1 INFO-VAX	Thu, 07 Nov 2002	Volume 2002 : Issue 616       Contents:5 A problem telnet to OpenVMS box and run cobol program  Advanced Server and UAF  Re: Advanced Server and UAF 1 Alpha network performance issues over leased line 5 Re: Alpha network performance issues over leased line 5 Re: Alpha network performance issues over leased line + Re: Alpha versus Vax linker PSECT addresses  Re: CHARON-ALPHA - soon ???? Creating PDF files on VMS  Re: Creating PDF files on VMS  Re: Creating PDF files on VMS  Re: Creating PDF files on VMS  Re: Creating PDF files on VMS  Re: Creating PDF files on VMS  Re: Creating PDF files on VMS  Re: Creating PDF files on VMS H Re: DECtalk A Guide To Voice 1985 First Printing EB 26375 56/06 23 100.0 Detecting non-Files-11 disk  Re: Detecting non-Files-11 disk  Disk usage analysis  Re: Disk usage analysis  Re: Disk usage analysis  Re: Disk usage analysis  Re: Disk usage analysis  Email Filter RE: Email Filter Re: Email Filter# Re: Encompass Electioneering Begins # Re: Encompass Electioneering Begins  Re: ftp questions 
 ftp questions  Re: ftp questions  Re: Goldfax performance/support  Re: Goldfax performance/support ( Re: Here we go again...where is OpenVMS?( Re: Here we go again...where is OpenVMS?( Re: Here we go again...where is OpenVMS?( Re: Here we go again...where is OpenVMS? Re: HP Advocacy Site Re: HP Advocacy Site RE: HP Advocacy Site Re: HP Advocacy Site RE: HP Advocacy Site Re: HP keyboard problems7 Re: Intel Conceding 64-Bit Desktops to IBM, AMD - eWeek % Re: mouse with a scroller for OpenVMS % Re: mouse with a scroller for OpenVMS % Re: mouse with a scroller for OpenVMS % Re: mouse with a scroller for OpenVMS P Northlake Printkit queues in "busy" state, with jobs which say "printing" but no Re: OpenVMS clustering Re: OpenVMS clustering Re: OpenVMS clustering Re: OpenVMS clustering Re: OpenVMS clustering OSU server cache Re: OSU server cache Re: OSU server cache Re: OSU server cache Re: Palladium article < Re: PGFLQUOTA increase on Alpha System VMS with java process& recommend a data/disk recovery service7 Re: SHOW PROCESS/PARENT (was: SHOW PROCESS/PARSE_STYLE)   Re: Slow performance with BACKUP  Re: Slow performance with BACKUP  Re: Slow performance with BACKUP  sysgen>set shadow_sys_disk %4001' trying to replace pathworks with samba.  VaxStation 3100 and M76/SPX  Re: VaxStation 3100 and M76/SPX ( Re: VMS on Itanium not mission critical?0 Re: VMS PHP connect to local Oracle Rdb Database0 Re: VMS PHP connect to local Oracle Rdb Database0 Re: VMS PHP connect to local Oracle Rdb Database7 Re: VMS to support KDE over CDE in which future version  Re: Voting Systems ?$ Re: Was OpenVMS left out on purpose?$ Re: Was OpenVMS left out on purpose?$ Re: Was OpenVMS left out on purpose?$ Re: Was OpenVMS left out on purpose?$ Re: Was OpenVMS left out on purpose?$ Re: Was OpenVMS left out on purpose?$ Re: Was OpenVMS left out on purpose?$ Re: Was OpenVMS left out on purpose?$ Re: Was OpenVMS left out on purpose?! Re: What is DAP status code 5067?   F ----------------------------------------------------------------------  # Date: Thu, 07 Nov 2002 16:53:53 GMT . From: Joseph Norris <jozefn@newbolt.sonic.net>> Subject: A problem telnet to OpenVMS box and run cobol programH Message-ID: <Pine.LNX.4.40.0211070848490.31863-100000@newbolt.sonic.net>   Here is my problem.   C I am using Net::Telnet to telnet into a OpenVms box and run a cobol E program. I have set the script up as below. The Cobol program at this 4 point is a test program that just does two displays.  C The problem is that the displays from the Cobol program do not show : up and so my waitfor's do not make it and the script dies.  G The last entry in the log file just shows the name of the Cobol program ? but it appears that nothing is running and then the perl script , proceeds to the next statement and logs off.  = Is this a buffering problem on the part of the Cobol program? C Is the Cobol program in this OS understanding that it is not really K talking to a telnet session but a pro-grammatically invoked telnet session?   C I can telnet from my box to the vms box and run the program on line  with no problem.    @ I know Cobol and perl but mainly on Hp3000 and Unix boxes - very new to OpenVms.   , Thanks to all who can give me any direction.@ Any and all help would be appreciated - need to get this to run.     use Net::Telnet ();   $username = 'dpsub_1';   $passwd = 'some_pass';     $filename = 'telnet.log';    unlink $filename;  &  $t = new Net::Telnet (Timeout => 60);     $fh = $t->dump_log;      $fh = $t->dump_log($fh);"     $fh = $t->dump_log($filename); print "dump set\n"; '     $t->open("machine.ip.kept.secret");  print "at open\n";B     $t->login(Name=>$username, Password=>$passwd, Prompt=>'/\$/'); print "at dir cmd\n"; !     $t->print("run testcob.exe");  	@lines = $t->getlines;      $t->print("logout");     $t->close;   my $cnt = 0;
 for (@lines){  	$cnt++; 	print "$cnt: $_\n"; }  exit(0);    ; #Joseph Norris (Perl - what else is there?/Linux/CGI/Mysql) I print @c=map chr $_+100,(6,17,15,16,-68,-3,10,11,16,4,1,14,-68,12,1,14,8, C -68,4,-3,-1,7,1,14,-68,-26,11,15,1,12,4,-68,-22,11,14,14,5,15,-90);    ------------------------------  $ Date: Thu, 7 Nov 2002 10:46:53 -0600* From: "Dave Gudewicz" <dgudewicz@core.com>  Subject: Advanced Server and UAF8 Message-ID: <3dca98fc$0$1451$1dc6e903@news.corecomm.net>  G Can't find this in the docs.  Are there any circumstances where an A.S. F (v7.3) w/LanManager *only* security needs to find anything in the UAF?  I I did read that external auth. and the combined VMS and LM security model C needs to look at UAF, but the third model "LM only" is my question.    -- Dave...   G It is noble to teach oneself, but still nobler to teach others-and less  trouble. -----Mark Twain    ------------------------------   Date: 7 Nov 2002 11:14:05 -0600 - From: Kilgallen@SpamCop.net (Larry Kilgallen) $ Subject: Re: Advanced Server and UAF3 Message-ID: <kHsmO99GZWIa@eisner.encompasserve.org>   e In article <3dca98fc$0$1451$1dc6e903@news.corecomm.net>, "Dave Gudewicz" <dgudewicz@core.com> writes: I > Can't find this in the docs.  Are there any circumstances where an A.S. H > (v7.3) w/LanManager *only* security needs to find anything in the UAF?  E I thought it mapped Microsoft identities into VMS identities in order H to know who should be the file owner.  It cannot do that without SYSUAF.   ------------------------------  * Date: Thu, 7 Nov 2002 16:02:18 +0000 (UTC)* From: "Maverick" <maverick902@hotmail.com>: Subject: Alpha network performance issues over leased line0 Message-ID: <aqe2qa$khj$1@sparta.btinternet.com>   Hello,  I We have a strange network performance problem.  We have a 2Mb leased line H terminated at each end by a Cisco 2600 router.   A Digital Alpha DS20 isI present in both locations, along with a network of NT boxes. One Alpha is J loaded with VMS 7.1-2 and the other with 7.2-1.  The network at either end is Switched ethernet.   
 So we have   | NT boxes|   |DS20 (7.2-1)|' ____|___________|______      Location A                     |       |Baystack 350 |        |10/100 switch |       |Cisco 2651 router|                    |                    | !                   |    (2Mb link)                    |                    |      |Cisco 2621 router|          |Cisco Catalyst|       |  3500 switch |$ _________|____________    Location B!          |                      |  | NT boxes|   |DS20 (7.1-2)|    L All boxes can connect over IP, but performance varies dramatically depending on the destination box.   C In summary, performance declines dramatically when the Alpha is the L recipient of the data, and is OK otherwise. The alpha in location B has onlyK recently moved there.  They used to talk quite happily to each other before K this, when the second Alpha was in another location again.  The only things D that have changed are the routers at each end (used to be Nortel ARNH routers) and the seond Alpha now plugs into a Cisco switch rather than a+ Nortel Baystack 350 it plugged into before.   G There appears to be no congestion on the link.  Can anyone think of any K reason why the choice of router (Cisco) would have such an effect, but only I on traffic to the Alphas? Or should I be looking at something else?  Some / measurements on performance are given below.  .   L - An FTP put from the DS20 in A to the DS20 in B is dreadful at about 25kB/sJ - An FTP put from the DS20 in B to the DS20 in A is also not good at about 50kB/sL - An FTP put from any NT box to any NT box (over the link) is  about 170kB/s in both directionsG - An FTP put from an NT box in A to the DS20 in B is again about 25kB/s J - An FTP get from an NT box on one side of the link to a DS20 on the other is about 170kB/sK - Local performance from NT boxes to Alpha's (both PUT and GET) seems OK at  about 4MB/s in both locations    Any help greatly appreciated!!   ------------------------------  $ Date: Thu, 7 Nov 2002 18:21:57 +02005 From: "Leonid Rosenboim" <My_1st_name@Consultant.Com> > Subject: Re: Alpha network performance issues over leased line7 Message-ID: <newscache$06s75h$5t9$1@lnews.actcom.co.il>   > Well, the first thing I'd check is the PHY negotiation between6 the Alpha and the Cat3500 switch. It sometimes happens; that the switch things its Half duplex while the NIC thinks ' its un full deuplex mode or vice versa.      --G ----------------------------------------------------------------------- I    Leonid Rosenboim     Visit:  http://www.masada2000.org/historical.html K        Consultant                Email: my first name at consultant dot com     5 "Maverick" <maverick902@hotmail.com> wrote in message * news:aqe2qa$khj$1@sparta.btinternet.com... > Hello, > K > We have a strange network performance problem.  We have a 2Mb leased line J > terminated at each end by a Cisco 2600 router.   A Digital Alpha DS20 isK > present in both locations, along with a network of NT boxes. One Alpha is L > loaded with VMS 7.1-2 and the other with 7.2-1.  The network at either end > is Switched ethernet.  >  > So we have >  > | NT boxes|   |DS20 (7.2-1)|) > ____|___________|______      Location A  >                    | >       |Baystack 350 |  >       |10/100 switch | >  >     |Cisco 2651 router|  >                   |  >                   | # >                   |    (2Mb link)  >                   |  >                   |  >     |Cisco 2621 router|  >  >       |Cisco Catalyst| >       |  3500 switch |& > _________|____________    Location B# >          |                      |  > | NT boxes|   |DS20 (7.1-2)| >  > D > All boxes can connect over IP, but performance varies dramatically	 depending  > on the destination box.  > E > In summary, performance declines dramatically when the Alpha is the I > recipient of the data, and is OK otherwise. The alpha in location B has  onlyF > recently moved there.  They used to talk quite happily to each other beforeF > this, when the second Alpha was in another location again.  The only thingsF > that have changed are the routers at each end (used to be Nortel ARNJ > routers) and the seond Alpha now plugs into a Cisco switch rather than a- > Nortel Baystack 350 it plugged into before.  > I > There appears to be no congestion on the link.  Can anyone think of any H > reason why the choice of router (Cisco) would have such an effect, but onlyK > on traffic to the Alphas? Or should I be looking at something else?  Some 1 > measurements on performance are given below.  .  > G > - An FTP put from the DS20 in A to the DS20 in B is dreadful at about  25kB/sL > - An FTP put from the DS20 in B to the DS20 in A is also not good at about > 50kB/sF > - An FTP put from any NT box to any NT box (over the link) is  about 170kB/s  > in both directionsI > - An FTP put from an NT box in A to the DS20 in B is again about 25kB/s L > - An FTP get from an NT box on one side of the link to a DS20 on the other > is about 170kB/sJ > - Local performance from NT boxes to Alpha's (both PUT and GET) seems OK at > about 4MB/s in both locations  >   > Any help greatly appreciated!! >  >  >    ------------------------------  * Date: Thu, 7 Nov 2002 16:32:23 +0000 (UTC)* From: "Maverick" <maverick902@hotmail.com>> Subject: Re: Alpha network performance issues over leased line1 Message-ID: <aqe4in$ets$1@knossos.btinternet.com>    Thanks Leonid,  H I had heard this, and  I changed both the switch and the Alpha to manual% 100Mb full duplex with little effect.     @ "Leonid Rosenboim" <My_1st_name@Consultant.Com> wrote in message1 news:newscache$06s75h$5t9$1@lnews.actcom.co.il... @ > Well, the first thing I'd check is the PHY negotiation between8 > the Alpha and the Cat3500 switch. It sometimes happens= > that the switch things its Half duplex while the NIC thinks ) > its un full deuplex mode or vice versa.  >    ------------------------------  $ Date: Thu, 7 Nov 2002 09:54:08 -0500& From: "Ed Vogel" <ed.vogel@compaq.com>4 Subject: Re: Alpha versus Vax linker PSECT addresses. Message-ID: <3dca7e91$1_1@hpb10302.boi.hp.com>  3 "John Reagan" <john.reagan@hp.com> wrote in message , news:yCgy9.9$qm4.1207350@news.cpqcorp.net... | G | The only languages (that I know) which promise layout order are BLISS  | and MACRO. | C     Compaq C (V6.2 and later) provides this ability (on Alpha) via:   :     #pragma extern_model strict_refdef  "<name>" noreorder       Ed Vogel     Compaq C Engineering   ------------------------------  $ Date: Thu, 7 Nov 2002 12:35:50 +01002 From: "Robert Boers" <robert.boers@softresint.com>% Subject: Re: CHARON-ALPHA - soon ???? * Message-ID: <3dca5019$1@news.deckpoint.ch>  E Correct. The IPF translator will translate both 'native' and 'VESTed' 	 binaries.   
