1 INFO-VAX	Thu, 06 Jul 2000	Volume 2000 : Issue 375       Contents:) ANNOUNCING: MegaPOVRay v0.5a For OpenVMS.   Re: Charon-VAX Hobbyist Question$ DCL for detecting a finished process( Re: DCL for detecting a finished process Re: DCL vs. "real" programming Re: DCL vs. "real" programming Re: DCL vs. "real" programming Re: DEC C Question *variable ++  Re: DEC C Question *variable ++  Re: DEC C Question *variable ++  Re: DEC C Question *variable ++  Re: DEC C Question *variable ++  Re: DEC C Question *variable ++  Re: DEC C Question *variable ++  Re: DEC C Question *variable ++ ; DEC C Question *variable ++ (possible compiler enhancement) ? Re: DEC C Question *variable ++ (possible compiler enhancement)  Decserver unreachable  Re: Decserver unreachable 7 Re: Disk I/O Performance (was Re: OpenVMS loses big...) 7 Re: Disk I/O Performance (was Re: OpenVMS loses big...) 7 Re: Disk I/O Performance (was Re: OpenVMS loses big...) ) Re: highwater marking, speed vs. security ) Re: highwater marking, speed vs. security ) Re: highwater marking, speed vs. security 8 Re: How can I "replicate" Volume Shadowing on a Sun box? Re: LA36 Escape codes ' Re: Multilingual System Dictionary(ies) ' Re: Multilingual System Dictionary(ies)   Multinet port numbers being used$ RE: Multinet port numbers being used$ RE: Multinet port numbers being used( Re: Netscape Fasttrack for OpenVMS 7.2-12 Re: OpenVMS loses big, was:  RE: Compaq advertises2 Re: OpenVMS loses big, was:  RE: Compaq advertises2 Re: OpenVMS loses big, was:  RE: Compaq advertises1 Re: OpenVMS loses big, was: RE: Compaq advertises , RE: Oracle Press Release, incl. OpenVMS info Problems with DCOM for OpenVMS" Re: Problems with DCOM for OpenVMS Test!! Ingnore!! Re: Trapping a STOP/ID Re: Trapping a STOP/ID Re: Trapping a STOP/ID Re: Trapping a STOP/ID Re: Trapping a STOP/ID Re: Trapping a STOP/ID Re: Trapping a STOP/ID1 Re: Upgrade to VMS 7.2 fails on Help Message File  Vax in Canada ?  RE: Vax in Canada ?  Re: Vax in Canada ?  Re: VMS employment Re: We need VMS people RE: We need VMS people  F ----------------------------------------------------------------------  % Date: Thu, 06 Jul 2000 01:21:49 -0500 1 From: Robert Alan Byer <byer@mail.ourservers.net> 2 Subject: ANNOUNCING: MegaPOVRay v0.5a For OpenVMS.3 Message-ID: <3963DF2D.7A6EA4B4@mail.ourservers.net>   ( Announcing MegaPOVRay v0.5a for OpenVMS., (Yes, it's a month late, but I've been busy)    Here's what's new/fixed/removed.     REMOVED GIF SUPPORT 6 	(potential legal issued reguarding the Unisys patnet)    fixed docs for pigment_pattern.   fixed disc problems (removed fastdisc patch)+   fixed bug with photons on average texture .   fixed bug with media method 3 when samples=1?   fixed bug with parsing backwards compatibility and unofficial  	version number    added fastsphere patch   added fastpoly patchC   fixed "bug" with mesh (MegaPOV didn't ignore degenerate triangles ( 	if uv mapping was used in the triangle)   fixed bug with tff   fixed bug with csqr   B Again, currently only the Alpha platform is supported with MegaPOV> because it requires the IEE_FLOAT mode which VAX dosen't have.  $ You can get the latest version at...  ) 	http://www.ourservers.net/openvms_ports/   ; If you have any problems or questions, feel free to e-mail.    --    @  +------------------+--------------------------+---------------+@  | Robert Alan Byer | byer@mail.ourservers.net | ICQ #65926579 |@  +------------------+--------------------------+---------------+@  | Send an E-mail request to obtain a copy of my PGP key.      |@  +-------------------------------------------------------------+@  | "It is by caffeine alone I set my mind in motion.  It is by |@  |  cans of cola the thoughts aquire speed, the hands aquire   |@  |  shakes, the shakes become a warning.  It is by caffeine    |@  |  alone I set my mind in motion."                            |@  +-------------------------------------------------------------+   ------------------------------  # Date: Thu, 06 Jul 2000 06:04:13 GMT - From: goathunter@goatley.com (Hunter Goatley) ) Subject: Re: Charon-VAX Hobbyist Question , Message-ID: <3964211f.63035119@news.wku.edu>  1 On Wed, 28 Jun 2000 02:35:43 GMT, "Zane H. Healy" # <healyzh@shell1.aracnet.com> wrote:  > J >That's the full blown commercial version.  They disabled ethernet supportK >and some other stuff on the Hobbyist version.  Which unfortunatly makes it   >almost worthless for hobbyists. > C Just FYI, following the suggestion someone here made, I was able to < get C-Kermit transferring files after I bumped up the SYSGENB parameters related to the typeahead buffers.  Once again, C-Kermit comes to the rescue!   Hunter ------9 Hunter Goatley, Process Software, http://www.process.com/ 9 goathunter@goatley.com     http://www.goatley.com/hunter/    ------------------------------  % Date: Thu, 06 Jul 2000 08:26:50 -0700 8 From: quicksand <quicksandNOquSPAM@chilis.co.uk.invalid>- Subject: DCL for detecting a finished process 9 Message-ID: <3a3e1d10.fbee851c@usw-ex0102-013.remarq.com>   ?   We are working on a script (.COM) to loop "SHOW SYS" commands = for a specific process...  it logs vital statistics about the A process to a file.  Right now it just loops until we kill it with 
 Cntrl-C. <br> @   We'd like to enhance it so that it detects when the process is@ gone (the SHOW SYS command would return nothing/blanks) and stop automatically.	 <br> <br> ?   I tried using a test of $STATUS, but evidentally it considers 9 all requests sucessful, even if the process is not found. > <br>   other trivia:  It defines sys$output to the desired logA file before attempting any "show sys/id='p1'/noheading" commands.  <br> <br> <br> Thanks very much in advance.      ; -----------------------------------------------------------   7 Got questions?  Get answers over the phone at Keen.com.  Up to 100 minutes free!  http://www.keen.com    ------------------------------  % Date: Thu, 06 Jul 2000 17:26:38 +0100 - From: Tim Llewellyn <tim.llewellyn@bbc.co.uk> 1 Subject: Re: DCL for detecting a finished process ) Message-ID: <3964B33E.8C36FDA8@bbc.co.uk>    quicksand wrote:  A >   We are working on a script (.COM) to loop "SHOW SYS" commands ? > for a specific process...  it logs vital statistics about the C > process to a file.  Right now it just loops until we kill it with  > Cntrl-C. <br> B >   We'd like to enhance it so that it detects when the process isB > gone (the SHOW SYS command would return nothing/blanks) and stop > automatically. > <br> <br> A >   I tried using a test of $STATUS, but evidentally it considers ; > all requests sucessful, even if the process is not found. @ > <br>   other trivia:  It defines sys$output to the desired logC > file before attempting any "show sys/id='p1'/noheading" commands.  > <br> <br> <br> > Thanks very much in advance.  F You have the PID of the process, why not use F$GETJPI instead? You can get everything. SHOW SYS tells you from F$GETJPI. For example:  ' $ write sys$output f$getjpi (212,"PID") & %SYSTEM-W-NONEXPR, nonexistent process  \212\ $ sh symb $status    $STATUS == "%X000008E8"  $ exit %x8e8& %SYSTEM-W-NONEXPR, nonexistent process  D So you can test for $STATUS equals %X8E8 to determine if the process is still running.   F Its either that or check the line count in your file that is receiving
