. >DG >I wondered if there was some better way - perhaps to have the RAID box F >do the work, rather than VMS ? (Having seen how blisteringly fast theB >RAID box bring a new shadow set member online, without noticeablyB >degrading performance, it looks like a better hardware solution).F >Alternatively perhaps to fail out a member & back that up, although I= >can't see how to access a member after I've swxcr failed it.r  C Please do not fail out members, there is no guarantee of what statelF the disk is in, did all I/Os complete for example, besides what if one. really fails while your chosen disk is failed?  I >All advice & suggestions appreciated - the pointy-haired ones might evena+ >be willing to part with money for this ...i  F It really depends on your configuration, what database product if any?E What percentage of the files are updated each day?  What time windowso? do you have for performance impairing maintenance?  Do you haveiE any quiet time at all over the week?  The non hardware solution is touH come up with a job that only backs up the key files daily, and do a full> back up once a week, on the weekend, assuming the weekends areG quieter.  Also consider doing the defrag once a week.  Generally though-E hardware is cheaper than paying people to change tapes or hang aroundo( just in case at night or on the weekend.  A The hardware solution can be anything from a mirror site to spareqF disk (to do a disk to disk backup first then to tape) or multiple tape@ drives.  It depends on what sort of window you normally have and6 what sort of performance hit you see and can tolerate.  = If you have a nightly window and want a full backup every days@ consider multiple tape drives doing different disks in parallel,< minimises back up time and also gives you redundancy, if one; tape drive fails, you do not have to rush replace your onlye< backup device.  If you have a very small window do a disk to disk backup then to tape.o  7 Again the answer really depends on what your site needs ? to ensure all data is safely backed up in a way that as much as,9 possible guarantees you can restore a working system withs intact data.   Geoffrey Sinclair    ------------------------------  % Date: Tue, 13 Jun 2000 09:23:57 -0400e( From: Ed Vogel <edward.vogel@compaq.com>8 Subject: Re: C bashing (was Re: VMS File Caching Futures* Message-ID: <394635ED.8427603F@compaq.com>  >     Just to make everyone aware, the Compaq C compiler has theF     ability to detect many of the cases that have been discussed here.2     If you compile /WARN=ENABLE=QUESTCODE the V6.22     compiler will issue diagnostics for the cases:       if (xyz = abc)	         {          do_this();	         }d       and:       /*     ** Try this nested comment     *\&     printf("This will never print\n");       /*     ** And there is no error     */       and:       do_this;    H     as well as many other common user "mistakes".  While it is true that8     using /WARN=ENABLE=QUESTCODE can also generate a lotF     of "noise", it can be also save developers a lot of time by having0     the compiler detect potential coding errors.  @                                                         Ed VogelL                                                         Compaq C Engineering   ------------------------------  # Date: Tue, 13 Jun 2000 16:07:34 GMTn* From: Art Rice <arice.NOcSPAM@ue.itug.org>G Subject: Re: Canadian Association of Compaq Users Symposium rescheduleds8 Message-ID: <dslckscfi6nhtcp45h70d9k2i3k5rpfvqe@4ax.com>  C On Mon, 12 Jun 2000 23:27:18 GMT, Kilgallen@eisner.decus.org.nospamm (Larry Kilgallen) wrote:  _ >In article <skaevqj9h5167@corp.supernews.com>, "Peter Weaver" <peter.weaver@stelco.ca> writes:eF >> Larry Kilgallen wrote in message <6iJNoVN1Cj$2@eisner.decus.org>... >>> ....@ >>>From Canada, however, the complaints have continued unabated.= >>>There was apparently a clearly named individual at CACU to ? >>>contact, and no reported success stories at getting into the  >>>VMS hobbyist plan.  >>> ...  >> aD >> I was able to get the Hobbyist licenses using my old DECUS Canada; >> membership number so that is at least one success story.h >'E >Thank you.  The reports of success seem to be made less eagerly thani >those of failure. >tF >> As a side note, are there any DECUS groups in the world that chargeG >> money for membership? Was the Canadian DECUS group the only one that I >> charged membership fees when it died? I was going to ask about the oldI >d@ >The US DECUS chapter had a voluntary higher level of membership> >lately that required a fee.  It may have gone away. DECUServe? >formerly had a fee, but as I understand the cost of collectinge >it overwhelmed the revenue. >-I >> "Canadian Tandem Users Group" but I went back to read Joseph's note torB >> make sure I had the name correct. In his note he said "I am theA >> current President for the Canadian Tandem Users Group, and theaE >> Director of Finance for the Canadian Association of Compaq Users."oF >> Does that mean that Tandem still has an user group? I was under theG >> impression that both DECUS and CTUG had both been killed to form the- >> new CACU????a > A >Not specific to Canada, but my impression is that DECUS and ITUG @ >have sufficiently different cultures and needs that there is no$ >simple way to just make them one.  B ITUG is self supporting through the members whereas DECUS receivedA their support through DEC/Compaq.  Individual membership fees arepA $150.  If that were lowered ITUG would be depending on Compaq fortF financial assistance and Compaq would be giving direction to ITUG.  AsC a separate entity, ITUG actually has the ability to put pressure on F Tandem/Compaq and let them know what the users actually want and need.  : On the other hand, Compaq held the purse strings of DECUS.  C As far as different cultures,  I believe that less today than I didn> three years ago.  Except for the Fab plants and some real-time9 installations,  OpenVMS and NSK are used in much the sametE environments,  Finance, Telco, Health Care, etc..  And in some cases,<B side by side.  Who would have thought 10 years ago that over 1,000E Tandem processors would be utilized to handle e-mail at a single siteaF for a single ISP? (a job normally handled by UNIX)  Both platforms areE expensive, (you get what you pay for), both are realiable and proven.kA Applications for either are expensive and have expensive training B courses.  At least VAX/VMS/OpenVMS was taught and utilized in many; colleges and universities.  Maybe that's where the culturali differences still exist.   >ITUG, for instance, has their@ >yearly symposium in San Jose, and I presume that is to maximize# >participation by Tandem engineers.n     -- w
 Art Rice   **S# Special Data Processing Corporationd& --------------------------------------' All opinions expressed are mine and do l% not reflect the views of my employer.l   ------------------------------  % Date: Tue, 13 Jun 2000 09:35:10 -0400 " From: Dan Sugalski <dan@sidhe.org>+ Subject: Re: CMQ/Dec with a sense of Humor? : Message-ID: <4.3.2.7.0.20000613093432.01ca85d0@24.8.96.48>  , At 10:50 PM 6/12/00 +0000, Mark Sterk wrote:     >Dan Sugalski wrote: >-. > > At 12:24 AM 6/12/00 -0400, JF Mezei wrote: > > >Antony Wardle wrote: 2 > > > > I wasn't sure if I had the hold screen, so1 > > > > I thought that I would try a control T tos" > > > > see if there was any life. > > > >.( > > > > Well, I got a message up sayine: > > > > ( > > > > dead_eater waiting for dead_beef > > >iG > > >I know that at least one DEC engineer was a fan of Monthy Python.  