 Robert Boers.   5 "Stanley F. Quayle" <stan@stanq.com> wrote in message ) news:3DC70320.17107.24831F87@localhost... 2 > On 4 Nov 2002 at 21:06, David J. Dachtera wrote:( > > Any chance of a VAX->IPF translator? > B > One of the specs is that the Alpha->IPF translator can translate > VAX->Alpha translated images.  >  > --Stan Quayle  > Quayle Consulting Inc. >  > ----------E > Stanley F. Quayle, P.E. N8SQ  +1 614-868-1363  Fax: +1 614 868-1671 3 > 8572 North Spring Ct. NW, Pickerington, OH  43147 ? > Preferred address:  stan@stanq.com       http://www.stanq.com  >  >    ------------------------------  $ Date: Thu, 7 Nov 2002 09:01:36 -0500/ From: "Hank Vander Waal" <hvanderw@mansply.com> " Subject: Creating PDF files on VMS; Message-ID: <000001c28666$30415790$cb96a8c6@manufact5l8vs8>   G I need to create PDF output files from applications running on my Alpha L system.   Does that require additional software or is there an easier way to
 do that ??   Any help would be appreciated.     Hank Vander Waal  / VMS - is there anything that DARES to compare?? ! Unix  - for hackers - by hackers!  You get what you pay for !   ------------------------------  $ Date: Thu, 7 Nov 2002 14:17:24 +00002 From: John Eisenschmidt <jweisen@eisenschmidt.org>& Subject: Re: Creating PDF files on VMS4 Message-ID: <20021107141724.A16342@eisenschmidt.org>  N Unless the Voices are Mistaken, Hank Vander Waal (hvanderw@mansply.com) Wrote:I > I need to create PDF output files from applications running on my Alpha N > system.   Does that require additional software or is there an easier way to > do that ??   http://www.cs.wisc.edu/~ghost/  5 Ghostscript gets hungry -- feed it a Postscript file.     > Any help would be appreciated. >  >  > Hank Vander Waal > 1 > VMS - is there anything that DARES to compare?? # > Unix  - for hackers - by hackers!  > You get what you pay for !   --  / John W. Eisenschmidt <jweisen@eisenschmidt.org> C    Public Key   |  http://www.eisenschmidt.org/jweisen/misc/pgp.asc D    Fingerprint  |  5F9B F916 5AD1 3295 CF99 BC1E 1F97 E6A3 37E3 BEF2J Is this mail an attachment? http://www.jensbenecke.de/misc/outlook.en.html  ; 			"If God were cheese, would we still worship on Sundays?"    ------------------------------   Date: 7 Nov 2002 15:32:09 +0100 ' From: huber@mppmu.mpg.de (Joseph Huber) & Subject: Re: Creating PDF files on VMS+ Message-ID: <N62+W7yh0En$@vms.mppmu.mpg.de>   m In article <000001c28666$30415790$cb96a8c6@manufact5l8vs8>, "Hank Vander Waal" <hvanderw@mansply.com> writes: I > I need to create PDF output files from applications running on my Alpha N > system.   Does that require additional software or is there an easier way to > do that ?? >   M Depending on the format of the applications output, there are freeware tools:   . a2ps     to produce postscript from ascii text, ghostscript   to produce PDF from postscriptG txt2pdf  a simple version posted here in 1998 (find it in /pds/stuff/c/           below my homepage.   G txt2pdf  a commercial tool , see postings from scanface (sic ?) in this           newsgroup.   : You may also want to look into http://www.planetpdf.com/ .   --: Joseph "Sepp" Huber, Max-Planck-Institut Physik,Muenchen, : http://wwwvms.mppmu.mpg.de/~huber                            ------------------------------  % Date: Thu, 07 Nov 2002 09:21:26 -0600 7 From: "Craig A. Berry" <craigberry@nospam.telocity.com> & Subject: Re: Creating PDF files on VMSG Message-ID: <craigberry-6CB777.09212607112002@news.directvinternet.com>   ; In article <000001c28666$30415790$cb96a8c6@manufact5l8vs8>, 1  "Hank Vander Waal" <hvanderw@mansply.com> wrote:   I > I need to create PDF output files from applications running on my Alpha N > system.   Does that require additional software or is there an easier way to > do that ??  G I adapted an program called txt2pdf.c for VMS and posted it previously   int this n.g.  See:   I <http://groups.google.com/groups?q=txt2pdf.c+group:comp.os.vms&hl=en&lr=& I ie=UTF-8&selm=craig.berry-1A37ED.15192930092001%40newsrump.sjc.telocity.n 
 et&rnum=1>   ------------------------------  % Date: Thu, 07 Nov 2002 16:35:56 +0100 9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com> & Subject: Re: Creating PDF files on VMS' Message-ID: <3DCA885C.9E1799BF@aaa.com>   @ I use the PDF "device driver" in Ghostscript to create PDF files= from the PS output from DECdocument. Works great and is fully 9 integrated in my DECdoc "build" scripts. And it's free...    Jan-Erik Sderholm.    Hank Vander Waal wrote:  > I > I need to create PDF output files from applications running on my Alpha N > system.   Does that require additional software or is there an easier way to > do that ?? >   > Any help would be appreciated. >  > Hank Vander Waal > 1 > VMS - is there anything that DARES to compare?? # > Unix  - for hackers - by hackers!  > You get what you pay for !   ------------------------------  # Date: Thu, 07 Nov 2002 15:51:01 GMT * From: Paul Anderson <paul.anderson@hp.com>& Subject: Re: Creating PDF files on VMS5 Message-ID: <071120021050352895%paul.anderson@hp.com>   G In article <000001c28666$30415790$cb96a8c6@manufact5l8vs8>, Hank Vander " Waal <hvanderw@mansply.com> wrote:  C > I need to create PDF output files from applications running on my F > Alpha system.   Does that require additional software or is there an > easier way to do that ??  F Just as some OpenVMS applications create their own PostScript code, an? application could create its own PDF code.  If you already have 9 PostScript code, the suggestions already made would work.    Paul   --    Paul Anderson   OpenVMS Engineering    Hewlett-Packard Company    ------------------------------   Date: 7 Nov 2002 10:58:01 -0600 - From: Kilgallen@SpamCop.net (Larry Kilgallen) & Subject: Re: Creating PDF files on VMS3 Message-ID: <YQebFQDPqqOE@eisner.encompasserve.org>   c In article <3DCA885C.9E1799BF@aaa.com>, Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com> writes: B > I use the PDF "device driver" in Ghostscript to create PDF files? > from the PS output from DECdocument. Works great and is fully ; > integrated in my DECdoc "build" scripts. And it's free...   D Thanks for the suggestion.  I have been moving my Document output to/ the Macintosh and back via DECnet but manually.    ------------------------------  % Date: Thu, 07 Nov 2002 19:45:16 +0100 9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com> & Subject: Re: Creating PDF files on VMS' Message-ID: <3DCAB4BC.DACD5906@aaa.com>   8 Just to add to my previous post, here are my commands to run DECdoc and GS :    $!& $ GS == "$UTIL:[GS.GS-6_01.BIN]GS.EXE" $!  4 $ document mydoc manual.reference ps /contents/index $!G $ gs "-sDEVICE=pdfwrite" "-sOutputFile=mydoc.pdf" "-dNOPAUSE" "-dBATCH"  mydoc.ps $!  1 I can't remember where I got the GS distribution. ; Before installing GS, I tried to convert the PS files using B Adobe Acrobat Destiller on my PC, but it crashed with some message about missing fonts, I think.   ) I'v had no problems with these PDF files.   E Now, there are two things I wish, more speed from my "Jensen" box and F that I'd get my SAMBA install to work better, but that's another story :-)    Regards  Jan-Erik Sderholm.    Larry Kilgallen wrote: > e > In article <3DCA885C.9E1799BF@aaa.com>, Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com> writes: D > > I use the PDF "device driver" in Ghostscript to create PDF filesA > > from the PS output from DECdocument. Works great and is fully = > > integrated in my DECdoc "build" scripts. And it's free...  > F > Thanks for the suggestion.  I have been moving my Document output to1 > the Macintosh and back via DECnet but manually.    ------------------------------   Date: 7 Nov 2002 13:55:00 GMT ( From: bill@cs.uofs.edu (Bill Gunshannon)Q Subject: Re: DECtalk A Guide To Voice 1985 First Printing EB 26375 56/06 23 100.0 5 Message-ID: <aqdrbj$95b05$1@ID-135708.news.dfncis.de>   + In article <aqcq3a$5ar$1@deeptow.ucsd.edu>, / 	cdl@deeptow.ucsd.edu (Carl Lowenstein) writes:  > 8 > Several DECtalk manuals are available as text files at0 > < http://www.tmesis.com/freeware/V5/DECTALK >. > 9 > I don't know if the one you are seeking is one of them. 4 > They are only text files, no illustrative figures.E > Some of them are peculiarly laid out, landscape on 8.5"x5.5" paper.  >   ? I've got the original books that appear on that site.  They are 
 8.5"x5.5".   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------   Date: 7 Nov 2002 07:40:04 -0800 ! From: jsewell@iu.net (Joe Sewell) $ Subject: Detecting non-Files-11 disk= Message-ID: <9e311bd1.0211070740.5012ca69@posting.google.com>   B I've been tasked to create a utility (written in DECwindows/Motif,E though that shouldn't matter here) that can copy hard drives.  I have F the logic in place -- and working, even :) -- to do VMS copies out theD wazoo.  I also have the logic in place to do disks created on a UN*X% system, thanks to help provided here.   C My current roadblock is in detecting when a disk isn't Files-11.  I B thought I could try mounting the disk (using SYS$MOUNT), and if it@ failed, mount it /FOREIGN and force a physical backup.  For someB reason, though, the mount hangs within SYS$MOUNT itself; I presumeE it's hunting down a valid Files-11 homeblock, although there's no I/O 
 on the drive.   @ Short of reading the first physical block on the drive (which, IF presume, would be done by opening a channel directly to the device andF doing a SYS$QIO[W]), is there a better way of detecting a non-Files-11 disk?    ------------------------------   Date: 7 Nov 2002 11:09:54 -0600 - From: Kilgallen@SpamCop.net (Larry Kilgallen) ( Subject: Re: Detecting non-Files-11 disk3 Message-ID: <F$31MpoHOo+P@eisner.encompasserve.org>   a In article <9e311bd1.0211070740.5012ca69@posting.google.com>, jsewell@iu.net (Joe Sewell) writes:   E > My current roadblock is in detecting when a disk isn't Files-11.  I D > thought I could try mounting the disk (using SYS$MOUNT), and if itB > failed, mount it /FOREIGN and force a physical backup.  For someD > reason, though, the mount hangs within SYS$MOUNT itself; I presumeG > it's hunting down a valid Files-11 homeblock, although there's no I/O  > on the drive.  > B > Short of reading the first physical block on the drive (which, IH > presume, would be done by opening a channel directly to the device andH > doing a SYS$QIO[W]), is there a better way of detecting a non-Files-11 > disk?   H I am not sure how reading the first block on the disk would do anything.F It certainly cannot tell the difference between Files-11 and ISO-9660,9 since neither of those define the use of the first block.   B In fact I have had a lot of experience lately with disks that haveD ISO-9660 laid down on top of a formerly Files-11 disk, and MOUNT hasE no trouble mounting it either way.  Attempting to access the Files-11 D data that was replaced by a valid ISO-9660 data structure would leadC to error conditions, but disk formats in general do not insist that / their own "not allocated" blocks are all blank.    ------------------------------   Date: 7 Nov 2002 10:55:28 GMT ' From: Andrew Black <a211@black1.org.uk>  Subject: Disk usage analysis; Message-ID: <Xns92BF6F64D52AAandrewblack1orguk@130.133.1.4>    HiI We have a disk that has a cluster size of 35 but a lot of small files (a  < few blocks).  I suspect that a lot of space is being wasted.  M Is there any tool out there that would analyse how much space is wasted and   G what the consequences of different cluster sizes would be (ie trade of  G because of increased to bitmap.sys etc v. wasted space in small files).   K I realise I can do part of this by doing DIREC/SIZ=ALL / GRAND but thought  F if someone has a program to process the info it was save a bit of time   Cheers.    ------------------------------  % Date: Thu, 07 Nov 2002 13:09:02 +0100 : From: Karl Rohwedder <extern.karl.rohwedder@volkswagen.de>  Subject: Re: Disk usage analysis/ Message-ID: <aqdl4n$btb1@doiweb4.volkswagen.de>    Andrew Black wrote:    > HiJ > We have a disk that has a cluster size of 35 but a lot of small files (a> > few blocks).  I suspect that a lot of space is being wasted. > C > Is there any tool out there that would analyse how much space is   > wasted andH > what the consequences of different cluster sizes would be (ie trade ofI > because of increased to bitmap.sys etc v. wasted space in small files).  > E > I realise I can do part of this by doing DIREC/SIZ=ALL / GRAND but  	 > thought H > if someone has a program to process the info it was save a bit of time > 	 > Cheers.   B The DFU utility has the REPORT function, which provides this data  (used/allocated). < DFU is on the FREEWARE CD or get it via the openvms webpages   --    - mit freundlichen Gruessen | with best regards    Karl Rohwedder   iT-Ingenieurteam Ellernbruch 11 D-38112 Braunschweig   Telefon: +49 (531) 515521  Mobil:   +49 (172) 5434843 Telefax: +49 (531) 515531 . E.Mail:  karl.rohwedder(at)it-ingenieurteam.de   fr Volkswagen AG  K-DOE-11  
 Volkswagen AG  Brieffach 1811 38436 Wolfsburg    Telefon: +49 (5361) 9-28724  Telefax: +49 (5361) 9-71854  http://www.volkswagen.de   ------------------------------  $ Date: Thu, 7 Nov 2002 13:36:43 +0100 From: "labadie" <g.g@127.0.0.1>   Subject: Re: Disk usage analysis* Message-ID: <aqdmp7$2ni$1@web1.cup.hp.com>  B and with Vms 7.3-1, it comes with Vms, you do not have to get it !  > > Cheers. > C > The DFU utility has the REPORT function, which provides this data  > (used/allocated). > > DFU is on the FREEWARE CD or get it via the openvms webpages >  > -- > / > mit freundlichen Gruessen | with best regards  >  > Karl Rohwedder >  > iT-Ingenieurteam > Ellernbruch 11 > D-38112 Braunschweig >  > Telefon: +49 (531) 515521  > Mobil:   +49 (172) 5434843 > Telefax: +49 (531) 515531 0 > E.Mail:  karl.rohwedder(at)it-ingenieurteam.de >  > fr Volkswagen AG 