 the outputA from SHOW SYS/NOHEAD/ID=, it should be zero if there process does 
 not exist.  -- 6 Tim Llewellyn, OpenVMS Infrastructure, Remarcs Project0 MedAS at the BBC, Whiteladies Road, Bristol, UK.A Email tim.llewellyn@bbc.co.uk. Home tim.llewellyn@cableinet.co.uk   A I speak for myself only and my views in no way represent those of  MedAS or the BBC.    ------------------------------  # Date: Thu, 06 Jul 2000 12:00:52 GMT . From: Michael Austin <maustin@nc.prestige.net>' Subject: Re: DCL vs. "real" programming / Message-ID: <39648297.EB95F280@nc.prestige.net>   , This is a multi-part message in MIME format.& --------------9A5DBED70C72B38417FA649E* Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit   L I seem to recall an internal DEC program that I saw in "the Mill" that wouldM actually compile the DCL.  At the time I saw it, it would take a complex MVII R almost 20 minutes to boot.  With the compiled DCL, only 3 minutes.  I often wonderI what happened to those neat tools that you only saw as an employee of the Q company... and there were lot's of them.  Like FTSV which used VMS copy and if it N got interrupted, you could restart it and it would pick up at the block numberK where it left off.  Really great when the only interoffice connection was a  not-so-reliable 56K connection.    Michael AustinO Ex-Digit 156990 1984-1994  -- before being traded to Oracle.. and I didn't even  know I was a free agent!  9 "Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515" wrote:   0 > In article <8job8r$j8u$1@info.service.rug.nl>,6 >         helbig@astro.rug.nl (Phillip Helbig) writes: > [...] J > > It's not a one-shot (for which I would certainly use DCL), so it looksJ > > like it's F$ time.  This reminds me, someone once mentioned here a DCLJ > > compiler which would take a .COM file and convert it to the equivalentH > > FORTRAN with system calls etc and compile that.  Who can point me to	 > > this?  > J >         There is (or  was)  a  company  called Channel Islands Software,J >     really  just one guy, who had a product that did this.  I can't  forJ >     the life of me  remember  its  name.   Something  like  DCLFort?   IJ >     dunno...   It  did,  in  fact,  read  DCL  command  files and outputJ >     Fortran.  It was a bit pricey, like  US$900 or so for a single smallJ >     system.   Something  for the "big boys" who had huge DCL  procedures  >     that needed to be sped up. > J >         I talked to the  owner/programmer  on  the  phone a time or two.J >     One  of  the  issues  was that the source code  it  output  was  notJ >     particularly "maintainable".  Obfuscated I  take  it.   Another  wasJ >     that  you  needed  the  product's  RTL,  which provided the callableJ >     replacements for all  the  lexical  functions  that aren't otherwiseJ >     available  (of  course).  In other words, there was a lot more  thanJ >     just calls to VMS RTL and System Services routines.  At the  time  IJ >     couldn't justify the cost (I mostly wanted to just try it out, but aJ >     "demo"  was  really  available).   I  think there may have been someJ >     issues of keeping the product up  to  date with VMS releases as well, >     (i.e., adding the new lexicals, etc.). > J >         If there's interest, I'll see if  I  still have a flier or otherC >     reference information, but I expect it's a long-dead product.  >  >         -Ken > --O >  Kenneth H. Fairfield            |  Internet: Fairfield@SLC.Slac.Stanford.Edu < >  SLAC, 2575 Sand Hill Rd, MS 46  |  Voice:    650-926-2924< >  Menlo Park, CA  94025           |  FAX:      650-926-3515P >  -----------------------------------------------------------------------------D >  These opinions are mine, not SLAC's, Stanford's, nor the DOE's...  & --------------9A5DBED70C72B38417FA649E- Content-Type: text/x-vcard; charset=us-ascii;   name="maustin.vcf"  Content-Transfer-Encoding: 7bit , Content-Description: Card for Michael Austin  Content-Disposition: attachment;  filename="maustin.vcf"    begin:vcard  n:Austin;Michael   tel;work:704-947-1089  x-mozilla-html:FALSE org:Michael Austin, Inc 
 adr:;;;;;; version:2.1 + email;internet:michaelaustininc@hotmail.com  title:President  x-mozilla-cpt:;0 fn:Michael E. Austin	 end:vcard   ( --------------9A5DBED70C72B38417FA649E--   ------------------------------  % Date: Thu, 06 Jul 2000 15:31:22 +0100 - From: Tim Llewellyn <tim.llewellyn@bbc.co.uk> ' Subject: Re: DCL vs. "real" programming ( Message-ID: <3964983A.AAD00EF@bbc.co.uk>   Michael Austin wrote:   S > company... and there were lot's of them.  Like FTSV which used VMS copy and if it P > got interrupted, you could restart it and it would pick up at the block numberM > where it left off.  Really great when the only interoffice connection was a ! > not-so-reliable 56K connection.   ; Yeah, UK ACademia's Coloured Books used to do the same back P in the 80's, very useful feature, no resend on error, any part of the connection	 can fail.   F Then we moved to DECNET, and finally IP. Each time features were lost.  -- 6 Tim Llewellyn, OpenVMS Infrastructure, Remarcs Project0 MedAS at the BBC, Whiteladies Road, Bristol, UK.A Email tim.llewellyn@bbc.co.uk. Home tim.llewellyn@cableinet.co.uk   A I speak for myself only and my views in no way represent those of  MedAS or the BBC.    ------------------------------    Date: 06 Jul 2000 16:56:24 +0200G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de> ' Subject: Re: DCL vs. "real" programming H Message-ID: <y41z17pasn.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>   [resuming interupted transfers] L That's not a TCP/IP problem, it's an application problem. See WGET for help.   	Jan   ------------------------------   Date: 6 Jul 2000 02:13 CST' From: carl@gerg.tamu.edu (Carl Perkins) ( Subject: Re: DEC C Question *variable ++, Message-ID: <6JUL200002133258@gerg.tamu.edu>  1 JF Mezei <jfmezei.spamnot@videotron.ca> writes... 