 > Sounds like I > > >a reference to a famous sketch (where one is invited to eat his deado > > >mother in law). > >tL > > More likely it's looking for a quadword with the value 0xDEADBEEF in it.N > > One of the fun words and phrases you can actually spell entirely in hex... >s@ >No it's the name of the 'idle' process under the alpha console.K >If you press ^T you could see it if you are lucky, but you can better use   >the 'ps' command.  G Isn't that the one that scribbles 0XDEADBEEF in unused memory? Or am I .! mixing it up with something else?5   					Dan  I --------------------------------------"it's like this"-------------------o2 Dan Sugalski                          even samurai? dan@sidhe.org                         have teddy bears and evenB;                                       teddy bears get drunkd   ------------------------------  % Date: Wed, 14 Jun 2000 00:59:36 +1000 4 From: Huw Davies <Huw.Davies@kerberos.davies.net.au>+ Subject: Re: CMQ/Dec with a sense of Humor?nD Message-ID: <4.3.1.2.20000614005604.01d8ea70@kerberos.davies.net.au>  , At 09:35 13/06/00 -0400, Dan Sugalski wrote:  H >Isn't that the one that scribbles 0XDEADBEEF in unused memory? Or am I " >mixing it up with something else?  H ISTR that this was something MIPS did with RISC/OS back in the dim dark ; ages (at least that's when I first recall seeing DEADBEEF).u  A Again, not much VMS relevance, but I do recall asking a somewhat yK interesting question at an Australian DECUS when the DECstation 3100s were   announced along the line of   H     "How does it feel to announce Digital's fastest ever system using a  non-Digital chip?"  3 I know how to make friends and influence people :-)p  @ Huw Davies           | e-mail: Huw.Davies@kerberos.davies.net.au?                       | "If God had wanted soccer played in then=                       | air, the sky would be painted green" y   ------------------------------   Date: 13 Jun 2000 01:23 -0400c From: hein@eps.zko.dec.c*m" Subject: Re: CONV$RECLAIM question& Message-ID: <13JUN200001233897@miasys>  ^ In article <3943F2FB.2A62849E@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes...    I >Doing performance tests on a MV II (where you can see vast difference insJ >performance), I have noticed that a CONV$RECLAIM is almost instant for anK >almost empty file (even if lots of blocks have been allocated), whereas itwL >takes much longer if I wait a while and it have to reclaim lots of buckets.  8 Be sure to use the reclaim stats to quantify the effort.@ You may also want to limit efforts to key-0 for your application  3 Reclaim walks the data buckets for the index file.  L If an all-empty bucket is found, the next pointer to the prior bucket (whichK pointed to it) made to point around it to th enext bucket mentioned in the sC empty one. Then the index down pointer to the empty data bucket is oE removed and finally the bucket itself is put on the reclaimed bucket sH list in the area control block. (ANA/RMS/INT will readily confirm this).  I Not having looked up the details in a while ( how many buffers kept dirtynI how long?) it would appear that roughly the some amount of 'work' (read, fH scan, write) would be needed for a simalar amount of data. You just pick< between paying 100 times 1 units or 1 times a hundred units.  N Reclaiming for the whole day is problably a little more efficient in terms of H startup/shutdown/roundup. It has a risk however that you end up pushing / into an additional indexe layer during the day. I Reclaiming a little bit at the times, probably allows your system to find E bits and pieces in the various caches (vioc, HSx, disk). It may well - keep the index down to 1 level.u  I An approach might be to reclaim whenever the the application 'knows' thatRL an additional index layer is imminent (number of records queued, SYS$DISPLAY9 + AREAS, File grown beyond initial extend, whatever)...).   E Instead of a daily large reclaim, leaving behind a few valid records,eB why not do a full convert with FDL to create a clean file with the few remaining records.  L Or... You can possibly keep the down time very low by pre-creating the file;J Flipping the names (rename? logical?); force a re-open by the application H for the data entry stream; Have the data removal stream read to the end;L Verify no more record, have the data removal stream restart in the new file. Throw away the old file.  F [for one special application I once wrote a 'reset' tool to just blastM the index file header blocks with a clean template. The application just keptt@ on switching between 2 files, always on the same disk locations.J Heck... I'll append some code along those lines (not recently tested!)   ]  	 Have fun,c          Hein.  # #define OUTPUT_TYPE     "_TEMPLATE", /*C **  INDEXED_FILE_RESET.C    Hein van den Heuvel, DIGITAL, June 1996t **= **  This program effectivly clears an indexed file by copying A **  the prologue from a termplate file over the current prologue.nA **  A teplate file is used such that the program does not have tov4 **  understand indexed file RMS internal structures.5 **  The template file is created if it did not exist.-; **  The prologue and target file-id are meant to be cached,o9 **  with caches flushed upon failure to open target file.  **? **  The program could be run eventually as a server process for:= **  example an input mailbox to receive commands, and one for> **  output and errorsr **  Suggested commands are:.5 **      RESET <target-filespec> [<template-filespec>] * **      LIST  !output list of cached files+ **      CLEAR remove all cached informationh **; **  If a command line command is present, it is interpretede7 **  as a single shot reset. Actually... that's the onlyh **  part present right now.n ** */ #include <rms> #include iodef #include atrdefh #include <stdio> #include <descrip> #include fibdef 3 typedef struct { long count; void *address; } desc;e   typedef struct {     char  input_name[64];e     char  device[32];b     short fid[3];o     short chan;n&     variant_union  { int    file_size;B             struct { short  lo, hi;} file_size_words;} size_union;     int   template_size;-     char  template[5*512]; } file_block_type;a   main (int argc, char *argv[])a {d int             stat;g short           iosb[4]; desc            temp_desc;     int   template_size; file_block_type *file; struct fibdef   fib; struct atrdef   atr[2];  union {1     char fill[ATR$S_RECATTR];aB     short word[5];}fat; /* fake it, as fat.h not always present */  ,     file = malloc(sizeof (file_block_type));     file->fid[0] = 0;e     file->chan   = 0;      file->template_size = 0;(     strcpy (&file->input_name, argv[1]);  G     if (file->fid[0] == 0) stat = load_data ( argv[1], argv[2], file );g!     if (!(stat & 1)) return stat;      if (file->chan   == 0) {/            temp_desc.count   = file->device[0]; 0            temp_desc.address = &file->device[1];A            stat = sys$assign ( &temp_desc, &file->chan, 0, 0, 0);o(            if (!(stat & 1)) return stat;            }  "     temp_desc.count = sizeof(fib);     temp_desc.address = &fib; "     memset (&fib, 0, sizeof(fib));;     fib.fib$r_acctl_overlay.fib$l_acctl  = ( FIB$M_WRITE );AH     fib.fib$r_fid_overlay.fib$r_fid_fields.fib$w_fid_num = file->fid[0];H     fib.fib$r_fid_overlay.fib$r_fid_fields.fib$w_fid_seq = file->fid[1];P     fib.fib$r_fid_overlay.fib$r_fid_fields.fib$r_fid_rvn_overlay.fib$w_fid_rvn =         file->fid[2];        /*M     ** Need to get record attributes to set new eof size to make it look nice      */'     atr[0].atr$w_type  = ATR$C_RECATTR;l'     atr[0].atr$w_size  = ATR$S_RECATTR;n     atr[0].atr$l_addr  = &fat;     atr[1].atr$w_type  = 0;      atr[1].atr$w_size  = 0;b  B     stat = SYS$QIOW( 0, file->chan, IO$_ACCESS|IO$M_ACCESS, &iosb,9                      0, 0, &temp_desc, 0, 0, 0, &atr, 0); !     if (stat & 1) stat = iosb[0]; !     if (!(stat & 1)) return stat;e         /*E     ** Overwrite file's RMS prologue to init AREA and KEY descriptorss     */9     stat = SYS$QIOW( 0, file->chan, IO$_WRITEVBLK, &iosb,:M                      0, 0, &file->template, file->template_size, 1, 0, 0, 0);B!     if (stat & 1) stat = iosb[0];4!     if (!(stat & 1)) return stat;r       /*,     ** Set size to truncate to upon closing.8     ** Set EOF to truncated size swapping words as we go     */6     fib.fib$r_exctl_overlay.fib$w_exctl = FIB$M_TRUNC;&     fib.fib$l_exvbn = file->file_size;+     fat.word[5] = file->file_size_words.lo;e+     fat.word[4] = file->file_size_words.hi;h8     stat = SYS$QIOW( 0, file->chan, IO$_DEACCESS, &iosb,9                      0, 0, &temp_desc, 0, 0, 0, &atr, 0);D!     if (stat & 1) stat = iosb[0];t     return stat; }a          J load_data ( char *input_name, char *template_name, file_block_type *file ) {  int             stat, i;, char            rsa[256], default_name[128]; struct FAB      fab; struct RAB      rab; struct NAM      nam; struct XABSUM   sum; struct XABALL   all; struct XABKEY   *key;I void            **xabs;s fab = cc$rms_fab;e rab = cc$rms_rab;  nam = cc$rms_nam;l sum = cc$rms_xabsum; all = cc$rms_xaball;   fab.fab$l_nam = &nam;i fab.fab$l_xab = &sum;  fab.fab$l_fna = input_name; % fab.fab$b_fns = strlen( input_name );f rab.rab$l_fab = &fab;  nam.nam$l_rsa = &rsa;  nam.nam$b_rss = 255; sum.xab$l_nxt = &all;l  G /* First open this input file and have RMS fill in all the XAB present.tG ** RMS will also return a resultant file name with device + name parts.tG ** CHeck to make sure it is an indexed file with only one area defined.a */  I stat = SYS$OPEN ( &fab );   /* open, get expanded filename, get # keys */s if (!(stat&1)) return stat;a0 if (fab.fab$b_org != FAB$C_IDX) return RMS$_ORG;( if (sum.xab$b_noa != 1) return RMS$_AID;   /*A ** Init Xab Key for each defined key and insert in list (order!).o */J file->template_size = 512 * ( 1 + (sum.xab$b_nok+4)/5 ); /* area + keys */9 for (i = 0; i < 3; i++) file->fid[i] =  nam.nam$w_fid[i];k xabs = &fab.fab$l_xab;, key = malloc (sum.xab$b_nok * sizeof *key );$ for (i=0; i < sum.xab$b_nok; i++ ) {     *key = cc$rms_xabkey;l     key->xab$b_ref = i;e     key->xab$l_nxt = *xabs;y     *xabs = key;