 > K-DOE-11 >  > Volkswagen AG  > Brieffach 1811 > 38436 Wolfsburg  >  > Telefon: +49 (5361) 9-28724  > Telefax: +49 (5361) 9-71854  > http://www.volkswagen.de >    ------------------------------  % Date: Thu, 07 Nov 2002 09:23:52 -0500 & From: David M Smith <dsmit115@csc.com>  Subject: Re: Disk usage analysis8 Message-ID: <3qtksu08fp5j4uirecept36ho9rhvespcu@4ax.com>  D On 7 Nov 2002 10:55:28 GMT, Andrew Black <a211@black1.org.uk> wrote:  N >Is there any tool out there that would analyse how much space is wasted and  H >what the consequences of different cluster sizes would be (ie trade of H >because of increased to bitmap.sys etc v. wasted space in small files).  N A previous responder said you can use DFU's REPORT function to get an analysisO of the current state. The way I read your question, you would also like to know O what will happen if you reinitialize the disk with a different cluster size and $ reload the same set of files, right?  I Awhile ago, I put together a "throwaway" program to analyze the effect of K setting a different cluster size on a specific disk with its current set of  files. I include the doc below.   O If this looks like something you could use, I can ZIP the source and send it to N you (VMS Fortran). Or, someone here can tell me where/how to post this sort ofC thing -- I don't have a publicly accessible web site or FTP server.     L                                                         DISK_SHOW_USAGE_INFO    M           ___________________________________________________________________   H           DISK_SHOW_USAGE_INFO - Displays actual and computed disk spaceH                                 usage information for used and allocated+                                 disk blocks   L               This tool scans all files on a disk, totaling up the number ofJ               blocks in use and blocks allocated, and computes the minimumL               allocated blocks required for those files based on the current4               or a user-specified disk cluster size.  M           ___________________________________________________________________              Format  (               DISK_SHOW_USAGE_INFO  disk  M           ___________________________________________________________________              Parameters                 disk/               Specifies the disk to be scanned.   M           ___________________________________________________________________              Qualifiers  !               /CLUSTER_SIZE=value L               Specifies the cluster size to be used in computing the minimumJ               required allocation for files on the disk. If not specified,H               the current volume cluster size is used. By specifying al-J               ternate values, you can determine the effect of changing the5               volume cluster size on available space.                  /OUTPUT=filespecL               Allows output to be redirected from the standard SYS$OUTPUT to               a file.                  /STATISTICS (D)                /NOSTATISTICS G               The normal output of this tool ends with a summary of the I               buffered I/O, direct I/O, page faults, CPU and elapsed time H               required to perform the tests. The /NOSTATISTICS qualifier*               suppresses this information.  M           ___________________________________________________________________              Restrictions  3               This tool requires READALL privilege.   M           ___________________________________________________________________              Description   J               This tool is used to scan the files on a disk and determine,J               for those files and a specified cluster size, the amount al-I               located disk space is required to hold those files for that M               cluster size. It also sums the current blocks in use and blocks J               allocated for comparison. A summary report is produced after*               all files have been scanned.  M               The primary use of this tool is to compare the effect that will I               be obtained, for the existing set of files, by changing the K               disk cluster size. This cluster size change might be effected J               either by moving the files in their entirety to a newly ini-L               tialized disk, or by backing up the disk, changing the cluster,               size, and restoring the files.  J               It is important to note that the analysis is based purely onL               the existing files on the disk volume. If, subsequently, filesJ               are added with different size attributes the optimum clusterI               size may change. This is one of the reasons that disk spaceA+               management is so challenging!r  M           ___________________________________________________________________              Location                 xxxxxxxu  M           ___________________________________________________________________P             Promptsi  F           If the disk parameter is omitted, the prompt "Disk:" will beH           displayed. When this prompt is displayed, a full command line,8           including optional qualifiers, can be entered.  M           ___________________________________________________________________r             Return Values0                  o [TO BE SUPPLIED]  M           ___________________________________________________________________e             Examples  -               1. $ DISK_SHOW_USAGE_INFO KITS1iK                 DISK_SHOW_USAGE_INFO of _DSA3: (KITS1) at  9-NOV-2001 15:35e  !                 Command line was:e.                  $ DISK_SHOW_USAGE_INFO  KITS1  6                    Scan completed at  9-NOV-2001 15:36  6                    Available blocks:          177691776                    Maximum files:               8378026                    Cluster size:                    186                    Total files found:             6968@                    Actual blocks in use:      10966389    61.72%@                    Minimum allocation:        11014578    61.99%@                    Minimum wasted blocks:        48189     0.27%@                    Actual blocks allocated:   11015190    61.99%@                    Actual wasted blocks:         48801     0.27%  6                    File open errors:                 0  6                    Buffered I/O count:           219096                    Direct I/O count:              72216                    Page faults:                     366                    Charged CPU time:     0 00:00:05.776                    Elapsed time:         0 00:00:51.16  H                 This example shows a minimal command line, requesting anF                 analysis of disk KITS1 using its current cluster size.J                 From the summary report, this cluster size is 18 as shown.H                 Currently, there are 10966389 blocks in use and 11015190H                 blocks allocated, or 48801 wasted blocks (due to clusterG                 size roundup) as shown in the report. Based on the cur-lI                 rent cluster size, the minimum number of wasted blocks isHJ                 48189, indicating that one or more files are slightly overI                 allocated and could, perhaps, be trimmmed with a SET FILEp                 /TRUNCATE.    8               2. $ DISK_SHOW_USAGE_INFO KITS1 /CLUSTER=5K                 DISK_SHOW_USAGE_INFO of _DSA3: (KITS1) at  9-NOV-2001 15:37p  !                 Command line was:p9                  $ DISK_SHOW_USAGE_INFO  KITS1 /CLUSTER=5p  6                    Scan completed at  9-NOV-2001 15:38  6                    Available blocks:          177691776                    Maximum files:               8378026                    Cluster size:                     56                    Total files found:             6968@                    Actual blocks in use:      10966389    61.72%@                    Minimum allocation:        10981685    61.80%@                    Minimum wasted blocks:        15296     0.09%@                    Actual blocks allocated:   11015190    61.99%@                    Actual wasted blocks:         48801     0.27%  6                    File open errors:                 0  6                    Buffered I/O count:           219096                    Direct I/O count:              72306                    Page faults:                     426                    Charged CPU time:     0 00:00:05.546                    Elapsed time:         0 00:00:51.47  I                 This example shows the same disk as the previous example,nH                 but analyzed with a smaller cluster size for comparison.I                 The number of blocks in use, blocks allocated, and wasted G                 blocks (due to cluster size roundup) match the previous L                 example. However, due to the effect of the computations withI                 the smaller cluster size a smaller value is shown for the L                 minimum allocation and minimum wasted blocks. This shows theH                 effect of changing the disk to the smaller cluster size.I -------------------------------------------------------------------------tI David M. Smith 302.391.8533                       dsmit115 at csc dot com I Computer Sciences Corporation     (Opinions are those of the writer only)(I -------------------------------------------------------------------------    ------------------------------   Date: 7 Nov 2002 10:11:26 -0800t7 From: jones.computer.srv@worldnet.att.net (Daryl Jones)@  Subject: Re: Disk usage analysis= Message-ID: <8a646952.0211071011.66df83db@posting.google.com>h  j Andrew Black <a211@black1.org.uk> wrote in message news:<Xns92BF6F64D52AAandrewblack1orguk@130.133.1.4>... > HiK > We have a disk that has a cluster size of 35 but a lot of small files (a -> > few blocks).  I suspect that a lot of space is being wasted. > O > Is there any tool out there that would analyse how much space is wasted and  $I > what the consequences of different cluster sizes would be (ie trade of rI > because of increased to bitmap.sys etc v. wasted space in small files).d > M > I realise I can do part of this by doing DIREC/SIZ=ALL / GRAND but thought MH > if someone has a program to process the info it was save a bit of time > 	 > Cheers.i   Andrew,   ? Using DIREC/SIZE=ALL/GRAND isn't a bad wait of doing it. If the F difference between what is allocated and used is greater than 10% then1 the disk is a candiate for a cluster size change.i   Daryl Jones.   ------------------------------  $ Date: Thu, 7 Nov 2002 13:35:16 -00004 From: "rationalfan" <ptellep@vectortechnicalinc.com> Subject: Email Filter90 Message-ID: <9iyy9.24$dJ4.77053@news.uswest.net>  E I am looking for some ideals for SPAM filtering on an OpenVMS machinetI running Apache (Compaq Secure Web) with SMTP mail service.  What programsnB are available, or how hard would this be to write myself?  Thanks.   -- Peter Tellep ptellep@vectortechnicalinc.com   ------------------------------  $ Date: Thu, 7 Nov 2002 10:29:12 -0800$ From: Shane Smith <ssmith@icius.com> Subject: RE: Email Filterr0 Message-ID: <01C28648.A0F6C790@sulfer.icius.com>  D Hmmm, I think DELIVER off the freeware CD could probably be used forE this. It allows you to automatically process messages based on rules.r   Shanee   -----Original Message-----9 From: rationalfan [mailto:ptellep@vectortechnicalinc.com]a) Sent: Thursday, November 07, 2002 5:35 AMw To: Info-VAX@Mvb.Saic.ComO Subject: Email Filtert    E I am looking for some ideals for SPAM filtering on an OpenVMS machinee@ running Apache (Compaq Secure Web) with SMTP mail service.  What programsB are available, or how hard would this be to write myself?  Thanks.   -- Peter Tellep ptellep@vectortechnicalinc.com   ------------------------------  % Date: Thu, 07 Nov 2002 19:52:05 +0100 9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com>s Subject: Re: Email FilterC' Message-ID: <3DCAB655.537FE175@aaa.com>_  / Yes, DELIVER is OK for low to moderate volumes. ) Note that DELIVER creates a batch job for1 each mail reveived.a  - Nevertheless, during the last 10-15 years I'vc. implemented a number of application interfaces/ between different platforms where mail was usede- as the "transport" and DELIVER as the tool onK  the VMS side. Never failed me...  + But then, this wasn't realy SPAM filtering.    Jan-Erik Sderholm.a   Shane Smith wrote: > F > Hmmm, I think DELIVER off the freeware CD could probably be used forG > this. It allows you to automatically process messages based on rules.r >  > Shaneo >  > -----Original Message-----; > From: rationalfan [mailto:ptellep@vectortechnicalinc.com]e+ > Sent: Thursday, November 07, 2002 5:35 AMe > To: Info-VAX@Mvb.Saic.Comn > Subject: Email Filter  > G > I am looking for some ideals for SPAM filtering on an OpenVMS machinehB > running Apache (Compaq Secure Web) with SMTP mail service.  What
 > programsD > are available, or how hard would this be to write myself?  Thanks. >  > -- > Peter Tellep  > ptellep@vectortechnicalinc.com   ------------------------------  # Date: Thu, 07 Nov 2002 12:00:36 GMTo1 From: "Terry C. Shannon" <terryshannon@attbi.com>t, Subject: Re: Encompass Electioneering Begins, Message-ID: <EBsy9.70875$bG.51366@rwcrnsc53>  H "Bob Kaplow" <kaplow_r@eisner.encompasserve.org.TRABoD> wrote in message- news:2nb3by37lYCT@eisner.encompasserve.org...cA > In article <3dSx9.43901$bG.39372@rwcrnsc53>, "Terry C. Shannon"M  <terryshannon@attbi.com> writes:H > > Snagged from www.openvms.org. Opinions are those of Ken Farmer. As a. > > Director, I am not entitled to an opinion. >y > ROTFLMAO!s  C Glad to give you a laff riot. The fact remains, I cannot endorse or H electioneeer, nor would I. Whether or not you take this at face value isH your businesss, but be very, very, very careful when you make groundless/ acusations. Free advice, and worth every penny.i   >h3 > > Reminder: Vote for Encompass Board of DirectorslD > > You only have two weeks to vote for Encompass BOD's. OpenVMS.org supportsG > > Clay Denton who really seems to have a grasp on the issues and whatw needs toK > > be done to raise Encompass out of the ashes, not to mention nobody elsem hascE > > put in the same amount of volunteer effort as he has. In additioneG > > OpenVMS.org supports Kristi Browder. Both of the candidates will dov theiraL > > best, in our view, to fight the current bureaucracy that has bogged down) > > Encompass for the past several years.l >pI > Aren't both Clay and Kristi on the current "bogged down" board already?O Howp3 > does reelecting the incumbants solve the problem?2  G Actually, Kristi *WAS* on the BoD, is not now, has not been for a year.   J Ditto for Clay... he lost the election year, was appointed to the board to> handle CETS2002. Has no voting power in this position, though.   >wJ > BTW, I know Clay and think he's a fine person. I don't know what's going onF > with the board, but it clearly isn't representing my views or needs.  ) Well, did you run for the Board for 2003?r > H > OpenDECUS.org, if it existed, would support Dale E. Coy for the board.    He'll have his chance next year.   > 4 > Septemnber 9, 2001: A day that will live in infamy5 > Support Freedom: http://www.indefenseoffreedom.org/r > D >         They that can give up essential liberty to obtain a littleA >         temporary safety deserve neither liberty nor safety. --iC >         Benjamin Franklin Historical Review of Pennsylvania. 