 }*length ++ ;  }  }or  }  }*length = *length + 1 ; } B }Is there a reason why only the second variant works as expected ? } , }What does  " *length ++ "  supposed to do ? }  }DEC-C 6.0, VAX (VMS 7.2).  ? Compare the meanings of these two interpretations of your code:    *(length ++); 
 (*length) ++;   L I expect that the former is what you got, but the latter is what you wanted.  K I expect that some compilers would have interpreted it the way you expected  it to be interpreted.   D (I prefer to avoid this sort of syntax anyway - it has always seemedB like a bad plan to me. Probably all the Fortran in my background.)   --- Carl   ------------------------------  $ Date: Thu, 6 Jul 2000 09:54:20 +0100* From: "Richard Brodie" <R.Brodie@rl.ac.uk>( Subject: Re: DEC C Question *variable ++, Message-ID: <8k1hfu$19pm@newton.cc.rl.ac.uk>  [ "Carl Perkins" <carl@gerg.tamu.edu> wrote in message news:6JUL200002133258@gerg.tamu.edu...    > *(length ++);  > (*length) ++;  > N > I expect that the former is what you got, but the latter is what you wanted. > M > I expect that some compilers would have interpreted it the way you expected  > it to be interpreted.   L I hope not. Not after all the trouble the designers of C went to in defining# 15 levels of operator precedence ;)    ------------------------------  # Date: Thu, 06 Jul 2000 09:24:02 GMT 2 From: Shraga_Broyer@bmc.nospam.com (Shraga Broyer)( Subject: Re: DEC C Question *variable ++- Message-ID: <39644f2b.825992163@news.bmc.com>   , On Wed, 05 Jul 2000 22:05:18 -0400, JF Mezei  E According to the Compaq C Language Reference Manual (page 6-5) , Post F operators (like ++) have precedence over unary operator (like *). That causes the problem.   &  <jfmezei.spamnot@videotron.ca> wrote:   >myfunction (int *length)  >{
 >*length ++ ; 	 >return ;  >} >  >or  >  >myfunction (int *length)  >{   >*length = *length + 1 ;	 >return ;  >} > B >Is there a reason why only the second variant works as expected ? > , >What does  " *length ++ "  supposed to do ? >  >  >DEC-C 6.0, VAX (VMS 7.2).   ------------------------------   Date: 6 Jul 2000 14:00:50 GMT 1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon) ( Subject: Re: DEC C Question *variable ++, Message-ID: <8k23ei$1fgd$1@info.cs.uofs.edu>  , In article <3963E95E.E16E1E80@videotron.ca>,0  JF Mezei <jfmezei.spamnot@videotron.ca> writes: |> myfunction (int *length)  |> { |> *length ++ ;  |> return ;  |> } |>   |> or  |>   |> myfunction (int *length)  |> {   |> *length = *length + 1 ; |> return ;  |> } |>  D |> Is there a reason why only the second variant works as expected ? |>  . |> What does  " *length ++ "  supposed to do ?  6 Probably because of the " " between "length" and "++".   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>   0   ------------------------------  % Date: Thu, 06 Jul 2000 10:18:32 -0400r+ From: Tim Shoppa <shoppa@trailing-edge.com> ( Subject: Re: DEC C Question *variable ++1 Message-ID: <39645CF8.2A09F1B7@trailing-edge.com>r   JF Mezei wrote:e >  > myfunction (int *length) > {  > *length ++ ;
 > return ; > }  >  > or >  > myfunction (int *length) > {i > *length = *length + 1 ;n
 > return ; > }e > C > Is there a reason why only the second variant works as expected ?  > - > What does  " *length ++ "  supposed to do ?b  C Classic first-week-C-programmer problem.  From the comp.lang.c FAQ:p     Question 4.3  /   Does *p++ increment p, or what it points to? i    E   Unary operators like *, ++, and -- all associate (group) from rightl to   left. Therefore, *p++B   increments p (and returns the value pointed to by p before the  $   increment). To increment the valueH   pointed to by p, use (*p)++ (or perhaps ++*p, if the order of the side   effect doesn't matter).     References: K&R1 Sec. 5.1 p. 91  K&R2 Sec. 5.1 p. 95  ANSI Sec. 3.3.2, Sec. 3.3.3  ISO Sec. 6.3.2, Sec. 6.3.3 oI H&S Sec. 7.4.4 pp. 192-3, Sec. 7.5 p. 193, Secs. 7.5.7,7.5.8 pp. 199-200 s   --  B  Tim Shoppa                        Email: shoppa@trailing-edge.comG  Trailing Edge Technology          WWW:   http://www.trailing-edge.com/ 6  7328 Bradley Blvd                 Voice: 301-767-59176  Bethesda, MD, USA 20817           Fax:   301-767-5927   ------------------------------   Date: 6 Jul 2000 14:30:10 GMT 1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon) ( Subject: Re: DEC C Question *variable ++, Message-ID: <8k255i$1g2v$1@info.cs.uofs.edu>  , In article <8k23ei$1fgd$1@info.cs.uofs.edu>,  you write:o/ |> In article <3963E95E.E16E1E80@videotron.ca>,/3 |>  JF Mezei <jfmezei.spamnot@videotron.ca> writes:e |> |> myfunction (int *length) |> |> {/ |> |> *length ++ ; |> D9 |> Probably because of the " " between "length" and "++".n |> d  < I stand corrected.  Apparently the space is not significant.H I guess I just confused it with the autoincrement indicator in Macro-11. Sorry.   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>   g   ------------------------------  % Date: Thu, 06 Jul 2000 13:03:31 -0400)- From: JF Mezei <jfmezei.spamnot@videotron.ca>E( Subject: Re: DEC C Question *variable ++, Message-ID: <3964BBE1.FA2BA29D@videotron.ca>   Carl Perkins wrote:eA > Compare the meanings of these two interpretations of your code:o >  > *(length ++);  > (*length) ++;d > N > I expect that the former is what you got, but the latter is what you wanted.  L Well, obviously, it seems to be a problem of precedence. But I was wonderingJ if  the expected ANSI behaviour would have been to compile *length ++ into (*length) ++ . n  & Also, consider the difference between:  ( *(length ++) ;	and	myvar = *(length ++);  N If the standard is really to compile into *(length ++), shouldn't the compilerL have issued an error because *length ++ ; as a standalone statement does notM make sense ? Sure, it increments the pointer, but it then places the value ate< the new pointer on the stack but makes no use of that value.  > Wouldn't the compiler complain if I had a statements such as :  	 *length ;t 25 ; 2 + 2 ;9   ? ? ?-   ------------------------------  % Date: Thu, 06 Jul 2000 13:39:59 -0400i+ From: Tim Shoppa <shoppa@trailing-edge.com> ( Subject: Re: DEC C Question *variable ++1 Message-ID: <39648C2F.5F1BA790@trailing-edge.com>l   JF Mezei wrote:n@ > Wouldn't the compiler complain if I had a statements such as : >  > *length ;  > 25 ;	 > 2 + 2 ;   = In fact, CC/WARN=ENABLE=ALL will complain about all of these:    *length; ^oG %CC-I-EXPRNOTUSED, In this statement, the expression "*length" is never- used.-< at line number 3 in file DISK$USER:[SHOPPA.RESUME]MYFUNC.C;5   25;u ^0H %CC-I-EXPRNOTUSED, In this statement, the expression "25" is never used.< at line number 4 in file DISK$USER:[SHOPPA.RESUME]MYFUNC.C;5   2+2; ^kC %CC-I-EXPRNOTUSED, In this statement, the expression "2+2" is neverh used..< at line number 5 in file DISK$USER:[SHOPPA.RESUME]MYFUNC.C;5  1 But the default compiler options won't complain. e   > ? ? ?t  * That's not legal, but if instead you wrote   5 ? 3 ? 4 ? 3 : 4 : 5 : 7 ;b   it'd be perfectly legal!   Tim.   ------------------------------  $ Date: Thu, 6 Jul 2000 10:51:55 +0100* From: "Richard Brodie" <R.Brodie@rl.ac.uk>D Subject: DEC C Question *variable ++ (possible compiler enhancement)+ Message-ID: <8k1krs$pv8@newton.cc.rl.ac.uk>r  a "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message news:3963E95E.E16E1E80@videotron.ca...   - > What does  " *length ++ "  supposed to do ?   < Increment length and then dereference it. ++ has the highest precedence.a  @ I note that compiling the code gives me no errors, even with allA warnings enabled (CC 6.0). Naively parsing the code I read it as:    1) temp = length++ 2) *temp [redundant]  @ Would it be feasible to issue a warning when expressions are not3 used, even when portions of them have side effects?l  . e.g. "x * 3;" errors but "f(x) * 3;" does not.   ------------------------------  % Date: Thu, 06 Jul 2000 09:16:56 -0400d( From: Ed Vogel <edward.vogel@compaq.com>H Subject: Re: DEC C Question *variable ++ (possible compiler enhancement)* Message-ID: <396486C8.13A71641@compaq.com>   Richard Brodie wrote:    >>B > Would it be feasible to issue a warning when expressions are not5 > used, even when portions of them have side effects?s > 0 > e.g. "x * 3;" errors but "f(x) * 3;" does not.  9    I expect this would not be easy, but I have added this ?    request to our wish list.  I do note that lint on Tru64 willt/    detect these cases so it's certainly doable.-  1                                          Ed Vogel1=                                          Compaq C Engineering    ------------------------------  % Date: Thu, 06 Jul 2000 10:27:48 +0100C- From: Eric CHEVALLIER <chevalli@mpfrance.com>e Subject: Decserver unreachable, Message-ID: <39645114.69437713@mpfrance.com>   Hello,  @ I have a poblem all the decserver (200) dont work soudainly. The< material and the configuration is the same since five years.> When i make "SH K NODE" openVMS tell me that the decserver are unreachable.- with "SH CIRCUIT" i don't see EWA-0 status ONnA I have changed the Delni but the Decserver are still unreachable. * Do you know something about this problem ? Thank you in advance.v Eric   ------------------------------  % Date: Thu, 06 Jul 2000 11:54:57 +0200o5 From: Oswald Knoppers <Oswald.Knoppers@whitehouse.nl>s" Subject: Re: Decserver unreachable- Message-ID: <39645771.6155D19E@whitehouse.nl>    Eric CHEVALLIER wrote: >  > Hello, > B > I have a poblem all the decserver (200) dont work soudainly. The> > material and the configuration is the same since five years.@ > When i make "SH K NODE" openVMS tell me that the decserver are > unreachable.  < The decserver doesn't speak DECnet so it will always show as unreachable.  - You can try to connect to the server via mop:   E NCP> connect node xxx (works only if the hardware address is defined)o   or  5 NCP> connect via ewa-0 phys address xx-xx-xx-xx-xx-xxO  / The address should be printed on the decserver.    Regards,   Oswald   ------------------------------  / Date: Thu, 06 Jul 2000 09:10:15 +0200 (MET DST)d& From: Rudolf Wingert <win@fom.fgan.de>@ Subject: Re: Disk I/O Performance (was Re: OpenVMS loses big...)/ Message-ID: <200007060616.IAA06978@fom.fgan.de>e   Hello,   Bill Todd wrotes:-   >>>-L The normal approach to obtaining good streaming performance from a SCSI diskK is to use asynchronous multi-buffering such that the disk always has one or3J more outstanding requests ready to be satisfied, and can just keep sendingD or receiving the data without pause.  This does not require enablingI write-back caching for writes to perform well.  You can ask RMS to handleaG this for you (using its read-ahead/write-behind options) in some cases.i  H If you weren't doing this, the reason you obtained decent streaming-readH performance was because the disk was automatically pre-fetching for you.K Depending on the intelligence of its algorithm, you might or might not haveoB been able to do somewhat better by doing your own multi-buffering. <<<   K Tis is not our expirience. Multi-buffering brings a little bit more perfor- H mance, but that growth that the disk onboard write cache brings. We haveI tested it with WINNT. Disbaling the disk cache (read or write) brings the6J performance down from 15.2MB/s to 3.6MB/s (in our test environment; 18.2GBH IBM DGHS disk). This value is relative system independent. So you can doD with RMS what you want, you can't speed up over the meassured value.   Regards Rudolf Wingert  E P.S. The OpenVMS test was a little bit unrealistic, because we used acD      contigeous file. But I think that on all systems non-contigeous.      files will bring a performance degration.   ------------------------------    Date: 06 Jul 2000 10:13:47 +0200G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de> @ Subject: Re: Disk I/O Performance (was Re: OpenVMS loses big...)H Message-ID: <y4n1jv1xs4.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  ( Rudolf Wingert <win@fom.fgan.de> writes:  ) [RMS multibuffering and asynchronous I/O]FM > Tis is not our expirience. Multi-buffering brings a little bit more perfor-uJ > mance, but that growth that the disk onboard write cache brings. We haveK > tested it with WINNT. Disbaling the disk cache (read or write) brings therL > performance down from 15.2MB/s to 3.6MB/s (in our test environment; 18.2GB< > IBM DGHS disk). This value is relative system independent.G > P.S. The OpenVMS test was a little bit unrealistic, because we used a" >      contigeous file.d  N But that's what the original post was talking about: writing data sequentiallyJ to files that are a large fraction of a disk's capacity, and (much) largerJ than any on-disk cache. In that case, the cache might as well not exist asL long as you can keep the device fed with operations to perform. Ergo, if youK performance drops when you disable cache in this scenario, you're not doingt# the I/O properly in the first case.    	Jan   ------------------------------  $ Date: Thu, 6 Jul 2000 11:51:53 -0400' From: "Bill Todd" <billtodd@foo.mv.com>s@ Subject: Re: Disk I/O Performance (was Re: OpenVMS loses big...)( Message-ID: <8k29rl$4d1$1@pyrite.mv.net>  K Either you do not understand how to do asynchronous multi-buffering, or younK have a SCSI driver that won't queue multiple operations concurrently to the  disk.e  H Given that you obtained the same result on NT (which does have competent$ SCSI drivers), I suspect the former.   - bill  1 Rudolf Wingert <win@fom.fgan.de> wrote in message-) news:200007060616.IAA06978@fom.fgan.de...- > Hello, >- > Bill Todd wrotes:i >n > >>>iI > The normal approach to obtaining good streaming performance from a SCSI3 diskJ > is to use asynchronous multi-buffering such that the disk always has one orL > more outstanding requests ready to be satisfied, and can just keep sendingF > or receiving the data without pause.  This does not require enablingK > write-back caching for writes to perform well.  You can ask RMS to handleeI > this for you (using its read-ahead/write-behind options) in some cases.o >iJ > If you weren't doing this, the reason you obtained decent streaming-readJ > performance was because the disk was automatically pre-fetching for you.H > Depending on the intelligence of its algorithm, you might or might not haveD > been able to do somewhat better by doing your own multi-buffering. > <<<. >0E > Tis is not our expirience. Multi-buffering brings a little bit more  perfor-tJ > mance, but that growth that the disk onboard write cache brings. We haveK > tested it with WINNT. Disbaling the disk cache (read or write) brings theeL > performance down from 15.2MB/s to 3.6MB/s (in our test environment; 18.2GBJ > IBM DGHS disk). This value is relative system independent. So you can doF > with RMS what you want, you can't speed up over the meassured value. >n > Regards Rudolf Wingert > G > P.S. The OpenVMS test was a little bit unrealistic, because we used a F >      contigeous file. But I think that on all systems non-contigeous0 >      files will bring a performance degration. >. >u   ------------------------------  $ Date: Thu, 6 Jul 2000 04:06:20 -0400' From: "Bill Todd" <billtodd@foo.mv.com> 2 Subject: Re: highwater marking, speed vs. security( Message-ID: <8k295b$3iu$1@pyrite.mv.net>  J Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de> wrote inJ message news:y4ya3hvuy2.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de...8 > "Robert Deininger" <rdeininger@mindspring.com> writes: > I > > But recently I caught highwater marking causing a big performance hit.G > > on our oldest microvax (with slow disks that can't afford the extra  pain).H > > The files in question were sequential with fixed-length records, and wereJ > > created from Fortran using direct access mode (i.e. by record number). EvenK > > though the file was written almost totally in sequential order, RMS was B > > definitely doing something to every block on each file extend. >aK > When the file is extended, RMS must mark all the new records added to theo > file as empty.  F I assume you mean 'all the new potential locations for records' above.  H But it still should not be necessary when the records are being added inE physically-sequential order - any more than it is when they are addedwI sequentially (for that matter, the ACP shouldn't know the difference - so A perhaps RMS is simply neglecting to special-case this situation).l   - bill   >a > Jan    ------------------------------    Date: 06 Jul 2000 17:51:21 +0200G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de> 2 Subject: Re: highwater marking, speed vs. securityH Message-ID: <y4k8ez2r5y.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  ) "Bill Todd" <billtodd@foo.mv.com> writes:a  M > > When the file is extended, RMS must mark all the new records added to thei > > file as empty.H > I assume you mean 'all the new potential locations for records' above.  L No. AFAIK, RMS direct access files carry a "record is valid" indicator. WhenM the file is extended - and it is always extended by at least a cluster, whichfL usually will contain more than one record - all those new records have to beI marked as invalid. As RMS doesn't have a metadata item (equivalent to theuH high-water mark used by the file system), it must write all those blocksG before it moves the EOF mark. Anyway, that is how it _should_ work 8-).    	Jan   ------------------------------  $ Date: Thu, 6 Jul 2000 12:50:52 -0400' From: "Bill Todd" <billtodd@foo.mv.com> 2 Subject: Re: highwater marking, speed vs. security( Message-ID: <8k2da9$aon$1@pyrite.mv.net>  J Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de> wrote inJ message news:y4k8ez2r5y.