     key++;     }r stat = SYS$DISPLAY ( &fab );  if (!(stat&1)) return stat;   stat = SYS$CLOSE ( &fab ); if (!(stat&1)) return stat;i   /*? ** Now figure out the template file name and create or open it.i? ** Re-use the fab from the input file with all xabs filled  in.o */& fab.fab$b_fac = FAB$M_BRO | FAB$M_GET; fab.fab$l_fop = FAB$M_CIF;  file->device[0] = nam.nam$b_dev;9 strncpy (&file->device[1], nam.nam$l_dev, nam.nam$b_dev);wH strncpy (default_name, nam.nam$l_name, nam.nam$b_name + nam.nam$b_type);0 default_name[nam.nam$b_name + nam.nam$b_type]=0;# strcat (default_name, OUTPUT_TYPE);  fab.fab$l_dna = default_name;-( fab.fab$b_dns = strlen ( default_name ); fab.fab$l_fna = template_name; fab.fab$b_fns = 0;< if (template_name) fab.fab$b_fns = strlen ( template_name );     stat = SYS$CREATE ( &fab); if (!(stat&1)) return stat;u  M file->file_size = fab.fab$l_alq; /* review if template is to be truncated! */a rab.rab$l_rop = RAB$M_BIO;$ rab.rab$w_usz = file->template_size;# rab.rab$l_ubf = &file->template[0];  stat = SYS$CONNECT ( &rab);c if (!(stat&1)) return stat;y   stat = SYS$READ ( &rab); if (!(stat&1)) return stat;    stat = SYS$CLOSE ( &fab);t if (!(stat&1)) return stat;   & free (fab.fab$l_xab);   /* key xabs */ return stat; }t   ------------------------------  # Date: Tue, 13 Jun 2000 15:00:48 GMTm+ From: Paul Anderson <panderson@genicom.com>s  Subject: Re: DCPS 1.7 and HP 6MPC Message-ID: <panderson-9447B9.11005113062000@news.bellatlantic.net>v  C In article <FE1835D68492D311BF7900508B5BEB0D0DE7EF@petra.WPI.EDU>,  . "Mitchell, David R." <mitchell@WPI.EDU> wrote:  D > I suspect it is a lack of resources (memory - as it only has 3MB) G > that is causing the problem as smaller documents print successfully. -D > But at this point I'm just guessing as all I have to go on is the  > DCPS error: : > "%DCPS-E-FLUSHING, Rest of Job (to EOJ) will be ignored"  H If possible, turn on the printer's PostScript error reporting feature.  > This may give you more information about the PostScript error.   Paul   -- a"    Paul Anderson, DCPS Engineering"    GENICOM Corporation, Gardner MA   ------------------------------  % Date: Tue, 13 Jun 2000 08:42:48 -0700g+ From: Steven Xie <r33300@email.sps.mot.com>. Subject: DECnet help!e1 Message-ID: <39465678.1BA1244A@email.sps.mot.com>e   Hello,  ?                    I have a new system, OS is V7.2-1. I want top configure it to be                    runningC                    DECnet-plus. This is the first time I try to use  DECnet-plus, can                    somebodyiD                    show me a example for how to configure and how to design the names                    for that.E                    Looking at the new manual for DECnet-plus confusedg me.                         Thanks,                    StevenC   ------------------------------  % Date: Tue, 13 Jun 2000 19:02:14 +0200d= From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com>e Subject: Re: DECnet help!v) Message-ID: <39466915.16A6C2E0@gtech.com>e   Steven Xie wrote:aA > I have a new system, OS is V7.2-1. I want to configure it to befT > running DECnet-plus. This is the first time I try to use DECnet-plus, can somebody3 > show me a example for how to configure and how toh > design the name for that.o8 > Looking at the new manual for DECnet-plus confused me.   From memory:   $ @SYS$MANAGET:NET$CONFIGURE basic configurationa simple 6 character names phase IV address  = and you should have DECnet-plus up and running in 10 minutes.i   Arne   ------------------------------  # Date: Tue, 13 Jun 2000 16:10:36 GMT? From: jbecker@ui.urban.org Subject: Fun VMS Facts?i) Message-ID: <8i5mdp$4eg$1@nnrp1.deja.com>d  4 I'm putting together an orientation presentation for9 internal end users who are new to OpenVMS Alpha. For part 8 of the presentation, I'm going to include material along3 the lines of "What's OpenVMS" and "What's an Alpha"I6 because most of these people are completely unfamiliar with either.  6 Here's where you come in. What fun VMS facts would you7 include in such an orientation? I'm looking for tidbitsu6 that can be stated briefly and effectively to (in this6 situation) a general but educated audience. I want the- items to show that VMS is a good place to be.a  5 I have several ideas of my own, but I thought I'd aski the VMS community.   --
 Jim Becker+ The Urban Institute (http://www.urban.org/) 3 DECUS ESILUG (http://eisner.decus.org/lugs/esilug/)n    & Sent via Deja.com http://www.deja.com/ Before you buy.    ------------------------------  % Date: Tue, 13 Jun 2000 12:39:04 -0400p" From: Dan Sugalski <dan@sidhe.org> Subject: Re: Fun VMS Facts?o: Message-ID: <4.3.2.7.0.20000613123826.01c983d0@24.8.96.48>  6 At 04:10 PM 6/13/00 +0000, jbecker@ui.urban.org wrote:7 >Here's where you come in. What fun VMS facts would yous8 >include in such an orientation? I'm looking for tidbits7 >that can be stated briefly and effectively to (in thisu7 >situation) a general but educated audience. I want thee. >items to show that VMS is a good place to be.  I I'd make a quick scan through the CERT archives and pop up the number of aG CERT security advisories on a per-OS basis. That one's always fun... :)s   					Dan  I --------------------------------------"it's like this"------------------- 2 Dan Sugalski                          even samurai? dan@sidhe.org                         have teddy bears and even ;                                       teddy bears get drunkD   ------------------------------   Date: 13 Jun 2000 16:28:18 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)- Subject: Re: Hobbyist vms 7.2 alpha questions 6 Message-ID: <8i5nf2$bsn$1@mailint03.im.hou.compaq.com>  U In article <cancel.XhH%4.113998$WA1.9368@news2.mts.net>, spambot@stopspam.org writes:h. :I know how to do that [upgrade to V7.2 -srh].* :how do I upgrade without erasing my data?  <   You choose "upgrade" or "preserve" as part of the process.  H :the docs I have (manual with openvms 7.2) just restore a save set which :erases the hard drive.o  @   Huh?  That's not what _my_ manual says -- what manual are you    looking at here?  E   I'm guessing you are looking at some documentation specific to the iC   kit(s) provided by the hobbyist program?  If the hobbyist kit is yC   simply an image (BACKUP/IMAGE) of the distribution kit, then you eD   will need a spare disk to use as intermediate (temporary) storage @   during the upgrade (or installation) -- you will be recreatingD   the CD distribution media on this disk, then booting and using it D   for the OpenVMS upgrade (or installation) on the real system disk.  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Tue, 13 Jun 2000 13:21:26 GMT = From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)u2 Subject: how to stop echo on DS90M telnet listener0 Message-ID: <009EB8A1.36A04126@SendSpamHere.ORG>  I I have a DS90M.  I have enabled a telnet listener port tied to one of theaK RJ45 ports and the RJ45 is then connected to a serial port on a VAXstation.hK When I TELNET to the port on the DS90M, I connect to the VAXstation withouth/ a problem however, I get a strange echo back...M   Username: system system- Password: type_password_here_and_it's_visibler  K I can't seem to find a port characteristic which will stop this.  Any help?a   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM0   ------------------------------    Date: 13 Jun 2000 15:43:12 +0200* From: eplan@kapsch.net (Peter LANGSTOEGER)6 Subject: Re: how to stop echo on DS90M telnet listener* Message-ID: <39463a70$1@news.kapsch.co.at>  p In article <009EB8A1.36A04126@SendSpamHere.ORG>, system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) writes:J >I have a DS90M.  I have enabled a telnet listener port tied to one of theL >RJ45 ports and the RJ45 is then connected to a serial port on a VAXstation.L >When I TELNET to the port on the DS90M, I connect to the VAXstation without0 >a problem however, I get a strange echo back... >e >Username: systemv >systemc. >Password: type_password_here_and_it's_visible >IL >I can't seem to find a port characteristic which will stop this.  Any help?  L Mhhm. ECHO is negotiated between TELNET server (the DS90M) and TELNET client$ (what is this, UCX/TCPIP/TCPware ?).  