1759n   ------------------------------  # Date: Thu, 07 Nov 2002 12:21:48 GMTr1 From: "Terry C. Shannon" <terryshannon@attbi.com>h, Subject: Re: Encompass Electioneering Begins> Message-ID: <wVsy9.115401$bt.179381@rwcrnsc52.ops.asp.att.net>  H "Bob Kaplow" <kaplow_r@eisner.encompasserve.org.TRABoD> wrote in message- news:2nb3by37lYCT@eisner.encompasserve.org... A > In article <3dSx9.43901$bG.39372@rwcrnsc53>, "Terry C. Shannon"i  <terryshannon@attbi.com> writes:H > > Snagged from www.openvms.org. Opinions are those of Ken Farmer. As a. > > Director, I am not entitled to an opinion. >e > ROTFLMAO!e  C Glad to give you a laff riot. The fact remains, I cannot endorse or H electioneeer, nor would I. Whether or not you take this at face value isH your businesss, but be very, very, very careful when you make groundless/ acusations. Free advice, and worth every penny.v   >e3 > > Reminder: Vote for Encompass Board of Directors D > > You only have two weeks to vote for Encompass BOD's. OpenVMS.org supportsG > > Clay Denton who really seems to have a grasp on the issues and whats needs toK > > be done to raise Encompass out of the ashes, not to mention nobody else  hastE > > put in the same amount of volunteer effort as he has. In addition!G > > OpenVMS.org supports Kristi Browder. Both of the candidates will do6 their+L > > best, in our view, to fight the current bureaucracy that has bogged down) > > Encompass for the past several years.e >gI > Aren't both Clay and Kristi on the current "bogged down" board already?r Hows3 > does reelecting the incumbants solve the problem?u  G Actually, Kristi *WAS* on the BoD, is not now, has not been for a year.n  J Ditto for Clay... he lost the election year, was appointed to the board to> handle CETS2002. Has no voting power in this position, though.   >bJ > BTW, I know Clay and think he's a fine person. I don't know what's going onF > with the board, but it clearly isn't representing my views or needs.  ) Well, did you run for the Board for 2003?- >-H > OpenDECUS.org, if it existed, would support Dale E. Coy for the board.    He'll have his chance next year.   > 4 > Septemnber 9, 2001: A day that will live in infamy5 > Support Freedom: http://www.indefenseoffreedom.org/o >wD >         They that can give up essential liberty to obtain a littleA >         temporary safety deserve neither liberty nor safety. --SC >         Benjamin Franklin Historical Review of Pennsylvania. 1759l   ------------------------------   Date: 7 Nov 2002 07:09:29 -0600e From: briggs@encompasserve.org Subject: Re: ftp questions3 Message-ID: <$eHfDVyxKcEp@eisner.encompasserve.org>   T In article <aqdmrr$qeq$1@news1.kornet.net>, "David Lee" <phongle@kornet.net> writes:J > I have a couple networking questions that I wonder if anyone can give meF > some guidance here.  I have a PC running win2k connected to my Alpha< > workstations running VMS 7.2 (with TCPIP Ver 5 installed). > N > 1) From my PC, I can ftp to the Alpha workstations and able to use the "get"M > command to get files from it to the PC, but unable to use the "put" command M > to transfer file from my PC to the Alpha workstations, why?? The only thingsJ > I can think of is "the account that I am using on my PC, do not exist onF > Alpha workstations"  could it be the case? how do I get around that?  ! What error message(s) do you get?k  F My guess would be a file protection issue.  The user name and password@ that you are specifying when the ftp client prompts you for userC name and password does not have write access to the Alpha directoryp= that you are referencing.  If you are using some hand-holding0< GUI client, perhaps it is defaulting you into an "anonymous"; login on the Alpha.  That would result in read-only access,t< probably enforced at the ftp server level and never reaching) the point of file system security checks.   > The account that you are using on your PC is irrelevant unless? it does not have read access to the files you are transferring.   A If your PC files have sufficiently baroque syntax (e.g. spaces inN3 the file name), that could conceivably be an issue.s  > Look at file and directory protection with $ DIR /SECURITY and6 look at audit alarms with $ REPLY /ENABLE=SECURITY and2 $ SET AUDIT /ALARM /ENABLE=FILE_ACCESS=FAILURE=ALL  + (Hope I remembered the syntax right there).o  N > 2) From the Alpha workstations, I can not rlogin, telnet, nor ftp to the PC.J > I can do the "ping" from either the Alpha or the PC and it executes just# > fine, what am I doing wrong here?e  C If your PC is not running an rlogin server, you can't use rlogin toa connect to it.  B If your PC is not running a telnet server, you can't use telnet to connect to it.  E If your PC is not running an ftp server, you can't use ftp to connecti to it.  G Out of the box, Microsoft PCs do not run rlogin, telnet or ftp servers.o  B ICMP Echo reply service is built essentially every TCP stack thereB is so PING tends to work without you having to install any special server software.   	John Briggs   ------------------------------  $ Date: Thu, 7 Nov 2002 22:06:20 +0900& From: "David Lee" <phongle@kornet.net> Subject: ftp questions+ Message-ID: <aqdmrr$qeq$1@news1.kornet.net>:  H I have a couple networking questions that I wonder if anyone can give meD some guidance here.  I have a PC running win2k connected to my Alpha: workstations running VMS 7.2 (with TCPIP Ver 5 installed).  L 1) From my PC, I can ftp to the Alpha workstations and able to use the "get"K command to get files from it to the PC, but unable to use the "put" command K to transfer file from my PC to the Alpha workstations, why?? The only thingoH I can think of is "the account that I am using on my PC, do not exist onD Alpha workstations"  could it be the case? how do I get around that?  L 2) From the Alpha workstations, I can not rlogin, telnet, nor ftp to the PC.H I can do the "ping" from either the Alpha or the PC and it executes justK fine, what am I doing wrong here?  I can see the PC ip address listed in myeK host database.  I wonder if there is anything else that I need to do on thep- Alpha workstations besides doing the follwing-) $tcpip set host my_pc/address=193.15.55.9   
 Any ideas?   Thanks   ------------------------------   Date: 7 Nov 2002 06:48:03 -0600C- From: koehler@encompasserve.org (Bob Koehler)r Subject: Re: ftp questions3 Message-ID: <0HyzRDAU8Bfn@eisner.encompasserve.org>4  T In article <aqdmrr$qeq$1@news1.kornet.net>, "David Lee" <phongle@kornet.net> writes:J > I have a couple networking questions that I wonder if anyone can give meF > some guidance here.  I have a PC running win2k connected to my Alpha< > workstations running VMS 7.2 (with TCPIP Ver 5 installed). > N > 1) From my PC, I can ftp to the Alpha workstations and able to use the "get"M > command to get files from it to the PC, but unable to use the "put" commandAM > to transfer file from my PC to the Alpha workstations, why?? The only thing-J > I can think of is "the account that I am using on my PC, do not exist onF > Alpha workstations"  could it be the case? how do I get around that?  H    I do this all the time.  You need to give us the exact error messagesH    if we're going to figure out what's going wrong on your systems.  AreF    the files actually not transfered, or do you have problems with theG    file after transfer?  If the latter it's most likely the "automatic" 7    setting is getting binary vs. ASCII transfers wrong.u  N > 2) From the Alpha workstations, I can not rlogin, telnet, nor ftp to the PC.J > I can do the "ping" from either the Alpha or the PC and it executes justM > fine, what am I doing wrong here?  I can see the PC ip address listed in myxM > host database.  I wonder if there is anything else that I need to do on the-/ > Alpha workstations besides doing the follwing + > $tcpip set host my_pc/address=193.15.55.9o  G    You probably don't have rlogin, telnet, or ftp servers (aka daemons)%    running on the PC.    ------------------------------  $ Date: Thu, 7 Nov 2002 08:51:15 -00002 From: "Chris Sharman" <chris.sharman@sorry.nospam>( Subject: Re: Goldfax performance/support4 Message-ID: <aqd9f4$f9a$1$8300dec7@news.demon.co.uk>  = "JF Mezei" <jfmezei.spamnot@vl.videotron.ca> wrote in messageu) news:3DC9DEAE.C6A807F4@vl.videotron.ca...r > Chris Sharman wrote:J > > 1) The postscript queue appears to have a memory leak, which causes it toE > > thrash after a few hundred jobs. Initially it works fine, rippingn several L > > pages per second. Performance deteriorates however - I've seen it taking all ( > > of a DS20E and around 1 minute/page. >g > Is it based on ghostscript ?   Yes, that's right.   ------------------------------  % Date: Thu, 07 Nov 2002 07:58:24 -0500 ) From: Ken Garber <kgarber@stclairc.on.ca> ( Subject: Re: Goldfax performance/support. Message-ID: <3DCA6370.AF43CEDB@stclairc.on.ca>   Bob Ceculski wrote:r  p > "Chris Sharman" <chris.sharman@sorry.nospam> wrote in message news:<aqbk54$rak$1$8300dec7@news.demon.co.uk>...N > > I've a couple of Goldfax performance issues - anyone out there still using8 > > it on VMS, or has it become a windoze only product ? > >l  G We use it only on the cluster.  The Windows client has never been used.0  * The modem is connected to a DecServer 200.   Kena   ------------------------------  # Date: Thu, 07 Nov 2002 11:28:47 GMT.. From: peter@langstoeger.at (Peter LANGSTOEGER)1 Subject: Re: Here we go again...where is OpenVMS? 5 Message-ID: <P7sy9.182158$aa2.2167185@news.chello.at>   Y In article <aqbg59$mmd$1@aquila.mdx.ac.uk>, david20@alpha1.mdx.ac.uk (David Webb) writes:eb >In article <wyhoSx0E3DgW@eisner.encompasserve.org>, young_r@encompasserve.org (Rob Young) writes: >>E >>	They made the obviously correct decision.  HP-UX has a much largero@ >>	installed base and more applications.  They are able to take I >>	technically superiour Tru64 features (ADVfs, etc.) and swing that over, >>	to HP-UX.   >>A >>	It was/is far more important to keep the HP-UX customers happyB >>	from a business standpoint. >sN >No they made the simplest decision. Applications are just applications. They 9 >have to be ported anyway since your going to a new chip.iM >Using a more modern, more secure Kernel would have been the better solution.v  H Exactly. And finally don't forget the world is becoming Little-Endian...   -- r Peter "EPLAN" LANGSTOEGER % Network and OpenVMS system specialist. E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  # Date: Thu, 07 Nov 2002 13:31:28 GMT-. From: peter@langstoeger.at (Peter LANGSTOEGER)1 Subject: Re: Here we go again...where is OpenVMS?a5 Message-ID: <QWty9.183377$aa2.2182673@news.chello.at>2  b In article <3dc967a2$1@news.si.com>, "Brian Tillman" <tillman_brian@notnoone.notnohow.com> writes:H >>Dag gone it!  That teenager son of mine changed my newsgroup id again! >He'slL >>been bugging me about a new PC for Christmas.  If he keeps this up I think >>he's gonna get one.o > = >In my house, he'd be minus a hand (figuratively, of course).l  I Here, he has his own account (still without a password - because he is 5)t
 instead...   -- x Peter "EPLAN" LANGSTOEGERt% Network and OpenVMS system specialist  E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  # Date: Thu, 07 Nov 2002 12:41:00 GMTt1 From: "Terry C. Shannon" <terryshannon@attbi.com>n1 Subject: Re: Here we go again...where is OpenVMS?.9 Message-ID: <wbty9.490$Dc4.168@rwcrnsc51.ops.asp.att.net>2  $ <kenrbnsn1@rcn.com> wrote in message2 news:410-220021125204730211@M2W072.mail2web.com...I >Although it's good to see Tru64 get some well deserved credit...where is 	 >OpenVMS?r >e >u1 >InternetWeek: Mac OS, SCO, Tru64 Are Safest OSesh1 >http://www.internetwk.com/story/INW20021101S0002h  L Take it from the source from which it comes. Most of the younger "reporters" don't know what VMS is...    ------------------------------  # Date: Thu, 07 Nov 2002 17:38:16 GMTe# From: "John Smith" <a@nonymous.com>t1 Subject: Re: Here we go again...where is OpenVMS?oF Message-ID: <cyxy9.8090$oRV.3829@news04.bloor.is.net.cable.rogers.com>  < "Terry C. Shannon" <terryshannon@attbi.com> wrote in message3 news:wbty9.490$Dc4.168@rwcrnsc51.ops.asp.att.net...  >s& > <kenrbnsn1@rcn.com> wrote in message4 > news:410-220021125204730211@M2W072.mail2web.com...K > >Although it's good to see Tru64 get some well deserved credit...where iso > >OpenVMS?  > >n > >i3 > >InternetWeek: Mac OS, SCO, Tru64 Are Safest OSesr3 > >http://www.internetwk.com/story/INW20021101S0002  >-B > Take it from the source from which it comes. Most of the younger "reporters"p > don't know what VMS is...     E That that's just the visible symptom of the underlying problem of note advertising and marketing.   ------------------------------   Date: 7 Nov 2002 06:42:03 -0600t- From: koehler@encompasserve.org (Bob Koehler)k Subject: Re: HP Advocacy Siten3 Message-ID: <UF0w++Lqyc5V@eisner.encompasserve.org>d  O In article <4jmy9.19988$sP2.7562@sccrnsc02>, "Tom M" <kryios@attbi.com> writes:bM > The most obvious problem with accessing RMS files through JDBC is that somepK > sort of data dictionary is required.  JDBC simply allows you to issue SQL/L > statements against a database so you need a means of defining column namesH > and data types for fields in the RMS files.  A programmer viewing dataM > through JDBC would also expect to be connected to an engine that could join-G > data from different files.  Supporting a JDBC connection to RMS filess8 > essentially requires that you construct a DBMS engine.  E    In that case, Andrew (as usual) is wrong.  There is no standard in D    the Java world crying out for any RMS access not already provided    by the JDK.   ------------------------------  * Date: Thu, 7 Nov 2002 15:13:43 +0000 (UTC)+ From: david20@alpha1.mdx.ac.uk (David Webb)  Subject: Re: HP Advocacy Site + Message-ID: <aqdvv7$ie6$1@aquila.mdx.ac.uk>r  O In article <4jmy9.19988$sP2.7562@sccrnsc02>, "Tom M" <kryios@attbi.com> writes: L >The most obvious problem with accessing RMS files through JDBC is that someJ >sort of data dictionary is required.  JDBC simply allows you to issue SQLK >statements against a database so you need a means of defining column namesoG >and data types for fields in the RMS files.  A programmer viewing datatL >through JDBC would also expect to be connected to an engine that could joinF >data from different files.  Supporting a JDBC connection to RMS files7 >essentially requires that you construct a DBMS engine.c >t/ Well PERL manages with the DBD and DBI modules.eG You can even manipulate a flat file as a database using the CSV module.a  J As far as I'm aware noone has written a DBD module to handle indexed files/ but there is the VMS::INDEXEDFILE module which oD "implements DBM-like functionality using OpenVMS RMS Indexed files."  N I don't see that you actually need a separate data dictionary just some way ofD associating column names with positions in the indexed files records4 (just as with pretty much any programming language).    M I admit I don't know very much about JDBC but I don't see why there should bes< any insuperable problems in developing a JDBC RMS connector.  