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de...+ > "Bill Todd" <billtodd@foo.mv.com> writes:e >dK > > > When the file is extended, RMS must mark all the new records added to  the7 > > > file as empty.J > > I assume you mean 'all the new potential locations for records' above. >.I > No. AFAIK, RMS direct access files carry a "record is valid" indicator.D  K Depends on what you mean by an "RMS direct access" file:  the original postaG suggests that Fortran uses RMS Sequential files for this purpose, but I* can't vouch for that.g  E RMS Relative files indeed have a 'record exists' bit in each record'sgG control byte (and a 'record deleted' bit as well, so that a user random G access failure can differentiate between the two conditions).  But thatWK doesn't intrinsically require that they be treated any differently than RMSwH Sequential files as far a high-water marking and initialization of spaceJ beyond EOF is concerned (though I'd expect the algorithms not to leave the, last bucket in the file partially formated).  E RMS Sequential files with fixed-length records, which can be accessediE randomly in pretty much the same manner as Relative files (at least I1E *think* I remember writing code in RMS-11 to provide 'relative recordRJ number' access to fixed-format SEQ files; it's certainly easy enough to doL externally, though you'd have to round up record lengths to an even boundaryH and take into account whether they could cross block boundaries), do notG associate any such control information with each record.  But as I saidyI above, it's not clear that the presence of such control information wouldsG affect the need for high-water-mark operations:  as long as records aremF added in sequential order (as the original post stated was usually the7 case), it should be able to be handled just as a normall- sequentially-accessed Sequential file can be.e   - bill    WhenmI > the file is extended - and it is always extended by at least a cluster,o whicheK > usually will contain more than one record - all those new records have to  beK > marked as invalid. As RMS doesn't have a metadata item (equivalent to thecJ > high-water mark used by the file system), it must write all those blocksI > before it moves the EOF mark. Anyway, that is how it _should_ work 8-).  >  > JanS   ------------------------------  % Date: Thu, 06 Jul 2000 09:41:05 -0500e/ From: Scott Vieth <svieth@ameritech.net.nospam>dA Subject: Re: How can I "replicate" Volume Shadowing on a Sun box?B4 Message-ID: <39649A81.F616D035@ameritech.net.nospam>  > Normally, these approaches would work fine but we like to haveA a totally separate copy of the database as a "safety net".  WhileDH we are backing-up the database disks that were split off from the shadowF sets, our nightly offline processing is running against the disks thatE are still mounted.  If the nightly processing blows up, we can alwaysiB go back to the point right before the processing started by simply? dismounting the database drives and re-mounting the drives thatgB were split off before processing started.  It gives us an easy way to get things running again.   Thanks,  -Scott   "Richard B. Gilbert" wrote:M  ? >         I'm currently doing "hot" backups under Oracle 7.1.3.c >-M >         From memory, it's something like "ALTER TABLESPACE BACKUP" and thenoH > do a normal VMS backup of the files in the tablespace.  You have to be6 > running in "logging" mode in order for this to work. >-( > Message text written by Warren SpencerL > >Another approach to this type of problem is to use Oracle's native backupE > facilities, which permit "hot" backups.  I believe this facility iso > availablea7 > in 7.3 and later versions.  Hot backups are done on an > tablespace-by-tablespace: > basis, and do not require that you quiesce the database. > <    ------------------------------  % Date: Thu, 06 Jul 2000 06:07:53 -0700l4 From: mapryan <mike.ryanNOmiSPAM@redmar.com.invalid> Subject: Re: LA36 Escape codes9 Message-ID: <02a4ad00.a741c8c4@usw-ex0105-038.remarq.com>g   Paul  < It's the new Compaq LA36. What I needed was the simple codes@ that set stuff like page length, CPI, etc. I think that I've got7 most from getting a similar list for the LA50 from DSN.e   Cheers   Mike  9 Please excuse the crappy, automatically generated adverts, below...Sent from remarq.com !; -----------------------------------------------------------t    ; -----------------------------------------------------------   7 Got questions?  Get answers over the phone at Keen.com.  Up to 100 minutes free!- http://www.keen.coma   ------------------------------   Date: 6 Jul 2000 10:07:42 -0500n, From: koehler@eisner.decus.org (Bob Koehler)0 Subject: Re: Multilingual System Dictionary(ies)+ Message-ID: <V0DZoobYbD$k@eisner.decus.org>e  \ In article <39639D2B.B224B2F2@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes: > O > since the file is essentially a fixed length sequential file, isn't it just aRO > matter of adding your records (padded with nulls, I beleive) and then sortingu > the file ? >   F    Gee, on all my systems, it's a keyed-indexed file.  Appending to it    should be sufficient:  A    VMS$PASSWORD_DICTIONARY.DATA;1            File ID:  (2297,1,0)]      [...]  6    File organization:  Indexed, Prolog: 3, Using 1 key'                              In 2 areasu   ------------------------------  % Date: Thu, 06 Jul 2000 13:13:26 -0400l- From: JF Mezei <jfmezei.spamnot@videotron.ca>t0 Subject: Re: Multilingual System Dictionary(ies), Message-ID: <3964BE33.7AF6CDFC@videotron.ca>   Bob Koehler wrote:H >    Gee, on all my systems, it's a keyed-indexed file.  