J AFAIK, the DS90M port telnet server can disable echo negotiation (but whatO does it do then ? Echo or No-Echo ?), but I think it can't disable echo itself.l+ Maybe the TELNET client can/should do it...    --  < Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651; Network and OpenVMS system manager  Fax.    +43 1 81111-888 < FBFV/Information Services           E-mail  eplan@kapsch.netF <<< KAPSCH AG  Wagenseilgasse 1     PSImail PSI%(0232)281001141::EPLANH A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a realist"N "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998   ------------------------------  # Date: Tue, 13 Jun 2000 14:57:16 GMTp= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)n6 Subject: Re: how to stop echo on DS90M telnet listener0 Message-ID: <009EB8AE.99CEDD27@SendSpamHere.ORG>  W In article <39463a70$1@news.kapsch.co.at>, eplan@kapsch.net (Peter LANGSTOEGER) writes:aq >In article <009EB8A1.36A04126@SendSpamHere.ORG>, system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) writes: K >>I have a DS90M.  I have enabled a telnet listener port tied to one of thetM >>RJ45 ports and the RJ45 is then connected to a serial port on a VAXstation.cM >>When I TELNET to the port on the DS90M, I connect to the VAXstation withouta1 >>a problem however, I get a strange echo back...a >> >>Username: system >>system/ >>Password: type_password_here_and_it's_visible, >>M >>I can't seem to find a port characteristic which will stop this.  Any help?  >aM >Mhhm. ECHO is negotiated between TELNET server (the DS90M) and TELNET cliento% >(what is this, UCX/TCPIP/TCPware ?).i  J Same results w/ Multinet V4.2 Rev A-X, UCX Version V4.2 and TCP/IP Version V5.0A.     --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMp   ------------------------------  % Date: Tue, 13 Jun 2000 16:32:57 +0100   From: steven.reece@quintiles.com6 Subject: Re: how to stop echo on DS90M telnet listener> Message-ID: <802568FD.00558227.00@qedilc01.qedi.quintiles.com>   Peter Langstoeger wrote/quoted:hO >>>>When I TELNET to the port on the DS90M, I connect to the VAXstation withouts0 >a problem however, I get a strange echo back... >a >Username: systeml >systemo. >Password: type_password_here_and_it's_visible >]L >I can't seem to find a port characteristic which will stop this.  Any help?  L Mhhm. ECHO is negotiated between TELNET server (the DS90M) and TELNET client$ (what is this, UCX/TCPIP/TCPware ?).  J AFAIK, the DS90M port telnet server can disable echo negotiation (but whatO does it do then ? Echo or No-Echo ?), but I think it can't disable echo itself.n. Maybe the TELNET client can/should do it...<<<  > This is the reason I moved to DS700s with DNAS on flash cards.I - You can telnet onto the console port on the system really quickly after N powerup (such as after a power outage), usually quicker than the time taken toP get the host up so you can watch it booting successfully (or not, but then againI there is no or not cos we're talking VMS here aren't we boys and girls?); O - There's no annoying echo from the DECserver, just the one from the VMS systeml sitting behind it.  M There may be other ways, but I got fed up trying so a DS700 with DNAS won the  day.   Steve.   ------------------------------  % Date: Tue, 13 Jun 2000 00:57:41 -0400g* From: David A Froble <davef@tsoft-inc.com>' Subject: Re: Info on VAXstation 400-VLC - Message-ID: <3945BF45.E2B43082@tsoft-inc.com>t   Bill Bradford wrote: > ? > Anybody know where I can find a user's manual for this thing? 7 > (VAXstation 4000-VLC).  Specifically, I need to know:n > C >         - Type of RAM (I know its 72pin parity, but can I fill it  >           with 16s?)  P This system supports 24 MB of memory.  The 6 slots can accept 4 MB 72 pin parityN SIMMs.  You MUST put the SIMMs in in pairs, ie; 2, 4, or 6 SIMMs, and in slots 1&2, 3&4, and 5&6.  C >         - Position of "S3" switch to enable serial console (all Iu >           have is a VT320).   J Not sure, it's on the right side, toward the rear.  That's 'right side' ifN you're facing the front of the unit, and then your right.  Think it's UP.  Try each, it's one or the other.  A > Of course, if anyone has an owner's manual for this unit or cany= > point me somewhere online, it would be greatly appreciated.a > ; > Thanks!  This machine will soon be running the decvax.orgr > web site. 8-)    Dave   --  4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com6 T-Soft, Inc.  170 Grimplin Road  Vanderbilt, PA  15486   ------------------------------  # Date: Tue, 13 Jun 2000 15:42:21 GMT  From: d.webb@mdx.ac.uk  Subject: Re: Internet VMSCluster) Message-ID: <8i5kos$3ma$1@nnrp1.deja.com>   + In article <8i2clg$tds@newton.cc.rl.ac.uk>,f-   "Richard Brodie" <R.Brodie@rl.ac.uk> wrote:a >iC > "Stephen Eickhoff (remove the - to reply)" <operagost@e-mail.com>a wrote in message& > news:39444BF0.530159DF@e-mail.com... > E > > However, perhaps I'm mistaken in my belief that load balancing ise
 supported. >tB > Yes. If you want load balancing you have to use DNS round robin. >  >    No.a  C TCPIP Services aka UCX supports 3 different types of cluster alias.]    C 1) The "default" cluster alias. This is what you get if you run the0F    @sys$manager:ucx$config or @sys$manager:tcpip$config procedure in aE    cluster and answer yes to the questions about setting up a cluster 
    interface.1>    This provides failover capabilities only. No round robin or    load-balancing.   2) Round-robin. G    For this you need a DNS. This can be running on any system but needs &    to be authorative for your systems.A    You then setup multiple address records for the cluster alias.   2    Connections then cycle between these addresses.E    This provides a crude load-balancing (ie each system gets an equals    share of connections).   F    Does not provide failover. If a system goes down then when its turn?    comes around clients will still try to connect to this node..   3) Load-balancing.  9    This uses a load-broker which gathers information fromh@    cluster members (each of which runs a metric server to gether    information on its loading).&E    The load-broker then constructs a list of multiple address records*@    (as in round-robin) and updates the authorative DNS with thisA    information. This dynamic updating is done on a regular basis. 5    Systems which are down are removed from this list. H    The load-broker orders the list according to the load on the systems.  F    This gives you a reasonable measure of load-balancing together with6    fairly quick removal of downed nodes (ie failover).     For more information see :  8 Digital TCP/IP Services for OpenVMS Management chapter 6  H http://www.openvms.compaq.com:8000/72final/6526/6526/profile_contents_00 1.html    
 David Webb VMS and Unix team leader CCSS Middlesex University    & Sent via Deja.com http://www.deja.com/ Before you buy.l   ------------------------------  % Date: Tue, 13 Jun 2000 17:52:59 +0100e* From: "Richard Brodie" <R.Brodie@rl.ac.uk>  Subject: Re: Internet VMSCluster+ Message-ID: <8i5ota$tsq@newton.cc.rl.ac.uk>   G <d.webb@mdx.ac.uk> wrote in message news:8i5kos$3ma$1@nnrp1.deja.com...X  = >>If you want load balancing you have to use DNS round robin.*E > TCPIP Services aka UCX supports 3 different types of cluster alias./   > 2) Round-robin.  > 3) Load-balancing.  I You're right, of course. What I meant to say was that for load balancing,.B some sort of DNS based scheme would be required. One can certainly/ do better than the simple round-robin approach.+   ------------------------------  # Date: Tue, 13 Jun 2000 17:16:59 GMTl From: d.webb@mdx.ac.uk  Subject: Re: Internet VMSCluster) Message-ID: <8i5q9p$7c6$1@nnrp1.deja.com>e  ) In article <8i5kos$3ma$1@nnrp1.deja.com>,(   d.webb@mdx.ac.uk wrote: - > In article <8i2clg$tds@newton.cc.rl.ac.uk>,d/ >   "Richard Brodie" <R.Brodie@rl.ac.uk> wrote:i > >fE > > "Stephen Eickhoff (remove the - to reply)" <operagost@e-mail.com>E > wrote in message( > > news:39444BF0.530159DF@e-mail.com... > >tG > > > However, perhaps I'm mistaken in my belief that load balancing ist > supported. > >mD > > Yes. If you want load balancing you have to use DNS round robin. > >$ > >e >; > No.t >eE > TCPIP Services aka UCX supports 3 different types of cluster alias.n > E > 1) The "default" cluster alias. This is what you get if you run the H >    @sys$manager:ucx$config or @sys$manager:tcpip$config procedure in aG >    cluster and answer yes to the questions about setting up a cluster  >    interface./@ >    This provides failover capabilities only. No round robin or >    load-balancing. >i > 2) Round-robin. C >    For this you need a DNS. This can be running on any system butf needst( >    to be authorative for your systems.C >    You then setup multiple address records for the cluster alias.( >t4 >    Connections then cycle between these addresses.G >    This provides a crude load-balancing (ie each system gets an equal, >    share of connections).r > H >    Does not provide failover. If a system goes down then when its turnA >    comes around clients will still try to connect to this node.v >  > 3) Load-balancing. >3; >    This uses a load-broker which gathers information fromRB >    cluster members (each of which runs a metric server to gether! >    information on its loading). G >    The load-broker then constructs a list of multiple address records B >    (as in round-robin) and updates the authorative DNS with thisC >    information. This dynamic updating is done on a regular basis.o7 >    Systems which are down are removed from this list.oA >    The load-broker orders the list according to the load on thee systems. >rH >    This gives you a reasonable measure of load-balancing together with8 >    fairly quick removal of downed nodes (ie failover). >  > For more information see : > : > Digital TCP/IP Services for OpenVMS Management chapter 6 >e >eH http://www.openvms.compaq.com:8000/72final/6526/6526/profile_contents_00 > 1.html >h > David Webb > VMS and Unix team leader > CCSS > Middlesex University > ( > Sent via Deja.com http://www.deja.com/ > Before you buy.s >i  ) Sorry correction to URL that should be :-   F www.openvms.compaq.com:8000/72final/6526/6526profile_contents_001.html    