 David Webb VMS and Unix team leader CCSS Middlesex University     >d; >"Bob Koehler" <koehler@encompasserve.org> wrote in message . >news:jPloAi5B9mc8@eisner.encompasserve.org...G >> In article <3DC8F992.2070308@nospamn.sun.com>, Andrew Harrison SUNUKf >Consultancy writes: >> >I >> > I don't think that RMS is the same as a DBMS it is more analagous toi >C-ISAM. >> >G >> > But even at this low level its clearly the responsibility of RMS'shG >> > supplier to provide ODBC or JDBC support for it. It is yet anothertE >> > example of HPQ not providing proper support for Java on OpenVMS.h >> >E >> > One of OpenVMS's apparent selling points is RMS, but you have to D >> > get a third party to provide support or code it yourself if you2 >> > want to access RMS intelligently from Oracle. >>I >>    Unfortunately we're used to it.  There's no simple interface to RMSeJ >>    features from C like there is from Fortran.  Now that's an extensionJ >>    to Fortran, but one that at least Solaris and HP-UX support in their
 >>    f77. >>K >>    We're lucky enough that you can access these features from any nativesJ >>    language, but Java doesn't currently compile to native instructions. >>E >>    On the other hand C programmers haven't missed what they're not F >>    accustomed to.  No reason Java programmers should really miss it
 >>    either.  >>H >>    VMS has always simply used RMS to support the standard features ofG >>    the language, with IIRC the exception of the now defacto industry G >>    standard Fortran extensions.  You want to program in C, you get CoD >>    I/O, you want to program in Java, you get Java I/O, ...; never4 >>    mind RMS is their underneath the language RTL. >>I >>    But I do agree that a bevcoming-standard interface like JDBC should / >>    be supported by RMS if RMS is sufficient.1 >>I >>    Anybody know RMS and JDBC well enough to know if RMS is sufficient?- >> >> >- >-   ------------------------------  $ Date: Thu, 7 Nov 2002 08:16:13 -0800$ From: Shane Smith <ssmith@icius.com> Subject: RE: HP Advocacy Site30 Message-ID: <01C28635.F1540800@sulfer.icius.com>  G There are already products that do this, ConnX (sp?) comes to mind. TherD availability of these may well be why the VMS guys haven't spent the time doing one themselves.   Shane    -----Original Message-----% From: Tom M [mailto:kryios@attbi.com].* Sent: Wednesday, November 06, 2002 8:51 PM To: Info-VAX@Mvb.Saic.Como Subject: Re: HP Advocacy Site     F The most obvious problem with accessing RMS files through JDBC is that someE sort of data dictionary is required.  JDBC simply allows you to issuea SQLeD statements against a database so you need a means of defining column names F and data types for fields in the RMS files.  A programmer viewing dataF through JDBC would also expect to be connected to an engine that could joinE data from different files.  Supporting a JDBC connection to RMS files 6 essentially requires that you construct a DBMS engine.    : "Bob Koehler" <koehler@encompasserve.org> wrote in message- news:jPloAi5B9mc8@eisner.encompasserve.org...eF > In article <3DC8F992.2070308@nospamn.sun.com>, Andrew Harrison SUNUK Consultancy writes:_ > > H > > I don't think that RMS is the same as a DBMS it is more analagous to C-ISAM.  > > F > > But even at this low level its clearly the responsibility of RMS'sF > > supplier to provide ODBC or JDBC support for it. It is yet anotherD > > example of HPQ not providing proper support for Java on OpenVMS. > >nD > > One of OpenVMS's apparent selling points is RMS, but you have toC > > get a third party to provide support or code it yourself if you 1 > > want to access RMS intelligently from Oracle.t >cH >    Unfortunately we're used to it.  There's no simple interface to RMSI >    features from C like there is from Fortran.  Now that's an extension I >    to Fortran, but one that at least Solaris and HP-UX support in their_	 >    f77._ >_J >    We're lucky enough that you can access these features from any nativeI >    language, but Java doesn't currently compile to native instructions.  > D >    On the other hand C programmers haven't missed what they're notE >    accustomed to.  No reason Java programmers should really miss it  >    either. > G >    VMS has always simply used RMS to support the standard features ofaF >    the language, with IIRC the exception of the now defacto industryF >    standard Fortran extensions.  You want to program in C, you get CC >    I/O, you want to program in Java, you get Java I/O, ...; never 3 >    mind RMS is their underneath the language RTL.  > H >    But I do agree that a bevcoming-standard interface like JDBC should. >    be supported by RMS if RMS is sufficient. > H >    Anybody know RMS and JDBC well enough to know if RMS is sufficient? >  >m   ------------------------------  % Date: Thu, 07 Nov 2002 16:12:29 +0000I' From: Andrew Harrison SUNUK Consultancys Subject: Re: HP Advocacy Sited. Message-ID: <3DCA90ED.4030907@nospamn.sun.com>   Bob Koehler wrote:Q > In article <4jmy9.19988$sP2.7562@sccrnsc02>, "Tom M" <kryios@attbi.com> writes:_ > M >>The most obvious problem with accessing RMS files through JDBC is that somesK >>sort of data dictionary is required.  JDBC simply allows you to issue SQL_L >>statements against a database so you need a means of defining column namesH >>and data types for fields in the RMS files.  A programmer viewing dataM >>through JDBC would also expect to be connected to an engine that could joinkG >>data from different files.  Supporting a JDBC connection to RMS filesi8 >>essentially requires that you construct a DBMS engine. >  > G >    In that case, Andrew (as usual) is wrong.  There is no standard in F >    the Java world crying out for any RMS access not already provided >    by the JDK. >   1 In that case the commercial providers of JDBC forr+ RMS have engineered their product to meet a  requirement that doesn't exist.T  ) Oh dear, either they are wrong or you are    regards  Andrew Harrisont   ------------------------------   Date: 7 Nov 2002 11:11:23 -0600s- From: Kilgallen@SpamCop.net (Larry Kilgallen)t Subject: RE: HP Advocacy Sitet3 Message-ID: <bGX$XkF9EvN6@eisner.encompasserve.org>n  W In article <01C28635.F1540800@sulfer.icius.com>, Shane Smith <ssmith@icius.com> writes:mI > There are already products that do this, ConnX (sp?) comes to mind. TherF > availability of these may well be why the VMS guys haven't spent the > time doing one themselves.  @ It is to the disadvantage of VMS if VMS Development were to take? on the task of driving all the third party software vendors out  of the VMS business.   ------------------------------  % Date: Thu, 07 Nov 2002 08:16:55 -0500_2 From: rdeininger@mindspring.com (Robert Deininger)! Subject: Re: HP keyboard problemsoJ Message-ID: <rdeininger-0711020816550001@1cust27.tnt3.nashua.nh.da.uu.net>  < In article <01C285B5.BA988E10@sulfer.icius.com>, Shane Smith <ssmith@icius.com> wrote:   @ >Here's an interesting one. This new LK463 isn't wireless is it? >_G >http://www.aftenposten.no/english/local/article.jhtml?articleID=431326  > F >Apparently a couple of guys in Norway have HP wireless keyboards thatG >can type on each others' machines 150 metres (roughly yards) apart. SoSF >HP replaced them, and it's still happening. Not something you want to >type passwords on...   8 LK463 is USB and (using an included adapter widget) PS2.   ------------------------------  % Date: Thu, 07 Nov 2002 16:32:36 +0000 ' From: Andrew Harrison SUNUK Consultancy @ Subject: Re: Intel Conceding 64-Bit Desktops to IBM, AMD - eWeek, Message-ID: <3DCA95A4.20902@nospamn.sun.com>   Bill Todd wrote:5 > "Rick Jones" <foo@bar.baz.invalid> wrote in message & > news:aq9fn4$ca4$1@web1.cup.hp.com... >  > ...7 >  > ? >>Given that SPECweb99 was not obsoleted when SPECweb99_SSL was H >>announced (as SPECweb96 appears to have done), thus giving overlap andF >>forcing submittors to make zero-sum resource choices, and given thatE >>the number of SPECweb99* results published in the first 7 months oflH >>each benchmark's life is roughly the same, and the non-trivial overlapB >>in processor families, it would seem that the two benchmarks are6 >>indeed of the same order of magnitude of popularity. >  > I > Just to complete this particular topic:  not exactly.  19 of the 28 SSLaK > submissions came from a single vendor - HP - and all but one of them weremL > for PA-RISC or Itanic2:  flooding by a single vendor is hardly any kind ofI > broad proof of 'popularity', however you may wish to define it.  And ifsM > you're inclined to respond that about half of all the non-SSL tests are foriI > some flavor of IA32, note that they are spread across platforms from ats: > least 5 different vendors and a range of IA32 sub-types. >   D I am still not sure what the point is anyway. Most people who have a@ SSL requirement put a Crypto accelerator into their web servers.  E They may need to regardless of the levels of SSL traffic through the oH site because they may also need HSM's many of which double up as Crypto 
 accelerators.   H Sun's crypto does 4300 SSL operations per second and costs 2,700 dollarsG a much better deal if you have a very heavy SSL requirment than fillingi your system with more CPU's.   RegardsA Andrew HarrisonL   ------------------------------  % Date: Thu, 07 Nov 2002 09:18:51 +0100  From: munk@home.nl. Subject: Re: mouse with a scroller for OpenVMS8 Message-ID: <5b7ksuk8qp8g31htn6shh0l8adutif9e0c@4ax.com>   Forrest & Kerry,  A There is something I don't understand. How many LK keyboards does8D Compaq sell for direct connection to a VMS system ? It can't be thatF many ! The keyboards sold with a workstation will most likely be used,E the keyboards sold with servers will hardly be used, just for booting 
 purposes etc.c  D Now how do most users connect to a VMS systems these days ? With theD use of a terminal emulator on a PC. And those users are stuck with a? crippled keyboard layout that is difficult to use with many VMS 
 applications.   D So on the one hand we have Compaq producing VMS software that reliesC on the use of a 'special' keyboard layout, and on the other hand we:D have Compaq telling customers who want to use that software with theC help of a terminal emulator on a PC (as 90% will be doing) "we willoD not give you a keyboard (driver) that will make that software easilyF accessible, even if we have a suitable keyboard". Somehow that doesn't sound very logical to me.    Regards,   Dirk        0 On Wed, 06 Nov 2002 22:59:45 GMT, Forrest Kenney# <Forrest.Kenney@hp.com.doom> wrote:f   >Shane,c >uL >    Thanks for the offer I know what values are being sent to the PC by the	 >keyboard K >Fred and I approved them when the specification was written for the LK463.  >It is what L >the default Microsoft USB keyboard driver that then decides what to do with	 >them.  I-J >am fairly certain that for things it has not mapping for it ignores them. > H >    We have 0 plans to put any effort into making the LK463 work on any
 >thing butJ >VMS.  Not that it would not be nice to have it work for Tru64 or for a PC
 >just thatK >we have way more work than we can do already.  The company views this as a1 >specialH >one off keyboard just for VMS.  If someone want to try and make it work
 >better undersD >windows Fred and I will gladly give them the information they need. >r >B >Forrest Kenney. >OpenVMS Development   ------------------------------  + Date: Thu, 07 Nov 2002 09:51:25 +0100 (MET)k9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com>-. Subject: Re: mouse with a scroller for OpenVMS; Message-ID: <01KOKYS6RC8UA23BZ7@sysdev.deutsche-boerse.com>d  C > There is something I don't understand. How many LK keyboards doesfF > Compaq sell for direct connection to a VMS system ? It can't be thatH > many ! The keyboards sold with a workstation will most likely be used,G > the keyboards sold with servers will hardly be used, just for bootingh > purposes etc.r  @ They might be rarely used, but a) they are sold and b) they are 
 necessary.  F > Now how do most users connect to a VMS systems these days ? With theF > use of a terminal emulator on a PC. And those users are stuck with aA > crippled keyboard layout that is difficult to use with many VMS- > applications.   F Sic transit gloria mundi.  However, is this really the case?  I don't  know.c  F On a related note, I was recently at a furniture store (not Ikea) and G was pleasantly surprised to see a VT 520 and an LK411 at each checkout MH stand.  I'm told that Ikea has a similar system.  The Dutch police also D use VT's and LK keyboards.  Thus, there might be a lot of "terminal  folks" out there, even today.    ------------------------------  % Date: Thu, 07 Nov 2002 15:14:52 +0100h From: munk@home.nl. Subject: Re: mouse with a scroller for OpenVMS8 Message-ID: <74tksug81cimk0j92t09s0884dbacnr6c5@4ax.com>  8 On Thu, 07 Nov 2002 09:51:25 +0100 (MET), Phillip Helbig+ <HELBPHI@sysdev.deutsche-boerse.com> wrote:w  D >> There is something I don't understand. How many LK keyboards doesG >> Compaq sell for direct connection to a VMS system ? It can't be thataI >> many ! The keyboards sold with a workstation will most likely be used,iH >> the keyboards sold with servers will hardly be used, just for booting >> purposes etc. >MA >They might be rarely used, but a) they are sold and b) they are c >necessary.g   No dispute there !   > G >> Now how do most users connect to a VMS systems these days ? With the@G >> use of a terminal emulator on a PC. And those users are stuck with a B >> crippled keyboard layout that is difficult to use with many VMS >> applications. >jG >Sic transit gloria mundi.  However, is this really the case?  I don't D >know. >KG >On a related note, I was recently at a furniture store (not Ikea) and 1H >was pleasantly surprised to see a VT 520 and an LK411 at each checkout I >stand.  I'm told that Ikea has a similar system.  The Dutch police also  E >use VT's and LK keyboards.  Thus, there might be a lot of "terminal h >folks" out there, even today.  F Very true, but mostly because people working with those terminals willC only use that specific application. Workers in a office environmento> will need all kind of other applications, and they will have a" billybox with a terminal emulator.   ------------------------------  + Date: Thu, 07 Nov 2002 15:37:50 +0100 (MET)e9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com>2. Subject: Re: mouse with a scroller for OpenVMS; Message-ID: <01KOLAYZAHI49ZLF2M@sysdev.deutsche-boerse.com>S  H > Very true, but mostly because people working with those terminals willJ > only use that specific application. Workers in a office environment willI > need all kind of other applications, and they will have a billybox witht > a terminal emulator. -  H You can do it one way or the other.  I remember once using WORD under a B Citrix Metaframe Windows emulation---from an LK keyboard on a VMS  machine!   ------------------------------   Date: 7 Nov 2002 07:03:37 -0800o/ From: kevin_heikkila@yahoo.com (Kevin Heikkila)tY Subject: Northlake Printkit queues in "busy" state, with jobs which say "printing" but no3; Message-ID: <b64ef43.0211070703.14f2226@posting.google.com>>  F I have a situation where several Printkit version 2.2-3 queues go into= a "busy" state, with jobs that ultimately go into an infinite E "printing" status. Nothing ever gets printed. This doesn't happen alle the time, just once in a while.l  C The workaround is to go into KITCP and stop/start the queue and the 
 jobs startE to print. This normally works, but on several occasions this needs tow be doney8 several times so that all jobs in the queue get printed.  5 Has anyone see this behaviour before? Is there a fix?p   This is on VMS 6.1.    ------------------------------   Date: 6 Nov 2002 23:56:41 -0800u7 From: jones.computer.srv@worldnet.att.net (Daryl Jones)t Subject: Re: OpenVMS clusteringo= Message-ID: <8a646952.0211062356.6b025dc3@posting.google.com>a  s "Yong Liu" <fdu9774@rogers.com> wrote in message news:<Lrhy9.5575$oRV.3317@news04.bloor.is.net.cable.rogers.com>...o > Hi,, > J > I am interested in knowing how to apply OPENVMS clustering technology inI > alpha based servers. Specifically, I have an application running on twodN > alpha servers. How do I use clustering technology to ensure fault tolerance?N > Specifically, how does the clustering software monitor and manager fail over. > once my application fails on primary server? >  > Thanks  	 Youg Lui,o  E Generally, by having your applications run on all nodes and all disks E mounted on all nodes in the cluster. Furthermore, you must have batchgE queues setup to be exact on all your nodes where logicals can be usedeC direct the job stream to queues on each node. You will take about arC 10% hit on performance by doing this. One more thing make sure youra@ LOCKDIRWT is the same on all nodes. This will allow the locks toF tranfer to the node that has the most activities on the resource. ThisE provides dynamic lock mastering and directory searches based on whichi) nodes has the greater need for the locks.i  E When the a node fails, the users will only have to log onto the othereF nodes and restart their process. Certain applications like Oracle have& an automatic failover if a node fails.   Daryl Jonese   ------------------------------  % Date: Thu, 07 Nov 2002 09:50:37 +0000g( From: Nic Clews <sendspamhere@127.0.0.1> Subject: Re: OpenVMS clusteringa) Message-ID: <3DCA376D.E6C927CA@127.0.0.1>t   Yong Liu wrote:2 > J > I am interested in knowing how to apply OPENVMS clustering technology inI > alpha based servers. Specifically, I have an application running on two N > alpha servers. How do I use clustering technology to ensure fault tolerance?N > Specifically, how does the clustering software monitor and manager fail over. > once my application fails on primary server?  G All the replies you've had so far are right, but you maybe need to takenD a slight step back and read the rather good book from Digital calledG "Building Dependable Systems" although it is out of date with regard toE@ the technologies now available to OpenVMS, and is a little briefF regarding techniques for application writing, but this, and the recent/ clustering documentation is the place to start.d  G I think you need to have an overview of what you want to achieve. AboveyG you describe a situation where the 'redundant' server takes over from an@ failed primary, this is how "clustering" is implemented in otherE systems, and I'm loathe to describe them as clusters in comparison to  what OpenVMS can do for you.  H Here's one alternative. Shared storage with before image and after imageG logging for your application transactions (remember RTR is available tonC you), and a 'heartbeat'. The secondary server takes over on loss of D heartbeat (WNT style) and 'catches up' with the view from the shared1 storage, or whatever you've determined is common.   D However, this is very primitive in view of what VMS can give you, ifF you've paid for a server to sit there waiting to be failed on, who not? use it to share the service load, but this requires applicationsB co-ordination as described and provided by the OpenVMS distributedH locking system, in a very powerful and flexible way. I'd fully recommendH understanding how best to use the lock manager, and what it can do, thenF laterally think that into the way you design your application. LockingF isn't as simple as locked or not locked, there are many levels betweenD shared and exclusive, which form a huge matrix of possibilities when' talking about locks and their sublocks.v  E OpenVMS does data integrity exceptionally well, and understanding howoG clusters do what they do, particularly with storage and quorum, is alsoqC key to getting exactly what you want. Ultimately the design and theaA success of that design is down to you and your application folks..D OpenVMS is not clustering out of the box, and rightly so, because itF provides flexibility, that good knowledge can provide systems that areD disaster tolerant. However all the ingredients are there, trust your> application to OpenVMS and the cluster will protect your data.  E These are the showcase systems where you can read testimonials on the  OpenVMS website.  F Just because you have two systems clustered, does not make them in any@ way provide application fault tolerance by default, it has to be designed in.  F It'll be worth the effort. Failover is the lazy alternative, clustered? application co-operation is where this king of clusters shines.e   -- d? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciencesu nclews at csc dot come   ------------------------------  % Date: Thu, 07 Nov 2002 10:05:00 +0000e( From: Nic Clews <sendspamhere@127.0.0.1> Subject: Re: OpenVMS clusteringe) Message-ID: <3DCA3ACC.2000C333@127.0.0.1>e   Nic Clews wrote: >    Links:    P http://www.openvms.compaq.com/doc/73final/documentation/PDF/BUILD_DEPEND_SYS.PDF  2 http://www.openvms.compaq.com/success-stories.html     Programming:  H http://www.openvms.compaq.com/doc/731FINAL/5841/5841PRO.HTML (Locking is
 Chapter 7)   Clustering (including concepts)r  < http://www.openvms.compaq.com/doc/731FINAL/6318/6318PRO.HTML$ (Then re-read programming concepts!)< http://www.openvms.compaq.com/doc/731FINAL/4477/4477PRO.HTML  A I'm not saying you should disregard the rest of the documentation 	 though...A   -- U? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer SciencesH nclews at csc dot comn   ------------------------------  * Date: Thu, 7 Nov 2002 14:36:46 +0000 (UTC)+ From: david20@alpha1.mdx.ac.uk (David Webb)r Subject: Re: OpenVMS clustering + Message-ID: <aqdtpu$hk4$1@aquila.mdx.ac.uk>n  b In article <3DC9E354.4A6EA85E@vl.videotron.ca>, JF Mezei <jfmezei.spamnot@vl.videotron.ca> writes: >Yong Liu wrote:K >> I am interested in knowing how to apply OPENVMS clustering technology iniJ >> alpha based servers. Specifically, I have an application running on twoO >> alpha servers. How do I use clustering technology to ensure fault tolerance?eO >> Specifically, how does the clustering software monitor and manager fail overw/ >> once my application fails on primary server?s >w .. .  .hO >There are various ways for processes to talk to each other across nodes. There0X >is DECNET, TCPIP, as well asn the faster ICC$ routines. (intra cluster communications). >eK >If you build a TCPIP based server, VMS provides load balancing for inboundo: >requests, so a variety of nodes can process transactions. >cK >And of course, file access from multiple nodes is transparent. If node A'sdK >process locks a record, that that record is also locked if Node B tries to" >access it.   J This use of a distributed lock manager to control access to files, recordsI and other objects is why most VMS cluster applications are not setup in a N failover fashion but are set up to run concurrently on multiple cluster nodes.    