Appending to it >    should be sufficient: > C >    VMS$PASSWORD_DICTIONARY.DATA;1            File ID:  (2297,1,0))  L Oops ... ever since I upgraded to 7.x, the new format of DIR/FULL has gottenM me into trouble... that Shelved state and caching attribute lines should haved been put at the end...  @ mea culpa ! It is indexed allright. I just didn't look properly.   ------------------------------  # Date: Thu, 06 Jul 2000 12:05:22 GMTr. From: Michael Austin <maustin@nc.prestige.net>) Subject: Multinet port numbers being used)/ Message-ID: <396483A6.9D90E78C@nc.prestige.net>t  , This is a multi-part message in MIME format.& --------------009AB9D49FAEB6F046F6C46D* Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit-  F does anyone know how to tell Multinet to show you what ports are beingB used... I am interested in knowing if another package is using the  default port for http (port 80).   Thanks,  Michael Austin  & --------------009AB9D49FAEB6F046F6C46D- Content-Type: text/x-vcard; charset=us-ascii;u  name="maustin.vcf"e Content-Transfer-Encoding: 7bit , Content-Description: Card for Michael Austin  Content-Disposition: attachment;  filename="maustin.vcf"    begin:vcard  n:Austin;Michael a tel;work:704-947-1089t x-mozilla-html:FALSE org:Michael Austin, Incw
 adr:;;;;;; version:2.1e+ email;internet:michaelaustininc@hotmail.com  title:President  x-mozilla-cpt:;0 fn:Michael E. Austin	 end:vcard   ( --------------009AB9D49FAEB6F046F6C46D--   ------------------------------  % Date: Thu, 06 Jul 2000 10:48:38 -0300C1 From: "Boyle, Darren" <boyledj@bankofbermuda.com>v- Subject: RE: Multinet port numbers being usedmK Message-ID: <9F664D538536D411BD3200508B6FF01A3335EB@bdant027.bda.bobda.com>o   Try MULTINET SHOW/CONNECTION - Darren   > ----------5 > From: 	Michael Austin[SMTP:maustin@nc.prestige.net]m( > Sent: 	Thursday, July 06, 2000 9:05 AM > To: 	Info-VAX@Mvb.Saic.Com, > Subject: 	Multinet port numbers being used >  > <<File: maustin.vcf>>hH > does anyone know how to tell Multinet to show you what ports are beingD > used... I am interested in knowing if another package is using the" > default port for http (port 80). > 	 > Thanks,s > Michael Austin >     F **********************************************************************C This message and any files transmitted with it are confidential and J may be privileged and/or subject to the provisions of privacy legislation.M They are intended solely for the use of the individual or entity to whom theytL are addressed. If the reader of this message is not the intended recipient, B please notify the sender immediately and then delete this message.I You are notified that reliance on, disclosure of, distribution or copyings of this message is prohibited.   Bank of Bermuda F **********************************************************************   ------------------------------  % Date: Thu, 06 Jul 2000 10:06:08 -0400,; From: Andrade Jay A CONT NPRI <AndradeJA@Npt.NUWC.Navy.Mil><- Subject: RE: Multinet port numbers being usedaT Message-ID: <CEB144BF16CCD2118F4800805FE6A25301450BA1@NPRI54EXC07.NPT.NUWC.NAVY.MIL>   MULTINET CONFIG/SERVER and then SERVER-CONFIG>SHOW   -----Original Message-----5 From: Michael Austin [mailto:maustin@nc.prestige.net]m% Sent: Thursday, July 06, 2000 8:05 AMs To: Info-VAX@Mvb.Saic.Comi) Subject: Multinet port numbers being usedj    F does anyone know how to tell Multinet to show you what ports are beingB used... I am interested in knowing if another package is using the  default port for http (port 80).   Thanks,e Michael Austin   ------------------------------  # Date: Thu, 06 Jul 2000 10:27:56 GMTo From: pavichal@my-deja.com1 Subject: Re: Netscape Fasttrack for OpenVMS 7.2-1c) Message-ID: <8k1mv5$o1o$1@nnrp1.deja.com>e  - In article <3963839B.EF5BD671@theblakes.com>,t*   Colin Blake <colin@theblakes.com> wrote:F > You might want to try enabling and applying "CGI File Type", just in case.f >i >b Excellent!! That worked.E I want to be able to perform other functions like pinging and viewing.H queues etc.. but I get insufficient privs. Do you know what the best wayF to allow fasttrack access to other information is without comprimising the system??    & Sent via Deja.com http://www.deja.com/ Before you buy.S   ------------------------------  $ Date: Thu, 6 Jul 2000 03:55:43 -0400' From: "Bill Todd" <billtodd@foo.mv.com>.; Subject: Re: OpenVMS loses big, was:  RE: Compaq advertisese( Message-ID: <8k1duq$7ln$1@pyrite.mv.net>  L You might be well-advised to read up on the existing discussion in this area' before putting your foot in your mouth.e   - bill  6 <yyyc186.illegaltospam_@flashcom.net> wrote in message3 news:3963f11b$2$lllp186$mr2ice@news.flashcom.com... 1 > In <8jvmhh$b7@gap.cco.caltech.edu>, on 07/05/00sD >    at 10:33 PM, mathog@seqaxp.bio.caltech.edu (David Mathog) said: >u >fJ > >For a UNIT text file operation, going RAMDISK to RAMDISK on OpenVMS, orL > >file cache to file cache on Linux, on identical DS10s, the result is thatK > >the Linux system is 2.5-6.5 times faster.  This is for an operation like K > >"read text record from input, write text record to output" - pure IO. It D > >doesn't matter if you do this in 1 file or in 1000 you're alreadyB > >starting out with the Unix systems "lighter" text file handlingI > >mechanisms 3X faster than those on OpenVMS.  And it goes downhill fromeE > >there, rapidly, because the lack of file effective file caching on  >kH > Sure, for accademia speed will always win out over integrity.  You areI > overlooking the stability FILES 11 gives you over the unix disk stream.eF > In the real world, where integrity is king, we want to know the dataE > actually made it to the disk (hopefully into a nice record orienteduK > indexed file and not some fprintf() stream) when the call returns.  