 David Webb VMS and Unix team leader CCSS Middlesex University    & Sent via Deja.com http://www.deja.com/ Before you buy.e   ------------------------------  % Date: Tue, 13 Jun 2000 07:59:00 -0400-- From: "news.rcn.com" <ccastiglia@engsint.com>   Subject: LATNET Wiring standards+ Message-ID: <8i57lj$sep$1@bob.news.rcn.net>c  F Running a LATNET network consisting of ten (10) Digital DECServer 90L+I terminal servers spaced out along approximately 750 feet of thinwire coaxn8 (there's a repeater at approximately the 500 foot mark).  G The 90L+ servers are typically located in the drop ceiling or a closet.oL Four-pair Cat 3 solid 24AWG cable (Black Box #EYN734A) runs from the serversG to wall jacks and then standard DEC "flat" interconnect cables are usedN* between wall jacks and terminals/printers.  E Question:  what is the maximum permissible length for the Cat 3 cablenI between server and wall jack/terminal?  As I approach the maximum length,,5 would solid UTP CAT 3/Cat 5 be a better cable choice?    ------------------------------  % Date: Tue, 13 Jun 2000 06:16:00 -0700 ? From: Mike Price <mike.priceNOmiSPAM@littlewoods.co.uk.invalid>e$ Subject: Re: LATNET Wiring standards9 Message-ID: <03030c9d.8bcb5e53@usw-ex0102-015.remarq.com>o  > This may not answer your question fully as I do not specialise@ in cable types etc. but I have had problems with terminal cables< in the past and this was relevant for me. I apologise if you already know this and more.o  ; The cable from the terminal server to the terminal (which It= believe is the bit in question) does not know or care whethert= the network is LAT, IP or anything else. This bit is simply av@ dec423 connection. This is like a RS232 connection combined with> and RS422. Therefore the cable distances you need are for this rather than LAt etc. These distances are as follows  # For DEC423, maximum distances are :o   Up to 19.2Kb/s - 1000 feet Up to 38.4Kb/s - 500 feet   + For 9600 Kb/s the distances are even longera  = I have had problems in the past pushing a 38400 terminal past-9 500 feet - on a screen paint the user would start getting:@ jibberish on the screen. We dropped the speed to 19200 and there? were no further problem. This was using cat 5 cabeling and I ame assured it was top quality.i@ Putting better quality cable in would not have fixed the problem in our casee   Hope this helpsd  
 Mike Price  L * Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *G The fastest and easiest way to search and participate in Usenet - Free!i   ------------------------------   Date: 13 Jun 2000 08:18:33 GMT+ From: "labadie" <gerard.labadie@compaq.com>e Subject: Re: Process StatesaC Message-ID: <01bfd50f$ed0ad8c0$521ebc10@w-glabadie.frq.cpqcorp.net>   3 lib/extract=$rsndef/out=rsn.mar sys$library:lib.mlb   ' will put in rsn.mar the list of states.    Regards-   Grard    F nleach@gol.com a crit dans l'article <8hqtap$u5u$1@nnrp1.deja.com>...G > I seem to remember that there used to be a file (in sys$system maybe)nH > that contained the list of process state acroynms and their full names% > (ie RWAST = Resource Wait For AST).n > D > Am I halucinating as it doesn't seem to be around anymore (v7.2) ? >  > Nigelw >  >  > ( > Sent via Deja.com http://www.deja.com/ > Before you buy.e >    ------------------------------  # Date: Tue, 13 Jun 2000 11:31:10 GMTs% From: "P.Lj" <plj@byron.ext.telia.se> 9 Subject: Re: Qt2.1.0 GUI library now available on OpenVMS 2 Message-ID: <39461B88.78229DBC@byron.ext.telia.se>   Hi,y   Just a %ONE-I-MSGe   I had to do a:  4 set file qt21.bck /attr=(RAT:NONE,LRL:32256,RFM:FIX)  F before I could use backup to restore the saveset, if someone else.....   /P.Ljk     Rainer Lehrig wrote:  B > Qt is a C++ library for developing GUI applications and more ...@ > The special thing about Qt is that it is platform independent.. > It runs on most unix systems and on windows.; > Now that I have ported Qt to OpenVMS it also runs on VMS.e> > You can get information about Qt on http://www.trolltech.comD > Qt is the base for the KDesktopEnvironment (KDE) known from Linux.C > I'am in contact with trolltech in order to support Qt for OpenVMS-E > officially. For this we need people willing to test Qt for OpenVMS.HF > Qt for OpenVMS is available under http://www.lehrig.de (click button > service).iH > It is a zip file containing a readme.txt and a backup saveset (33 MB). >I > Dr.-Ing. Rainer Lehrig > mailto: lehrig@t-online.de   ------------------------------  % Date: Tue, 13 Jun 2000 09:07:19 -0600n2 From: "Stephens, Diana" <Diana.Stephens@Coors.com>: Subject: Question about pbxgk-bb and vrc21-ha on as500/400L Message-ID: <0F5051B5F548D311AA1B00508B5A5715025372E6@jane.golden.coors.com>   Hi,_  H I have an alphastation 500/400. I need to do some "system swapping" withL older slower systems. I need to put my old vrc21-ha on the as500. I was toldG by my salesman that a pbxgk-bb (elsa gloria synergy graphics card - aka2H powerstorm3d) works with the vrc21-ha. When I attach the vrc21-ha to theK pbxgk-bb (with a bc13-10 cable), the screen and text "bow" on the right andpF left sides. Nothing I do with the monitor adjustments fixes this. This4 monitor works fine on the as250 with a zlxp-l1 card.  * Has anyone been able to get this to work?   	 Thank youi   Diana Stephens Coors Brewing Company5   ------------------------------  % Date: Tue, 13 Jun 2000 10:16:22 +0200 C From: Eberhard Heuser-Hofmann <vaxinf@chclu.chemie.uni-konstanz.de>w" Subject: RE: Recording CD's on VMSA Message-ID: <009EB8A8.E30FB77D.1127@CHCLU.CHEMIE.UNI-KONSTANZ.DE>_   Alex,b  6 may be you've got already response, but I think it is 5 important for you to know that everything is already b9 there and in contrast to other solutions the software is s free:U  6 try cdrecord for vms, if you have a Alpha-Workstation.5 there is cdcopy-program on the same ftp-server and a e java-based-GUI, too.  - ftp://v36.chemie.uni-konstanz.de/cdrecord_vmsb+ ftp://v36.chemie.uni-konstanz.de/diskreader-+ ftp://v36.chemie.uni-konstanz.de/burnit_vms    regards1 eberhard   ------------------------------  % Date: Tue, 13 Jun 2000 13:49:49 +0400n4 From: Valentin Likoum <valentin.likoum@ncc.volga.ru>6 Subject: Re: Serial port programming (TX, LTA, TELNET), Message-ID: <394603BD.57B26A4D@ncc.volga.ru>   JF Mezei wrote:  > J > For an application which needs to access a modem outbound (dialing out).D > I don' have a terminal server, so I can't reall test these things. >  > 17G > Is VMS now capable of handling a $ASSIGN to an LTA device without the P > IO$M_Access etc bit(eg: can the same code be used to assign a TX port as a LTA > port ?  G Yes, you can $assign to device and driver will automatically connect to D terminal server during first $qio[w] if destination for lta port has been> defined previously (mcr latcp set port ltaxxx: /node=node_name /port=port_name)   >  > 2vK > Since  terminal servers don't allow speed and other port parameters to be4L > changed, what happens if I do an IO$_SETMODE operation with a speed changeP > request it it to a LTA port ?  Does the driver simply ignore the parameters itP > doesn't handle and set those that are relevant to an LTA port, or will it fail > and report an error ?lO > (e.g. can the same IO$_SETMODE be used on TX and LTA ports without problem ?)e  C Even if it would work (I didn't try it) you will set speed only forn virtual B lta port and won't touch real port on terminal server. Really some	 terminal  C servers does allow setting speed on ports via snmp or connecting to  command G console via MOP or telnet, though it's not very easy and not compatible-	 with TXs.-   >  > 3 M > For systems who use TCPIP only, how does an application access a modem on aj? > TCPIP terminal server ?  What sort of parameters are needed ?8 > L > (Is it correct to assume that there is no MC LATCP DEFINE PORT LTA8373.... > equivalent for TCPIP?) > N > While I realise that serial ports are a dying breed and that permanent TCPIPP > connections to toasters and the rest of the world are what is coming, it wouldP > be nice to have a single $ASSIGN_PORT routine that an application can call andP > provide any/all parameters and the routine will use the parameters it needs toS > assign and set a serial port, independant of whether it is a TX, LTA, or TN port.    -- s     Valentin Likoumo'     valentin.likoum@ncc.volga.ru.nospam    ------------------------------  % Date: Wed, 14 Jun 2000 00:46:17 -0400E. From: Michael Austin <maustin@nc.prestige.net> Subject: Re: TAR FOR VMS?w/ Message-ID: <39470E19.D080DE4B@nc.prestige.net>h  , This is a multi-part message in MIME format.& --------------A9050F457FD662D694189864* Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7biti  M VMSTAR works GREAT... used it to move vms directory structure to a SunOS box.i   Bob Koehler wrote:  Z > In article <39412b5f.0@nnrp1.news.uk.psi.net>, andywright@psilink.co.uk (Andy w) writes:
 > > Hello,= > >       I need to transfer files from Openvms V7.2 on AlphaiO > > to a Tru64 system via tape (DLT), (no network so it has to be via tape!!!).-J > > Is there a tar utility for vms or something simillar so I can do this.J > > As they are both Compaq systems you would think there should be one!!! > >0 >tI > Both VMS and Tru64 understand ANSI/ASCII labeled tapes (VMS as it's own J > labeled tapes and Tru64 via lft).  This will preserve file names but not > directory hierarchy. > I > There are several tars available for VMS.  One in the POSIX kit, if youiI > happen to have that, and many can be found on the net.  I've been using E > Sid Penstone's, but I use his tarread more often than his tarwrite.  >6H > I don't have a URL for any, but I generally just start from the FAQ orJ > other links on www.openvms.digital.com when I want to find free software
 > for VMS. > H > ----------------------------------------------------------------------A > Bob Koehler                     | Computer Sciences Corporationh? > Hubble Space Telescope Payload  | Federal Sector, Civil Group G >  Flight Software Team           | please remove ".aspm" when replyinge  & --------------A9050F457FD662D694189864- Content-Type: text/x-vcard; charset=us-ascii;e  name="maustin.vcf"  Content-Transfer-Encoding: 7bit , Content-Description: Card for Michael Austin  Content-Disposition: attachment;  filename="maustin.vcf"    begin:vcard  n:Austin;Michael c tel;work:704-947-1089r x-mozilla-html:FALSE org:Michael Austin, Inc-
 adr:;;;;;; version:2.1a+ email;internet:michaelaustininc@hotmail.come title:Presidentj x-mozilla-cpt:;0 fn:Michael E. Austin	 end:vcardm  ( --------------A9050F457FD662D694189864--   ------------------------------  % Date: Tue, 13 Jun 2000 11:36:35 +0100i$ From: Andy Burns <andy@burns.uk.net> Subject: Re: VAX on Intel?8 Message-ID: <f3ubks41uepgecd9uutjb7brpu1r2tacgq@4ax.com>   David J. Dachtera wrote:  5 >OpenVMS: 450,000 sites.		Compaq stock price: < $40USp >e5 >Micro$hit: 200,000,000+ sites	M$ stock price: > $120s  J what does the price of a single respective share have to do with anything?F surely the market capitalisation is the meaningful figure ? (unless US2 stocks are rebased according to available volumes)     --  
 Andy Burns   ------------------------------  % Date: Tue, 13 Jun 2000 07:30:44 -0400o& From: "Jeff Killeen" <Jeff@Killeen.cc> Subject: Re: VAX on Intel?% Message-ID: <394619f1@news.toast.net>   J > Someone tell me something: under EXACTLY what conditions would expandingC > the OpenVMS user base harm Compaq's stock price or its investors?m  F If the expansion is achieve by doing something that resulted in CompaqL losing money.  For example dramatically reducing the cost of OVMS to competeL Windows and Linux while maintaining the same levels of support for OVMS theyI provide today when you buy a new license.  That would be a cost structurem/ that volume sales would never compensate for...i     --     Jeff Killeen - www.Killeen.ccsE =====================================================================aB "David J. Dachtera" <djesys.nospam@earthlink.net> wrote in message' news:3945A986.FD0E0421@earthlink.net...c > Larry Kilgallen wrote: > >eG > > In article <elkZ4.61$ru5.1966@typhoon.aracnet.com>, "Zane H. Healy" $ <healyzh@shell1.aracnet.com> writes:G > > > In comp.org.decus David J. Dachtera <djesys.nospam@earthlink.net>t wrote:H > > >> Looks like someone ELSE sees the value/potential of VMS on Intel. > > > 8 > > >> Wonder what they know that DEC/Compaq doesn't.... > > C > > They know that they (SRI) are not required to make a profit for  > > shareholders ? > 6 > DAMNIT I'M TIRED OF THESE "SHAREHOLDER" FANTASIES!!! > J > Someone tell me something: under EXACTLY what conditions would expandingC > the OpenVMS user base harm Compaq's stock price or its investors?v >i5 > OpenVMS: 450,000 sites. Compaq stock price: < $40USy >b6 > Micro$hit: 200,000,000+ sites M$ stock price: > $120 > J > ...and if the US Dept. of Justice succeeds in breaking up M$ before theyI > move to Canada, expect that stock price to more than double, even afterw/ > the inevitable drop at the time of the split.@ >pH > So, someone tell me: under EXACTLY what conditions would expanding theD > OpenVMS user base harm Compaq's stock price or its investors???!!! >iD > > For those who are really concerned about "support", the questionC > > would be how much will DEQ charge for a service contract on VMSf  > > running on an Intel machine. >OI > Until it's officially certified by the Q, consider it unsupported under-
 > Charon-VAX.- >-I > I find it interesting that the "Charon" name was chosen. If I'm not toocE > badly mistaken, that is the character in Greek mythology who is the I > Ferryman on the River Styx - guides (takes) the dead to the underworld, ? > y'know. Strangely prophetic, if true (but hardly unexpected)." >. > -- > David J. Dachtera  > dba DJE Systemsc$ > http://home.earthlink.net/~djesys/ >@< > Unofficial Affordable OpenVMS Home Page and Message Board:- > http://home.earthlink.net/~djesys/vms/soho/m   ------------------------------  # Date: Tue, 13 Jun 2000 15:00:23 GMT.% From: Alan Greig <agreig@my-deja.com>  Subject: Re: VAX on Intel?) Message-ID: <8i5i9r$237$1@nnrp1.deja.com>U  - In article <3945A986.FD0E0421@earthlink.net>, :   "David J. Dachtera" <djesys.nospam@earthlink.net> wrote: >oE > I find it interesting that the "Charon" name was chosen. If I'm notr toorE > badly mistaken, that is the character in Greek mythology who is the = > Ferryman on the River Styx - guides (takes) the dead to thei underworld, ? > y'know. Strangely prophetic, if true (but hardly unexpected).   G Yep and it says on the web site that they picked this name because they B wanted to give him something good to bring back to the land of the living.r   >s > -- > David J. Dachterat > dba DJE Systemst$ > http://home.earthlink.net/~djesys/ >T< > Unofficial Affordable OpenVMS Home Page and Message Board:- > http://home.earthlink.net/~djesys/vms/soho/e >o   -- --
 Alan Greig    & Sent via Deja.com http://www.deja.com/ Before you buy.e   ------------------------------   Date: 13 Jun 2000 17:00:27 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: VAX on Intel?6 Message-ID: <8i5pbb$ck2$1@mailint03.im.hou.compaq.com>  g In article <3945A986.FD0E0421@earthlink.net>, "David J. Dachtera" <djesys.nospam@earthlink.net> writes:dG :So, someone tell me: under EXACTLY what conditions would expanding the C :OpenVMS user base harm Compaq's stock price or its investors???!!!   H   Please make your case -- I'd love to see this port.  In the interrum, H   please see the OpenVMS FAQ, and specifically please see section VMS11.  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Mon, 12 Jun 2000 09:36:49 +0200wE From: "David Andreas Alderud" <aaldv97@student.removethispart.vxu.se>l" Subject: Re: VMS Security features& Message-ID: <8i23e7$k72$1@news.lth.se>  J > > I was not refering to VMS, only to OSs in general, I don't know enough aboutrI > > VMS to make such a statement :). Someone once told me that VMS didn't6 sufferL > > from buffer overruns the way other OS does, so I'm interested in how theJ > > stack is implemented. I'll guess I'll see how it's done when I buy the book.- >-F > Buffer overruns and stack implementations really aren't related, but > that's a separate issue.  F They are. Mybe not in VMS. But the different implementations I know of doesn't give full protection. G Just because you can't execute code on the stack doesn't mean you can'tr9 overrun something with a return adress to malicious code.aJ I don't know how much system cracking you've done, but I know these thingsJ very well I'm sad to admit, that's why I'm writing this report, to do some good. :)L Mybe VMS puts some stack-guards in the stack to, sometimes, make it possibleD to detect overruns, but it doesn't give full protection for apperent reasons.  K > In general, the pages allocated to the stack on VMS aren't executable. If L > you try and execute code on the stack your process will throw an error andA > die. Memory a process allocates is also generally not marked as-H > executable, and attempts to execute randomly allocated data will fail.E > Finally the pages of your process that are executable are generallyR; > marked as read-only, so attempts to alter them will fail.w  K But the important thing, how does VMS protect the executable process memory = space from beeing overwritten with malicious executable data?r Is this what the Monitor does?  