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------   Date: 7 Nov 2002 09:07:27 -0800 1 From: keithparris_NOSPAM@yahoo.com (Keith Parris)n Subject: Re: OpenVMS clustering = Message-ID: <cf15391e.0211070907.239a3dd8@posting.google.com>   s "Yong Liu" <fdu9774@rogers.com> wrote in message news:<Lrhy9.5575$oRV.3317@news04.bloor.is.net.cable.rogers.com>...wG > I am interested in knowing how to apply OPENVMS clustering technology   E Several folks have pointed to the documentation, which I think you'llm find will be very helpful.  @ For a gentle introduction to OpenVMS Clusters, you might find myD seminar notes on "OpenVMS Clusters: Basic and Intermediate Concepts" helpful:h http://www.geocities.com/keithparris/decus_presentations/f2000_vmscluster_seminar_basic_intermediate.ppt  A Probably the simplest example of automatic failover in an OpenVMS D Cluster is where you submit a job to a generic batch queue (and thatD generic batch queue points to all the possible nodes in the cluster,E but actually feeds jobs only to the nodes which are up and running at B any given time), with the /RESTART qualifier on the SUBMIT commandD pointing back to the same generic batch queue; this way, if the nodeE on which the job is running fails, the job is automatically restartedi) by the queue manager on a surviving node.    ------------------------------  $ Date: Thu, 7 Nov 2002 15:09:17 -00002 From: "Chris Sharman" <chris.sharman@sorry.nospam> Subject: OSU server cachee4 Message-ID: <aqdvju$gg8$1$8300dec7@news.demon.co.uk>  @ How do we prevent caching of pdf documents from our web server ?D There are some html http-equiv tags we're aware of (although they'reI documented as not working very well on ie), but we're not clear on how to 5 get them sent as true http headers for pdf documents.   F The documents are customer order proofs, which the customer reviews, &H requests amendments to, and we then amend & reproof. Most customers haveJ enormous difficulty getting around caching to see the revised copy, and weB then get hassled about why haven't we done their changes, etc etc.  J We'd like to specify a short life from the request time, or no caching, or something - any suggestions ?-   Using OSU server on VMS 7.3,   Thanks,  Chris5   ------------------------------  % Date: Thu, 07 Nov 2002 15:43:54 +0000l( From: Nic Clews <sendspamhere@127.0.0.1> Subject: Re: OSU server cacheo) Message-ID: <3DCA8A3A.12A53B84@127.0.0.1>o   Chris Sharman wrote:L > We'd like to specify a short life from the request time, or no caching, or > something - any suggestions ?h  E Are the web clients using SHIFT-RELOAD to view? This ought to force abC client cache refresh, but I've sometimes have to manually perform a H clear memory cache - clear disk cache in the browser. Nothing to do with the webserver.   -- p? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciences  nclews at csc dot com-   ------------------------------  $ Date: Thu, 7 Nov 2002 16:19:18 -00002 From: "Chris Sharman" <chris.sharman@sorry.nospam> Subject: Re: OSU server cache.4 Message-ID: <aqe3n7$kt1$1$8302bc10@news.demon.co.uk>  K > > We'd like to specify a short life from the request time, or no caching,1 or! > > something - any suggestions ?e >cG > Are the web clients using SHIFT-RELOAD to view? This ought to force atE > client cache refresh, but I've sometimes have to manually perform anJ > clear memory cache - clear disk cache in the browser. Nothing to do with > the webserver.  F Yes, that would do it, but only after our customers have got all hot & bothered & phoned us up.H We were hoping for a way to prevent our less web-literate customers from having problems.   Chrisg   ------------------------------  % Date: Thu, 07 Nov 2002 09:37:31 -060017 From: "Craig A. Berry" <craigberry@nospam.telocity.com>  Subject: Re: OSU server cache3G Message-ID: <craigberry-F064CA.09373107112002@news.directvinternet.com>e  4 In article <aqdvju$gg8$1$8300dec7@news.demon.co.uk>,4  "Chris Sharman" <chris.sharman@sorry.nospam> wrote:  L > We'd like to specify a short life from the request time, or no caching, or > something - any suggestions ?>  7 Read the chapter on caching in Alan Winston's new book.h   ------------------------------   Date: 7 Nov 2002 06:35:41 -0600y- From: koehler@encompasserve.org (Bob Koehler)u Subject: Re: Palladium article3 Message-ID: <bkEAch9+CnBi@eisner.encompasserve.org>U  b In article <3dc97180$1@news.si.com>, "Brian Tillman" <tillman_brian@notnoone.notnohow.com> writes:D >>   The TweakUI that I could find was for W95.  I'm running WNT and	 >>   W2K.E > M > Neither TweakUI nor X-Teq allows one to hide the "Hummingbird Neighborhood"VL > icon on the desktop.  I thought there was a way, but I can't find it right > now.  H    The Hummingbird Exceed 6.1 has an HNTUNE.exe which will move it from I    the desktop into "My Computer" where it's less obvious.  I know of no  3    documentation, but that's all that program does.2   ------------------------------   Date: 7 Nov 2002 00:10:23 -0800a7 From: jones.computer.srv@worldnet.att.net (Daryl Jones)eE Subject: Re: PGFLQUOTA increase on Alpha System VMS with java processg= Message-ID: <8a646952.0211070010.21430584@posting.google.com>   o mchaignon@lucent.com (marc chaignon) wrote in message news:<df8157c4.0211060117.1c3019e6@posting.google.com>...a > My configuration :  # > platform : DEC AXPVMS OpenVMS 7.3.' > Operating system : DEC AXPVMS VMS 7.3e& > process write in java with JDK 1.3.17 > This process communicates with Corba with JACORB V1.4 . > I execute my java process with this option : > -Xgc:compacting-# > for garbage collector activation.1@ > PGFLQUOTA max value is fixed at 600 000 for this java process. >  > My problem :F > The pgflquota of this process java increase indefinitely until 100%,@ > when the process is running. This quota increase mainly when a( > deconnection/connection corba arrived.; > In this java process, I trace the free memory with this :t. > java.lang.Runtime.getRuntime().freeMemory();0 > This free memory varries between 25 and 29 Mo., > Why this pgflquota increase indefinitely ? >  > Thank you for your help. >  > Bests regardsy >  > Marc Chaignono > Lucent Technologies France > mchaignon@lucent.com   Marc,a  A I was hoping someone else would give you more detail help by now.mD However, I will take a stab at it. My guess is that the Java programE is asking for more memory each time the deconn/conn cobra arrives andhC that memory can be page out to the page file, which causes the page1C file quota to be consumed. Unless the added memory is release aftermD deconn/conn has completed its task the page file quota will continueF to be consumed until the limit is reached. Again, this is only a quess on my part.    Daryl Jonest   ------------------------------  # Date: Thu, 07 Nov 2002 16:24:06 GMTc" From:   VAXman-  @SendSpamHere.ORG/ Subject: recommend a data/disk recovery servicef0 Message-ID: <00A169D8.A47E8A70@SendSpamHere.ORG>  J I have a RZ1CB-VW that appears to have a dead drive motor.  I have backupsJ from 22-AUG-2002 but I had some recent edits in progress since that backupJ that I did not get backed up.  I'd like to avoid retracing my footsteps to recreate it all.    H Has anybody here used a data recovery service that they would recommend?G I did some quick web searches but most of the hits take me to web sites H that are very PC/M$ centric.  I'd really like to trust my drive and dataH to a service that can at least spell VMS without having to spot them the "V" and/or the "M".s   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMf            R5   "Well my son, life is like a beanstalk, isn't it?" u   ------------------------------   Date: 7 Nov 2002 07:00:21 GMTa- From: djweath@attglobal.net (Dave Weatherall)c@ Subject: Re: SHOW PROCESS/PARENT (was: SHOW PROCESS/PARSE_STYLE)5 Message-ID: <DTiotGxQ0bj6-pn2-G1YWcM8HpndW@localhost>n  > On Tue, 5 Nov 2002 10:37:18 UTC, Kilgallen@SpamCop.net (Larry  Kilgallen) wrote:e  y > In article <01KOIB0061XM9QXTOZ@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes:lL > > SHOW PROCESS/PARSE_STYLE seems to work on 7.3 and 7.2.1-H1, but isn't inK > > HELP.  Is this a deliberate omission, i.e. an undocumented feature, or "' > > is it a bug that it is not in HELP?m > > M > > (Actually, I was hoping that SHOW PROCESS/PARENT existed.  :-)  Who else    > > has this on his wish list?)  > F > I see incredible semantic issues given the VMS habit of ignoring all1 > after the first four characters of a qualifier:i >  > 	SHOW PROCESS/PARENT > 	SHOW PROCESS/GRANDPARENTu! > 	SHOW PROCESS/GREAT_GRANDPARENTn3 > 	SHOW PROCESS/GREAT_GREAT_GREAT_GREAT_GRANDPARENTe   Are you sure it  does always?e  9 If you mean DCL I won't comment but consider the lines :-a   	RECORD CHANNEL WALLY  /ATT_IN 	RECORD CHANNEL WALLY  /ATT_OUTe  D I process these with the CLI$xxxx  routines. The ambivalence in the C first 4 chars (ATT_) means that the user _must_ enter the complete dD qualifier correctly. The CLI routines do differentiate though and I D can determine his intentions. /ATT or ATT_ brings forth a CLI error , message 'ambivalent qualifier' or such-like.  A That said, too many _GREAT's would definitely cause confusion :-)n  e -- x Cheers - Dave.   ------------------------------  * Date: Thu, 7 Nov 2002 08:41:38 +0100 (MET)& From: Rudolf Wingert <win@fom.fgan.de>) Subject: Re: Slow performance with BACKUPs6 Message-ID: <200211070741.IAA20578@sinet1.fom.fgan.de>   Hello,  C I am so sorry, that I answer so late. But I was out of office. Here@ my answers:.   Mike Rechtman wrotes:n  ; >>> a. Why not use a (much) larger /BLOCK than the default?C  B The default for disk savesets (32256) is the larges one I can use!  9 >>> b. Have you checked quotas (BYTLM, FILLM and friends)M  D Yes, I did. Years ago I did have a discussion with the TSC Muenchen.A They did say, that the values are to high. But I can't reduce thei$ values in case of SYSGEN parameters.  3 >>> c. Is there no/any/much other network activity?s  C Yes, the normal cluster activity. I did test the possibilities withiD DTSEND. The result: 95% bandwith by 100BaseTX and 54% by 1000BaseSX.C The Backup node is connected via GigaBit Ethernet. So six 100BaseTXnB can speak at the same time to the backup node. Also all connection are full duplex.   Bob Ceculski wrotes:   >>> /verify will slow it downr  I No, not realy. The backup needs eight day to copy the data to the saveset  and two days to verify.r   >>> /media=compact  8 If you use a tape, you are right, but I did use a disks.   John Santos wrotes:g  + >>> 2) Use a large blocksize.  /BLOCK=32256i   See above the answer to a)  > >>> 3) Use a large default extend quantity on the output disk.  G This would speed up the disk performance. But the max. block number for G network is much lower. Also BACKUP should be designed, that it open fordM IMAGE backup a saveset with a a size of useddiskblocks+savesetadministration.d I would do so.   David J. Dachtera wrotes:o  A >>> ...that such calculations assume exclusive access to the wire   F No. I did test all number with standard tools, during the normal time.H Ethernet with the DTSEND and a COPY from a MSCP served device to the NL:A device. I did see this performence over minutes without a change.t> Disk performance with a COPY of a contiguous file from disk toC disk. The write performance is the result of enableling the onboard $ write cache with a selfwritten tool.   Daryl Jones wrotes:m  B >>> Tuned each node that has these disk served to the backup node.  C This is a good idea, I have to check the MSCP buffers on all nodes.C  ! >>> Backup over a network is slowa  B That's not correct. If I do start only a few backups (four or so),C I will see a great performance. My feeling is, SCSI is the problem.oG If I do start a lot of writes and reads, the control and administrationME data will growth. This data will be served with 2.4MB/s only and slowoC down the disk access. During backup I did measure only 6MB/s spiraleD rate. But 6MB/s are 20GB/h, in case of this BACKUP must finish after  10hours and not after nine days.   TIA and reagrds Rudolf Wingert   ------------------------------  % Date: Thu, 07 Nov 2002 09:24:20 +0000e( From: Nic Clews <sendspamhere@127.0.0.1>) Subject: Re: Slow performance with BACKUPd) Message-ID: <3DCA3144.5E356437@127.0.0.1>i   Rudolf Wingert wrote:M >   ; > >>> b. Have you checked quotas (BYTLM, FILLM and friends)  > F > Yes, I did. Years ago I did have a discussion with the TSC Muenchen.C > They did say, that the values are to high. But I can't reduce thea& > values in case of SYSGEN parameters.  @ Some while back when we were doing some FibreChannel testing andF qualification, we had a little play around with BACKUP and the processD parameters, bearing in mind we had disks that could supply data at aE higher rate than ever before, a tape that could take data at a higherpD rate than ever before, and the DS20 which in itself could shift data quite rapidly.  D We discovered that exceeding the recommended process quotas actuallyC harmed (degraded) the backup speed. Backing them off to the maximumyE allowable, as according to the ancient 5.2 and backup tuning documentm= (and it's slightly more recent variant IIRC) enabled the bestoD throughput. (Disclaimer: This was done at 7.2-1 plus all the currentD patches at the time, and the test has not been repeated, mileage mayH therefore vary, the O/S has had a lot of work in the area of IO of late.E Also we did not experiment with varying each quota to determine where  the problem lay).C  @ I assume from your statement you have some rather enlarged PQL_MF settings, so perhaps it's upsetting the balance. Can you use any otherC techniques to address this so that the processes get their required B quotas? Are you in a cluster where a machine can be dedicated as a& backup engine (C/W reasonable quotas)?  w5 > >>> c. Is there no/any/much other network activity?r > E > Yes, the normal cluster activity. I did test the possibilities witheF > DTSEND. The result: 95% bandwith by 100BaseTX and 54% by 1000BaseSX.E > The Backup node is connected via GigaBit Ethernet. So six 100BaseTXiD > can speak at the same time to the backup node. Also all connection > are full duplex.  D I understand there is a later generation of gigabit cards which haveH improved performance. I have no idea if you're using one of those, or if it would affect your situation.0  ; Did you look at the distributed locking traffic statistics?9  3 Anyway, my selection of considerations for the pot.t   --  ? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer SciencesC nclews at csc dot como   ------------------------------   Date: 7 Nov 2002 09:23:02 -0800 7 From: jones.computer.srv@worldnet.att.net (Daryl Jones)f) Subject: Re: Slow performance with BACKUPe= Message-ID: <8a646952.0211070923.21d423b0@posting.google.com>C  d Rudolf Wingert <win@fom.fgan.de> wrote in message news:<200211070741.IAA20578@sinet1.fom.fgan.de>... > Hello, > E > I am so sorry, that I answer so late. But I was out of office. Heree
 > my answers:  >  > Mike Rechtman wrotes:  > = > >>> a. Why not use a (much) larger /BLOCK than the default?i > D > The default for disk savesets (32256) is the larges one I can use! > ; > >>> b. Have you checked quotas (BYTLM, FILLM and friends)e > F > Yes, I did. Years ago I did have a discussion with the TSC Muenchen.C > They did say, that the values are to high. But I can't reduce the-& > values in case of SYSGEN parameters. > 5 > >>> c. Is there no/any/much other network activity?K > E > Yes, the normal cluster activity. I did test the possibilities withrF > DTSEND. The result: 95% bandwith by 100BaseTX and 54% by 1000BaseSX.E > The Backup node is connected via GigaBit Ethernet. So six 100BaseTXtD > can speak at the same time to the backup node. Also all connection > are full duplex. >  > Bob Ceculski wrotes: >  > >>> /verify will slow it down  > K > No, not realy. The backup needs eight day to copy the data to the savesetg > and two days to verify.r >  > >>> /media=compact > : > If you use a tape, you are right, but I did use a disks. >  > John Santos wrotes:m > - > >>> 2) Use a large blocksize.  /BLOCK=32256j >  > See above the answer to a) > @ > >>> 3) Use a large default extend quantity on the output disk. > I > This would speed up the disk performance. But the max. block number foruI > network is much lower. Also BACKUP should be designed, that it open for O > IMAGE backup a saveset with a a size of useddiskblocks+savesetadministration.2 > I would do so. >  > David J. Dachtera wrotes:  > C > >>> ...that such calculations assume exclusive access to the wire  > H > No. I did test all number with standard tools, during the normal time.J > Ethernet with the DTSEND and a COPY from a MSCP served device to the NL:C > device. I did see this performence over minutes without a change.e@ > Disk performance with a COPY of a contiguous file from disk toE > disk. The write performance is the result of enableling the onboarde& > write cache with a selfwritten tool. >  > Daryl Jones wrotes:t > D > >>> Tuned each node that has these disk served to the backup node. > E > This is a good idea, I have to check the MSCP buffers on all nodes.t > # > >>> Backup over a network is slowo > D > That's not correct. If I do start only a few backups (four or so),E > I will see a great performance. My feeling is, SCSI is the problem. I > If I do start a lot of writes and reads, the control and administrationnG > data will growth. This data will be served with 2.4MB/s only and sloweE > down the disk access. During backup I did measure only 6MB/s spiralhF > rate. But 6MB/s are 20GB/h, in case of this BACKUP must finish after" > 10hours and not after nine days. >   > TIA and reagrds Rudolf Wingert   Rudolf,e   Bob Ceculski wrotes: >  > >>> /verify will slow it downr > K > No, not realy. The backup needs eight day to copy the data to the saveseto > and two days to verify.w  ? This statement about it only taking 2 days to verify the backup D definitely points the problem to data retrieval from the served diskF to the backup disk. From this statement, the backup should only take 2E days to complete. There is a bottleneck not a parameter problem beingi
 too small.   I hope this helps.   Daryl Jonesm   ------------------------------   Date: 7 Nov 2002 10:46:00 -0800 * From: jdimille@partners.org (John DimIlle)) Subject: sysgen>set shadow_sys_disk %4001e= Message-ID: <af3672e7.0211071046.4ed4e5a4@posting.google.com>u   Good Day   D  We are currently running OVMS 7.21h1 on a Cluste GS80. We have been< asked by HP Support to chaneg a Sysgen Dynamic parameter for SHADOw_sys_disk -1 to -14.  D  This will enable more information when shadowing gets an error backB fro a IO and then will fgo through Mount Verify. I have seen otherC options and curious to see if any other have set this to -14. IF sol6 can you explain why?. All patches are also up to date.   Regardsn   John Sr. Systems Managery   ------------------------------   Date: 7 Nov 2002 06:57:56 -0800c- From: merritt.robert@spsd.sk.ca (rob merritt)w0 Subject: trying to replace pathworks with samba.= Message-ID: <b6bf97d5.0211070657.40567be9@posting.google.com>p  D Hi I have a ovms axp machine with PW 6.1 installed my boss wanted to give samba a try aC (we need to upgrade our pwlmxxx7.02 lics to 7.03 to use the PW-AS).dE So I set about shutting down PW and making sure it would not restart.  then I install Samba and start it up and get:     $!
 $ set noon $ define/user sys$output nla0: $ define/user sys$error nla0:e $ tcpip disable service smbd $ tcpip set service smbd -         /protocol=TCP -e         /port=139 -a         /flags=listen -b         /user=SYSTEM -         /process=SMBD -e0         /file=SAMBA_ROOT:[BIN]SMBD_STARTUP.COM -:         /log=(FILE:SAMBA_ROOT:[VAR]SMBD_STARTUP.LOG,ALL) -         /limit=100& $ tcpip set config enable service smbd< %TCPIP-E-CONFIGERROR, error processing configuration request' -SYSTEM-F-BADPARAM, bad parameter value  $ tcpip enable service smbdw $!  ( is there some embedded PW that I missed    TCPIP> show config ena serv *e   Enable service=      LPD, PWIP_DRIVER, REXEC, RLOGIN, RSH, SMBD, SMTP, TELNETM   ------------------------------   Date: 7 Nov 2002 01:43:05 -0800.7 From: uwe.leinberger@t-systems.com (Dr. Uwe Leinberger)n$ Subject: VaxStation 3100 and M76/SPX= Message-ID: <b6b2072c.0211070143.1d29030c@posting.google.com>   	 Hi there!V  D Have a couple of these to change into "good hands". Location Munich,E Germany. Prices open for negotiation. Asking a good glas of beer.....n   ;-)    Uwe    ------------------------------   Date: 7 Nov 2002 14:00:51 GMTr( From: bill@cs.uofs.edu (Bill Gunshannon)( Subject: Re: VaxStation 3100 and M76/SPX5 Message-ID: <aqdrmi$95b05$2@ID-135708.news.dfncis.de>S  = In article <b6b2072c.0211070143.1d29030c@posting.google.com>,o: 	uwe.leinberger@t-systems.com (Dr. Uwe Leinberger) writes: > Hi there!h > F > Have a couple of these to change into "good hands". Location Munich,G > Germany. Prices open for negotiation. Asking a good glas of beer.....I >  > ;-)r >  > Uwem  ? If I was anywhere near Munich I would gladly buy you a glass ofS; beer for each of them and throw in a Jaegermeister as well.a  E But then, I'ld share them even if Vaxen weren't part of the deal. :-)   D I suppose it would take a keg to get them boxed up and put on a boat headed this way??a  B I can see it all now.  The first VAX Beowulf Cluster in the world.# Built entirely from VS3100's.   :-)    bill   -- sJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   a   ------------------------------  # Date: Thu, 07 Nov 2002 12:39:23 GMTl1 From: "Terry C. Shannon" <terryshannon@attbi.com>e1 Subject: Re: VMS on Itanium not mission critical?m, Message-ID: <%9ty9.71397$bG.50016@rwcrnsc53>  : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:3DC81670.5C9EA9CC@videotron.ca... > Bill Todd wrote:H > > That last comment seems to merit a response as well.  Since no Alpha systemK > > ever made has supported lock-step operation (EV7 may, but has no systemrK > > software that can make any use of it), are you suggesting that no AlphaL= > > system ever made is good enough for mission-critical use?n >  >'E > How did the VAXft handle the "fault tolerant" aspects ? Did it haven lockstep
 > in VAX ?  J Dual- rail, completely redundant, two copies of VMS, etc. That didn't stopH thr VAXft from going down when some bozo yankd the wrong power cord at a* DECUS some years back on the Left Coast... > I > Tandem's architecture requires lockstep, just like VMS requires certain  chipL > attributes (whcih were "missing" from alpha and incorporated as PAL code).< > Perhaps VMS achieves its robustness through other methods. >eH > It isn't so much the fact that IA64 has lockstep that is key here. But ratherH > the fact that when Tandem was lured into the Intel IA64 trap, it would haveH > given intel its requirements for a chip worthy of running in a Tandem. Hence,K > it is logical to conclude that IA64 was intented to have what it takes to  run4 > inside a serious machine.  >mI > What is interesting though is that at the time Intel was planning thosenH > features, Palmer had more or less told everyone not to bother with VMS because K > VMS wouldn't last much longer. So Intel wouldn't have factored VMS' needsi
 into IA64. >aK > So while the Tandem folks probably have a straightforwards port to a chipo thatJ > has the needed features to run NSK, are there features missing from IA64 which-D > are forcing the all-mighty VMS engineers to jump through hoops andJ > burning-loops to port something that was easy/straightforward on VAX and Alpha ?    ------------------------------   Date: 7 Nov 2002 06:03:26 -0800h% From: Franz.Lengel@EvoBus.com (Franz)l9 Subject: Re: VMS PHP connect to local Oracle Rdb Databaseu< Message-ID: <7eb9f7af.0211070603.63e4c19@posting.google.com>   Thanks Alan,  > Although unfortunately, your analysis is not very encouraging.  C Is there a portable scripting or programming language which runs onm- VMS and NT and allows similar access to Rdb ?W  E It might be Java I'm afraid. We haven't got experience with it, until-C now. And seems to be a lot of sweat to get so far, we are with Perle now.  B In Perl on NT we can access Rdb via ODBC. Even generating diagrams; based on the SQL query output is easy in Perl thanks to theD& ChartDirector from www.advsofteng.com.  E But, I would like to use same code on VMS and NT, as far as possible.u  
 Greetings, Franzs   ------------------------------  % Date: Thu, 07 Nov 2002 10:18:00 -0600t7 From: "Craig A. Berry" <craigberry@nospam.telocity.com>e9 Subject: Re: VMS PHP connect to local Oracle Rdb DatabaseTG Message-ID: <craigberry-B105F8.10180007112002@news.directvinternet.com>   < In article <7eb9f7af.0211070603.63e4c19@posting.google.com>,'  Franz.Lengel@EvoBus.com (Franz) wrote:   D > In Perl on NT we can access Rdb via ODBC. Even generating diagrams= > based on the SQL query output is easy in Perl thanks to the ( > ChartDirector from www.advsofteng.com. > G > But, I would like to use same code on VMS and NT, as far as possible.e  G The way the Perl DBI works you could very likely have code that is 99% <E the same on VMS and NT, but on VMS you'd be using DBD::Rdb under the l7 hood and on NT you'd be using DBD::ODBC under the hood.r   ------------------------------  % Date: Thu, 07 Nov 2002 11:07:49 -0600e7 From: "Craig A. Berry" <craigberry@nospam.telocity.com> 9 Subject: Re: VMS PHP connect to local Oracle Rdb DatabaseoG Message-ID: <craigberry-E70E1E.11074907112002@news.directvinternet.com>e  D Just another thought on this.  There appear to be a variety of open D source tools for gettting at Oracle databases, and if they're based E on SQL*Net it seems they could work with Rdb too.  Very much for the s7 do-it-yourselfer, but perhaps worth looking into.  See:e  + <http://fndapl.fnal.gov/~dbox/oracle/odbc/>u  * <http://www.oreillynet.com/oracle/os_dir/>   ------------------------------   Date: 7 Nov 2002 07:00:20 GMT - From: djweath@attglobal.net (Dave Weatherall)e@ Subject: Re: VMS to support KDE over CDE in which future version5 Message-ID: <DTiotGxQ0bj6-pn2-dXWSzz6VENjH@localhost>   3 On Tue, 5 Nov 2002 17:43:42 UTC, "Fred Kleinsorge" b$ <kleinsorge@star.zko.dec.com> wrote:   > Let me repeat. > J > We periodically consider the idea of Gnome or KDE.  We have not made anyF > decision to do either.  Even if we do, CDE/Motif will continue to be
 > offered. > J > When we consider the idea, the #1 topic is a business justification.  SoN > far, no strong business case has been made that would cause us to place thisI > above other work.  There are things on the horizon that "could" at some ) > point cause us to do it... but not yet.n <SNIP>  F Thanksvey much for the clarification Fred. Much appreciated. It means 5 that I can now discuss things from a firmer position.h   -- r Cheers - Dave.   ------------------------------  # Date: Thu, 07 Nov 2002 18:14:59 GMTs- From: "Carl Bennett" <carltbennett@yahoo.com>i Subject: Re: Voting Systems ?a7 Message-ID: <D4yy9.45577$7W2.2027@nwrddc01.gnilink.net>a  G Wasn't that one of the companies that merged into Unisys sometime aftera Eniac was born?o  ? "Atlant Schmidt" <atlantnospam@mindspring.com> wrote in messageo' news:3DC94F3B.A129E4F@mindspring.com...n > JF Mezei wrote:n >,E > > Also, I kinda giggled when they mentioned "paper tape". I figuredu someone mustI > > have insisted to keep at least one part of the legacy systems ... (ise papera0 > > tape older or younger than punched cards  ?) >o > It's an interesting question.' >u) > Remember that Hollerith developed punchi0 > cards in the 19th century for that very-famous, > 1890 census, and Jacquard looms (ca. 1801)  > predate *THAT* by quite a bit. >O2 > I'm no master of punched paper-tape history, but1 > marked paper tape (with, for example, dots-and-g) > dashes) existed no later than ca. 1850.m > / > And then there appears to be Basile Bouchon'se. > weaving loom that used punched paper tape to2 > control the pattern in 1725, but I've never seen > a picture of this device.v >y* > Extra credit: What became of Hollerith's > Tabulating Machine Company?w >  > Atlant >y >    ------------------------------  # Date: Thu, 07 Nov 2002 12:36:31 GMTt1 From: "Terry C. Shannon" <terryshannon@attbi.com>e- Subject: Re: Was OpenVMS left out on purpose?p> Message-ID: <j7ty9.115642$bt.179925@rwcrnsc52.ops.asp.att.net>  . "John Smith" <a@nonymous.com> wrote in messageD news:CG_x9.201780$%h2.172042@news02.bloor.is.net.cable.rogers.com... >e3 > "Shane Smith" <ssmith@icius.com> wrote in messaget, > news:01C284D9.53785580@sulfer.icius.com...J > > And /he/ couldn't get it changed? Or didn't want to? [Insert Jay's rap