A > All the system services that take arbitrary chunks of memory aseI > parameters, be they strings or buffers or whatnot, use a data structure K > called a descriptor to describe the buffer, and one of the pieces of thatPK > descriptor is a buffer length. The system service code (as well as prettySL > much all the languages on VMS except for C) can automagically handle these6 > things and also respect the buffer length parameter.  H I know, but isn't C the most common developent language on VMS, or is it highperformance FORTRAN?I I recomend Ada95, fast development and well protected programs, a bit bigl; runtime enviorment but for what it does it's well worth it.   I Thanks for your input, though, I don't agree with everything you said. :)p   ------------------------------  # Date: Tue, 13 Jun 2000 10:28:50 GMT = From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) " Subject: Re: VMS Security features0 Message-ID: <009EB889.1A16C7A7@SendSpamHere.ORG>  n In article <8i23e7$k72$1@news.lth.se>, "David Andreas Alderud" <aaldv97@student.removethispart.vxu.se> writes:K >> > I was not refering to VMS, only to OSs in general, I don't know enough- >aboutJ >> > VMS to make such a statement :). Someone once told me that VMS didn't >suffereM >> > from buffer overruns the way other OS does, so I'm interested in how thedK >> > stack is implemented. I'll guess I'll see how it's done when I buy the  >book. >>G >> Buffer overruns and stack implementations really aren't related, buta >> that's a separate issue.$ >$G >They are. Mybe not in VMS. But the different implementations I know ofi >doesn't give full protection.H >Just because you can't execute code on the stack doesn't mean you can't: >overrun something with a return adress to malicious code.  G You "might" be able to place code and data on the stack to execute sometG "malicious" code.  Chances are that the malefactor will only get as far-H as affecting the process running the program.  VMS maintains a process/-H job series of bits which describe system level function and/or the scopeI of these system level functions that the process is permitted to perform.uH The "malicious" code would likely fail if the process is not privied to G execute the functions needed to do its dirty work.  Can you or the mal-sF icous code affect these bit?  Not unless the process possesses certain of these privie bits.I  H In addition, the stack has guard pages, as well as, PTE protections thatH would make a stack overwrite into the areas of a more privileged (inner)G mode stack cause an ACCess VIOlation (ACCVIO).  Unless the code runningmF at usermode attempting to write over the stack knew of the location ofH these inner mode stacks and held the appropriate privies and attains theG appropriate access level to write to these stacks, the malevolence of alF stack overrun is likely to be rather benign.  In addition, the "guard"I pages are normally "NA -- No Access".  In otherwords, even the innermost nF privied access mode -- kernel -- cannot write to these pages without aG knowledge of the memory management sub-system and how to muck with the aG PTEs mapping these pages and invalidating any hardware TBs (Translationh  Buffers) related to these pages.    K >I don't know how much system cracking you've done, but I know these things5K >very well I'm sad to admit, that's why I'm writing this report, to do some0	 >good. :)   H Cracking unix is *not* cracking VMS.  You can crack a pistachio nut withH your hands and get inside but the same techniques you've employed to getI at the "kernel" there will not help you if you're working on say a Brazile nut.  M >Mybe VMS puts some stack-guards in the stack to, sometimes, make it possibleoE >to detect overruns, but it doesn't give full protection for apperente	 >reasons.h  G Guard pages and access mode protection.  Both will thwart your attemptseF to get at the inner modes to do damage unless you're privied to do so.  L >> In general, the pages allocated to the stack on VMS aren't executable. IfM >> you try and execute code on the stack your process will throw an error andeB >> die. Memory a process allocates is also generally not marked asI >> executable, and attempts to execute randomly allocated data will fail.dF >> Finally the pages of your process that are executable are generally< >> marked as read-only, so attempts to alter them will fail. >wL >But the important thing, how does VMS protect the executable process memory> >space from beeing overwritten with malicious executable data? >Is this what the Monitor does?i  F You're obviously not going to understand this until you spend a littleG time understanding VMS and, specifically, memory management and the PTE- access mode protection scheme.    B >> All the system services that take arbitrary chunks of memory asJ >> parameters, be they strings or buffers or whatnot, use a data structureL >> called a descriptor to describe the buffer, and one of the pieces of thatL >> descriptor is a buffer length. The system service code (as well as prettyM >> much all the languages on VMS except for C) can automagically handle thesen7 >> things and also respect the buffer length parameter.  > I >I know, but isn't C the most common developent language on VMS, or is its >highperformance FORTRAN?u  ' Gee, it's always been macro for me!  :)m  J >I recomend Ada95, fast development and well protected programs, a bit big< >runtime enviorment but for what it does it's well worth it. >:J >Thanks for your input, though, I don't agree with everything you said. :)   Is it disagreement or denial?  m  G If you are truly looking to write a paper on some aspect of system sec- G utity and you want to include/cover VMS, you'd be best to read some of gF the "wonderful" VMS documentation available or find a time to sit downG with a VMS (and I hate this term) guru and learn some of the nuances ofh3 the operating system used to maintain its security.t   --N VAXman- OpenVMS APE certification number: AAA-0001           VAXman@TMESIS.COM  L GNU Freeware -- What does the GNU *really* stand for?  Garbage!  Not Usable!   ------------------------------  % Date: Tue, 13 Jun 2000 07:31:26 -0400t2 From: "Richard B. Gilbert" <DRAGON@compuserve.com>" Subject: Re: VMS Security features7 Message-ID: <200006130732_MC2-A884-6563@compuserve.com>   G         The program is loaded by the O/S.  The memory, for the portionsaJ containing code and constants, is set to no write access from user mode. =  J If you try, you get an ACCVIO (sigsegv).  Unless your program contains an=  J exception handler, that's the end!   Variables can be overwritten with co= de which can then be executed.t  J         The malicious code is executing in user mode, with only the user'= sRJ privileges.  If the user is not privileged, there's not much evil his cod= eeG or your parasitic code can do.  You can corrupt or delete his files andoF that's about all.  You can't touch the O/S or files belonging to otherA users unless they are foolish enough to grant world write access.-  / Message text written by "David Andreas Alderud":E >But the important thing, how does VMS protect the executable processR memory= space from beeing overwritten with malicious executable data?h Is this what the Monitor does?<E   ------------------------------  % Date: Tue, 13 Jun 2000 07:59:46 -0400 , From: Howard S Shubs <hshubs@mindspring.com>" Subject: Re: VMS Security features> Message-ID: <hshubs-72AC96.07594613062000@news.mindspring.com>  1 In article <009EB889.1A16C7A7@SendSpamHere.ORG>, i system@SendSpamHere.ORG wrote:  H >Guard pages and access mode protection.  Both will thwart your attemptsG >to get at the inner modes to do damage unless you're privied to do so.   I The way to break into a VMS box is to use a priv'd account with a poorly i3 chosen password.  But that's not a technical issue.y  8 Another way would require physical access to the system.   -- a Howard S Shubs, the Denim Adeptb   ------------------------------   Date: 13 Jun 2000 11:15:43 CDT= From: wayne@tachysoft.xxx.044962.killspam.0138 (Wayne Sewell) " Subject: Re: VMS Security features. Message-ID: <8vwqpKujetuH@tachxxsoftxxconsult>  n In article <8i23e7$k72$1@news.lth.se>, "David Andreas Alderud" <aaldv97@student.removethispart.vxu.se> writes:   [stuff deleted]w > L >> In general, the pages allocated to the stack on VMS aren't executable. IfM >> you try and execute code on the stack your process will throw an error and B >> die. Memory a process allocates is also generally not marked asI >> executable, and attempts to execute randomly allocated data will fail. F >> Finally the pages of your process that are executable are generally< >> marked as read-only, so attempts to alter them will fail. > M > But the important thing, how does VMS protect the executable process memoryv? > space from beeing overwritten with malicious executable data?E  > Is this what the Monitor does? >     J The enforcement of readonly memory is not performed by any type of monitorN process.  The architecture handles it at the firmware/microcode/PAL level.  IfL an attempt is made to modify a read-only page, an exception called an accessI violation occurs.  You can trap the exception and prevent your image from ) dying, but the page was *not* modified.  5  O Even from kernel mode, you can't modify a page that is complete readonly, which6O is the case for the operating system code in system space.  Of course, a kernelsN mode stream could change the protection on the page, but it is not easy to get into kernel mode on vms.  J The code in process space is also completely readonly.  Again, even kernel can't modify it.  L Not surprisingly, programs with 70s-mentality self-modifying code don't work' too well on vms, at least not on alpha.n  O If you don't know this stuff, you should not be making *any* comments about the-H hackability of vms.  It just doesn't work the same way as eunuchs and in0 general the eunuchs vulnerabilities don't apply.     Waynej -- RO ===============================================================================eM Wayne Sewell, Tachyon Software Consulting  (281)812-0738  wayne@tachysoft.xxxa: http://www.tachysoft.xxx/www/tachyon.html and wayne.html  K change .xxx to .com in addresses above, assuming you are not a spambot  :-)eO =============================================================================== O Otter, on dining with Bluto:"It's perfectly safe if you keep your arms and legsa 			away from his mouth."   ------------------------------  # Date: Tue, 13 Jun 2000 17:08:43 GMT  From: d.webb@mdx.ac.uk" Subject: Re: VMS Security features) Message-ID: <8i5pqc$6ti$1@nnrp1.deja.com>=  & In article <8i23e7$k72$1@news.lth.se>,H   "David Andreas Alderud" <aaldv97@student.removethispart.vxu.se> wrote:  C > > All the system services that take arbitrary chunks of memory asyA > > parameters, be they strings or buffers or whatnot, use a dataw	 structure H > > called a descriptor to describe the buffer, and one of the pieces of thatF > > descriptor is a buffer length. The system service code (as well as prettyH > > much all the languages on VMS except for C) can automagically handle theset8 > > things and also respect the buffer length parameter. >hG > I know, but isn't C the most common developent language on VMS, or is  it > highperformance FORTRAN?    E Many languages are used for development on VMS. However the importantm? words in the above are "system services" ie the API to call thesD underlying OS routines. All strings or buffers passed into or out of8 system service routines have an associated length count.  B Null terminated strings are strictly forbidden. There was a recentE thread about how some of the new galaxy routines broke this standard. < After the furor about this Compaq are changing the routines.@ (The way they were being used did not actually pose any securityA problem but this was viewed as the thin end of a potentially veryh
 nasty wedge).n  H With most languages this is pretty much transparent. C is the exception.D If you wish to use the system services from C you need to explicitlyD create structures called descriptors to pass in strings and buffers.    6 Another major factor affecting security is privileges.  D VMS has a large number of separate privileges which can be assigned.5 This contrasts with the Unix all or nothing approach. ? Programs can be granted privileges through a mechanism known as' installing the program image.r  # The two facilities are used so that   > 1) A program can be run without giving the user running it any    privileges.G 2) The program need only be given the minimum privileges it needs to doh    it's job.F 3) Programs can turn on and turn off any granted privileges as needed.G 4) When a program exits whether normally or abnormally the system makes:<    sure the user running the program does not gain any extra    privileges.? 5) Daemons (in VMS parlence detached processes) can be run from.>    non-privileged accounts. The process then runs an installed	    image. 3 6) Privileges are needed to install program images.     2 This means that if there is a buffer overflow then  ? 1) If the program crashes (through hitting memory protection egwC    many failures will result in the execution of instructions which)B    reference low virtual address space addresses - the VMS virtual>    address space has guard pages at such addresses which causeB    programs to crash) then the user has not gained any privileges.  F 2) If somehow the program were to survive and the user were to be ableC    to overwrite part of the address space and get their code to run/G    it would at most have the privileges which the program was installed/    with.      
 David Webb VMS and Unix team leader CCSS Middlesex University  G > I recomend Ada95, fast development and well protected programs, a bitj bigR= > runtime enviorment but for what it does it's well worth it.m >.H > Thanks for your input, though, I don't agree with everything you said. :) >. >p    & Sent via Deja.com http://www.deja.com/ Before you buy.L   ------------------------------  % Date: Tue, 13 Jun 2000 13:52:47 -0400o" From: Dan Sugalski <dan@sidhe.org>" Subject: Re: VMS Security features: Message-ID: <4.3.2.7.0.20000613135007.01ca66c0@24.8.96.48>  . At 11:15 AM 6/13/00 -0500, Wayne Sewell wrote:K >The enforcement of readonly memory is not performed by any type of monitor-E >process.  The architecture handles it at the firmware/microcode/PAL zH >level.  If an attempt is made to modify a read-only page, an exception L >called an access violation occurs.  You can trap the exception and prevent 8 >your image from dying, but the page was *not* modified.  L This is enforced by the MMU hardware on the alphas, as it is with any other K processor that has MMU capabilities. (Just about all processors do at this o point)  C I expect some Unices also use hardware memory protection for their  9 executable code, though I don't know for sure which ones.d   					Dan  I --------------------------------------"it's like this"-------------------c2 Dan Sugalski                          even samurai? dan@sidhe.org                         have teddy bears and evene;                                       teddy bears get drunk    ------------------------------  % Date: Tue, 13 Jun 2000 13:35:26 +0000v- From: ezzaoudi med <m.ezzaoudi@digitem.co.ma>uD Subject: Re: Which cpu does Sun server use: I386, Mip, Ppc or Alpha?- Message-ID: <3946389E.2D2656A5@digitem.co.ma>t  F SPARC is the only processor used by SUN for the servers and the stati= onsg runing Solaris.k* There is also a Solaris version for Intel. ____  " lookgood2k@my-deja.com a =E9crit :   > Any help will be appreciated.n >l( > Sent via Deja.com http://www.deja.com/ > Before you buy.I   ------------------------------  # Date: Tue, 13 Jun 2000 16:13:41 GMTr* From: Art Rice <arice.NOcSPAM@ue.itug.org>D Subject: Re: Which cpu does Sun server use: I386, Mip, Ppc or Alpha?8 Message-ID: <68nckskfj0o3lda278pcamg629nubgdadl@4ax.com>  ? On Fri, 09 Jun 2000 14:59:30 GMT, lookgood2k@my-deja.com wrote:o   >Any help will be appreciated. >a >w' >Sent via Deja.com http://www.deja.com/  >Before you buy.   Thought they used SPARC    -- e
 Art Rice   ** # Special Data Processing Corporationt& --------------------------------------' All opinions expressed are mine and do s% not reflect the views of my employer.t   ------------------------------  # Date: Tue, 13 Jun 2000 15:13:56 GMT / From: "John Nixon" <jorlnixon@worldnet.att.net>t Subject: www.compaq.comuF Message-ID: <Ucs15.7757$ov1.557473@bgtnsc04-news.ops.worldnet.att.net>  J Last week there was a discussin thread about people that could not connect
 to COMPAQ.COM  Are there any updates on this?  J I am still having this problem from my office connection.  I can get thereK from my personal dialup session  but not from work.  I can get to all otheraH web sites, but not Compaq.com.  I am probably the only one in my companyG that actually cares about this so it is a low priority with our networkiG group.  Is there any advice I can give them to help me get this ability5	 restored.m   ------------------------------  % Date: Tue, 13 Jun 2000 18:01:02 +0000S- From: ezzaoudi med <m.ezzaoudi@digitem.co.ma>7" Subject: [Fwd: SF36 documentation]- Message-ID: <394676DE.4ED5DD63@digitem.co.ma>u  1 Il s'agit d'un message multivolet au format MIME.>& --------------575873A6418857F64DF3D132* Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit       & --------------575873A6418857F64DF3D132 Content-Type: message/rfc822 Content-Transfer-Encoding: 7bitn Content-Disposition: inlinea   X-Mozilla-Status2: 00000000'- Message-ID: <39325F29.78A904FA@digitem.co.ma>>% Date: Mon, 29 May 2000 12:14:33 +0000t- From: ezzaoudi med <m.ezzaoudi@digitem.co.ma> % X-Mailer: Mozilla 4.7 [fr] (WinNT; I)o X-Accept-Language: frr MIME-Version: 1.0  To: vax <Info-VAX@Mvb.Saic.Com>o Subject: SF36 documentationa* Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit      SF36 documentation   Hi,d& Where can I find documentation for the6 Digital product " SF36-HE " . It is a storage box that can contain  6 RF36 DSSI Disk. It 's an old product !  	 Thank youg    ( --------------575873A6418857F64DF3D132--   ------------------------------   End of INFO-VAX 2000.329 ************************