 > > here.] >oL > I don't think that web site is owned/operated by HP, so I believe that all+ > he could do is bitch like the rest of us.d > H > Tandem/NSK isn't his department, and I'm sure Pauline Nist has her ownI > agenda to follow, and the only thing he/she have in common is that they  aresJ > both fated to sink with Itanic unless the banks and stock exchanges tell HP > to change cpu's for NSK.  C Well, they have something else in common. They both report to ScottrJ Stallard. I doubt that bank and exchanges care about the underlying CPU soJ long as the hardware is NonStop. IPF has lockstepping support and the portC is well along. And note that banks.exchanges will be among the lates	 adopters.p   > H > Unfortunately for VMS there is no group of customers large enough with thatH > much clout left.   DOD maybe, but even with current generation Alpha'sL > J-Stars will not likely need much more single-cpu power well into the nextK > decade - they can cluster to add horsepower until the APU's can't providedI > any more electricity. At that point, they lower the wind turbine out ofa the F > belly for more electricity and air conditioning. They can always put smaller0? > and lighter crews on-board to save the additional weight. :-)S  K Yeah, but the ram-air-turbine will increase the radar cross signature, last ! thing you want on a J-Stars bird.c >a >o   ------------------------------   Date: 7 Nov 2002 13:20:57 GMTo( From: bill@cs.uofs.edu (Bill Gunshannon)- Subject: Re: Was OpenVMS left out on purpose?/5 Message-ID: <aqdpbo$8t57e$1@ID-135708.news.dfncis.de>   > In article <m1ty9.115544$bt.179925@rwcrnsc52.ops.asp.att.net>,4 	"Terry C. Shannon" <terryshannon@attbi.com> writes: > J >                                                                      COE > ensures 15 years longevity r >   G Why do people keep saying this??  Government contracts run for one yearkD from Sept 1 to August 31.  There is no commitment on the governmentsH part fo renew any of them.  Support for anything done under a governmentG contract can end suddenly at the end of any fiscal year.  I know, I haduI a job just disappear that way (no, I didn't become un-employed.  I merely3G worked on other projects until the Army saw the error of their ways andw7 negotiated a new contract with MMDS to get me back. :-)o     bill   -- SJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   o   ------------------------------  # Date: Thu, 07 Nov 2002 12:30:10 GMTs1 From: "Terry C. Shannon" <terryshannon@attbi.com>s- Subject: Re: Was OpenVMS left out on purpose? > Message-ID: <m1ty9.115544$bt.179925@rwcrnsc52.ops.asp.att.net>  . "John Smith" <a@nonymous.com> wrote in messageC news:QlSx9.197587$%h2.53842@news02.bloor.is.net.cable.rogers.com...s > > > "Terry C. Shannon" <terryshannon@attbi.com> wrote in message: > news:n9Rx9.106352$wG.405791@rwcrnsc51.ops.asp.att.net...J > > VMS was mentioned a total of ONCE at the analyst briefing last week in > Santas > > Barbara. >e >  > As in, > a) OpenVMS is owned by HP, ors& > b) OpenVMS will be retired by HP, or8 > c) OpenVMS was acquired by HP in the Compaq merger, or > d) something more vague?  L Umm, a bit more vague. Marcello was supposed to be at the meeting, but had a	 conflict.   H Seems to me (and I could be wrong) that the IPF port proceeds apace, COEG ensures 15 years longevity (although support for non Air Force Generals F could end in 2011), and that NEW customers will be encouraged to adoptK HP-UX. That's about all I got. No info on early retirement, in fact OpenVMSn> Engineering has been largely exempt from workforce reductions. >  >  > Hope your move went well.t >p >    ------------------------------  # Date: Thu, 07 Nov 2002 12:30:54 GMTe1 From: "Terry C. Shannon" <terryshannon@attbi.com>e- Subject: Re: Was OpenVMS left out on purpose?C> Message-ID: <22ty9.115554$bt.180099@rwcrnsc52.ops.asp.att.net>  : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:3DC810D8.3CDA657D@videotron.ca... > "Terry C. Shannon" wrote:_ > >mJ > > VMS was mentioned a total of ONCE at the analyst briefing last week in Santal > > Barbara. >uA > Well, that is an infinite improvement.... going from 0 to 1 ...u  $ Many improvements since 7 May, IMHO.   ------------------------------  # Date: Thu, 07 Nov 2002 12:31:49 GMTn1 From: "Terry C. Shannon" <terryshannon@attbi.com>i- Subject: Re: Was OpenVMS left out on purpose?s, Message-ID: <V2ty9.71282$bG.51648@rwcrnsc53>  0 No, he was notified telephonically and by email.  : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:3DC848BC.56505C06@videotron.ca... > John Smith wrote:  > >:+ > > I know that Rich Marcello was notified.n >rF > Are you a spy ? Are there such weaknesses in HP's wintel based email system+ > that you can read his incoming emails ???p >d	 > :-) :-)o >u >e >e >r >  >r >r >o >h >  >p >u >t >  >r >w >  >( >h. > or did you just send him an email yourself ?   ------------------------------  % Date: Thu, 07 Nov 2002 10:20:09 -0500s0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>- Subject: Re: Was OpenVMS left out on purpose?e/ Message-ID: <3DCA84A3.B0306B25@vl.videotron.ca>.   "Terry C. Shannon" wrote:sE > Well, they have something else in common. They both report to ScottnL > Stallard. I doubt that bank and exchanges care about the underlying CPU soL > long as the hardware is NonStop. IPF has lockstepping support and the portE > is well along. And note that banks.exchanges will be among the late  > adopters.-  K If MIPS continues to release new-and-improved versions, then Tandem is in anN comfortable position since it can stay on MIPS as long as it takes before IA64J does the job, or IA64 is officially abandonned and a new chip is selected.  M The final timer has already begun for Alpha. There is nothing in the pipelinerK post EV79. Should IA64 be abandonned, it would take a very long time to getcE EV8 going again. So VMS would be stuck on a progressively slower chipoG (compared to rest of industry) until it can migrate to some other chip.r  L HP/Compaq made a huge gamble on IA64. It seems that the best payoff that canI be expected is a "me too" in the 64 bit market, but the worse scenario isvN killing HP's enterprise systems, except for Tandem which is currently based on a chip HP has no control over.   ------------------------------  # Date: Thu, 07 Nov 2002 18:11:46 GMTy# From: "John Smith" <a@nonymous.com>o- Subject: Re: Was OpenVMS left out on purpose?iF Message-ID: <C1yy9.8327$oRV.3633@news04.bloor.is.net.cable.rogers.com>  < "Terry C. Shannon" <terryshannon@attbi.com> wrote in message8 news:m1ty9.115544$bt.179925@rwcrnsc52.ops.asp.att.net... > 0 > "John Smith" <a@nonymous.com> wrote in messageE > news:QlSx9.197587$%h2.53842@news02.bloor.is.net.cable.rogers.com...h > > @ > > "Terry C. Shannon" <terryshannon@attbi.com> wrote in message< > > news:n9Rx9.106352$wG.405791@rwcrnsc51.ops.asp.att.net...L > > > VMS was mentioned a total of ONCE at the analyst briefing last week in	 > > SantaA > > > Barbara. > >n > >.
 > > As in,! > > a) OpenVMS is owned by HP, or ( > > b) OpenVMS will be retired by HP, or: > > c) OpenVMS was acquired by HP in the Compaq merger, or > > d) something more vague? >iL > Umm, a bit more vague. Marcello was supposed to be at the meeting, but had al > conflict.r >eJ > Seems to me (and I could be wrong) that the IPF port proceeds apace, COEI > ensures 15 years longevity (although support for non Air Force Generals-H > could end in 2011), and that NEW customers will be encouraged to adoptE > HP-UX. That's about all I got. No info on early retirement, in fact@ OpenVMSc@ > Engineering has been largely exempt from workforce reductions.     Terry,  L What gets me about the 2011 date is that the FTC has a ruling or legislationH (at least I think it falls under the FTC) that deals with issues of whatH happens when your car or computer manufacturer discontinues a particularI model. The vendor is required to stock parts for repair/replacement for aHL period of 5 years form the date of the product termination. Arguably one canF call this 'support'. No new engineering is done, no design defects areF remedied, they just ship you an identical part to the one that failed.  8 Now lets skip over to software and HP's words about VMS.  0 Support until 2011...what does this really mean?  H Does it mean the all new development on VMS ceases in 2006, and that theI 'support' window kicks in after that for the next 5 years. If so, all oneaI could expect from HP in 2007-2011 is a new CD of VMS and layered productsoK frozen in time as of 2006, and a friendly voice at the end of the telephone B line asking you for your credit card number and a ship-to address.  L If this is what HP means by support to 2011, everyone using VMS better startK looking for a different hardware and o/s vendor today so you can finish allc your porting by 2006.L    L As to the new customers being shown by HP to the PH-UX door, I am drawn to aG paraphrase of Henry Ford's comment about the Model T, "You can have anyh@ operating system from us that you want, as long as it's PH-UX.".  J Given all the cues HP has been giving, if I were you I'd consider renamingH your newsletter to "Shannon Knows AIX", and move on to greener pastures.   ------------------------------  # Date: Thu, 07 Nov 2002 18:15:30 GMTs# From: "John Smith" <a@nonymous.com> - Subject: Re: Was OpenVMS left out on purpose?aF Message-ID: <65yy9.8354$oRV.8111@news04.bloor.is.net.cable.rogers.com>  < "Terry C. Shannon" <terryshannon@attbi.com> wrote in message8 news:j7ty9.115642$bt.179925@rwcrnsc52.ops.asp.att.net... > 0 > "John Smith" <a@nonymous.com> wrote in messageF > news:CG_x9.201780$%h2.172042@news02.bloor.is.net.cable.rogers.com... > >i5 > > "Shane Smith" <ssmith@icius.com> wrote in messageU. > > news:01C284D9.53785580@sulfer.icius.com...L > > > And /he/ couldn't get it changed? Or didn't want to? [Insert Jay's rap > > > here.] > > J > > I don't think that web site is owned/operated by HP, so I believe that all-- > > he could do is bitch like the rest of us.: > > J > > Tandem/NSK isn't his department, and I'm sure Pauline Nist has her ownK > > agenda to follow, and the only thing he/she have in common is that theyR > are L > > both fated to sink with Itanic unless the banks and stock exchanges tell > HP > > to change cpu's for NSK. >nE > Well, they have something else in common. They both report to ScottiL > Stallard. I doubt that bank and exchanges care about the underlying CPU soL > long as the hardware is NonStop. IPF has lockstepping support and the portE > is well along. And note that banks.exchanges will be among the latel > adopters.u >m > >vJ > > Unfortunately for VMS there is no group of customers large enough with > thatJ > > much clout left.   DOD maybe, but even with current generation Alpha'sI > > J-Stars will not likely need much more single-cpu power well into thei nextE > > decade - they can cluster to add horsepower until the APU's can't  provide0K > > any more electricity. At that point, they lower the wind turbine out ofR > theAH > > belly for more electricity and air conditioning. They can always put	 > smallerlA > > and lighter crews on-board to save the additional weight. :-)a > H > Yeah, but the ram-air-turbine will increase the radar cross signature, last# > thing you want on a J-Stars bird.d    L It's a converted 707 for chrissake. That's about as stealthy as a cubic mile	 of chaff.    ------------------------------  # Date: Thu, 07 Nov 2002 18:16:53 GMTt# From: "John Smith" <a@nonymous.com>t- Subject: Re: Was OpenVMS left out on purpose?sF Message-ID: <p6yy9.14259$YSz1.67@news01.bloor.is.net.cable.rogers.com>  = "JF Mezei" <jfmezei.spamnot@vl.videotron.ca> wrote in messagen) news:3DCA84A3.B0306B25@vl.videotron.ca...  > "Terry C. Shannon" wrote:aG > > Well, they have something else in common. They both report to ScottOK > > Stallard. I doubt that bank and exchanges care about the underlying CPUR soI > > long as the hardware is NonStop. IPF has lockstepping support and theo portG > > is well along. And note that banks.exchanges will be among the late/
 > > adopters.r >sK > If MIPS continues to release new-and-improved versions, then Tandem is inA a)K > comfortable position since it can stay on MIPS as long as it takes beforeI IA64L > does the job, or IA64 is officially abandonned and a new chip is selected. >iF > The final timer has already begun for Alpha. There is nothing in the pipelineI > post EV79. Should IA64 be abandonned, it would take a very long time to- get.G > EV8 going again. So VMS would be stuck on a progressively slower chip0I > (compared to rest of industry) until it can migrate to some other chip.c >iJ > HP/Compaq made a huge gamble on IA64. It seems that the best payoff that can K > be expected is a "me too" in the 64 bit market, but the worse scenario issG > killing HP's enterprise systems, except for Tandem which is currently  based on  > a chip HP has no control over.    I As they say on Mr. Roger's Neighborhood, "Can you spell bankrupt? Sure, I( knew that you could."s   ------------------------------  # Date: Thu, 07 Nov 2002 09:48:52 GMT., From: "Ed Dennison" <ed.dennison@compaq.com>* Subject: Re: What is DAP status code 5067?0 Message-ID: <8Gqy9.5$pM4.62526@news.cpqcorp.net>  	 Lawrence,c  I 5067 is indicating there is a problem with the size of the DAP buffer. IttH would be worth looking at the format of the file being read. What is theD largest record in the file? You could then experiment with the $ SETL RMS/NETWORK_BLOCK_COUNT = n command. A $ SHOW RMS command will probably showF a system default of 8. Does it make any difference if you issue "$ SET. RMS/NET=127" and then try the Open/Read again?           Ed Dennison   7 "Lawrence Bleau" <bleau@umtof.umd.edu> wrote in messagee* news:aqbvrb$5hc$1@grapevine.wam.umd.edu...J > I'm running OpenVMS V7.1-2 AXP on a local system (UMSTEP) and VMS V5.5-2E > (VAX) on the remote system (UMDSP).  I have a user that is having ai problem L > reading a data file with his program.  When the file is on a local disk ofJ > UMSTEP, the program works fine.  When the file is on the remote node theK > program hangs on a pending i/o.  The program also hangs if the file is oni aaC > local disk and the user just adds UMDSP::UMSTEP:: in front of theu	 filename.i > K > ANAL/SYS shows the NETxxxx: device is busy, but the process isn't loggingrJ > any buffered or direct i/o's.  The remote NETSERVER.LOG file, as well asB > the local NET$SERVER.LOG, say there's a DAP status code of 5067. > K > Does anyone know what this means, and what might be causing this problem? 	 > Thanks.o >i > Lawrence Bleau > University of Maryland$ > Physics Dept., Space Physics Group > 301-405-6223 > bleau@umtof.umd.edur   ------------------------------   End of INFO-VAX 2002.616 ************************