WriteiI > caching may seem like a wonderfull thing, but it makes check pointing a0 > nightmare. >  > Roland >o >  > --= > -----------------------------------------------------------eF > yyyc186@flashcom.net              To Respond delete ".illegaltospam"8 >                             MR/2 Internet Cruiser 1.52: >                             For a Microsoft free univers= > -----------------------------------------------------------  >    ------------------------------   Date: 6 Jul 2000 15:59:48 GMTe2 From: mathog@seqaxp.bio.caltech.edu (David Mathog); Subject: Re: OpenVMS loses big, was:  RE: Compaq advertisesa, Message-ID: <8k2adk$dlg@gap.cco.caltech.edu>  g In article <LoKsvv2fycO9@eisner.decus.org>, Kilgallen@eisner.decus.org.nospam (Larry Kilgallen) writes:s >hD >You do seem to claim that your complaint is with VMS in general andD >only incidentally does it slip out that your concerns are only withD >C.  You blame RMS and they reveal that you are actually calling RMS >through the C Runtime library.   G Well, yes, sure.  I can't distinguish between RMS, and C on top of RMS,dH based on tests written in C - but I can hardly run tests written to call@ RMS on Linux!  The closest I came to a test of "pure" RMS was a E comparison of APPEND on VMS, with >> on Linux, the latter being much n@ faster.  I suppose that one could compare RMS calls with write()E calls, but that's moving far away from the initial problem, which was.3 a comparison of portable code run on both systems. e   >b@ >Try some tests in Ada, Pascal or Fortran, or else make it quite? >clear in your strident posts that yours is a C-only viewpoint.N  ? Fair enough.  The only other language I have on both systems isaB Fortran 90.  Here's one simple test - all it does is write a large file:   
 ! MAKETESTF.flE !   makes a 16000 entry fasta file, each containing a 500 bp sequencea ! 2 ! more or less verbatim translation of maketest.c. !a         integer i,j  ! B         open(unit=10,name='test.nfa',form='FORMATTED',status='NEW'+         1 ,recl=132,carriagecontrol='LIST')o         do i=1,16000           write(10,1000),i           do j=1,10r             write(10,1005)           end do         end do 1000    format('>test',i5.5)D 1005    format('AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA')         stop         endg  D (Note that the file it produces is not identical at the binary level5 on the two platforms, but it is at the "text" level.)y  B Compaq F90 on both platforms, both DS10s.  Run times on a RAMdisk:  C    OpenVMS 7.2-1 using RAMdisk    2.05 seconds  (extend=0,buffer=0)IH    OpenVMS 7.2-1 using RAMdisk    1.78 seconds  (extend=8000,buffer=255).    Linux, using file cache        1.20 seconds  I So in this instance Linux beats OpenVMS by either 48% (tuned) or 71% (not H tuned).  OpenVMS is still slower, but not quite so egregiously as it is F with C.  I reran all the tests on physical disks +/- highwater markingG to put it all into perspective, using the IO benchmark testing programs  at:   <   http://seqaxp.bio.caltech.edu/pub/SOFTWARE/MYBENCHMARK.ZIP  D on two DS10s, one VMS 7.2-1, the other RH 6.2, both with intraserverK U2W controllers and LVD disks, lots of memory.  I reran these these partly oC because I've discovered a problem when PIPE is used to pull out thesC timing lines from the small number of output lines produced by the l- test procedure.  That is, when I ran it like:   , $ pipe @testsplit | search sys$input elapsed  6 some tests came out much worse than when I did it like   $ @testsplit  B Compare columns 4 and 5 to see the effect.  I knew PIPE throughputC was bad, but really didn't expect this result at all.  Keep in mind:3 that the output from the procedure just looks like:D  $ MAKETESTF: Create test file, fortranD  Elapsed time in 1/100ths of a second: 243, Device operations = 1127   MAKETEST:  Creating test file,CSD  Elapsed time in 1/100ths of a second: 954, Device operations = 1098   etc.,u  E so we're talking a mere handful of bytes moving through the pipe.  (IRB suspect that the RMS settings are somehow interacting with PIPE to produce this bizarre result.)     . TEST            1     2    3      4     5    65 OS            VMS   VMS   VMS   VMS   VMS  VMS  LINUXoB disk         hard  hard  hard  hard  hard  RAM  hard (file cached). buffer          0     0   255   255   255    0. extend          0     0  2000  2000  2000    0. highwater     Yes    NO   YES    NO    NO   NO. pipe?          NO    NO    NO    NO   YES   NO  5 MAKETESTF:  26.26 22.09  3.86  2.20 22.07 2.05   1.2045 MAKETEST:   25.60 21.28 10.90  9.15 21.17  .77    .30R5 MYSTART:      .04   .04   .03   .03   .03  .01    .0085 MYOPENIN      .02   .05   .10   .03   .03  .01    .00h5 MYREAD        .46   .49   .52   .57   .45  .38    .10g5 MYOPEN       7.69  7.78  7.74  7.75  7.68  .67    .10-5 MYSPLIT     42.83 37.75 54.75 25.25 37.56 1.38    .40s    A So, Larry is right, other languages are not so bad as C.  But I'mrE also right - you have to go to a lot of work to get the IO even closeeC on OpenVMS to what is obtained with no effort on Linux (Unix).  AndrB while it's nice to know that Fortran is quicker than C, it doesn't> help me at all because all the code I receive is written in C,D something derived from C (like Perl), or even C++ - been a long time1 since I've seen anything new written in Fortran. s   Regards,   David Mathog mathog@seqaxp.bio.caltech.eduo? Manager, sequence analysis facility, biology division, Caltech  J **************************************************************************J *                                RIP VMS                                 *J **************************************************************************   ------------------------------   Date: 6 Jul 2000 13:27:26 -0500e9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)e; Subject: Re: OpenVMS loses big, was:  RE: Compaq advertisese+ Message-ID: <p6MfUWVNP5Vv@eisner.decus.org>s  a In article <8k2adk$dlg@gap.cco.caltech.edu>, mathog@seqaxp.bio.caltech.edu (David Mathog) writes:   0 > TEST            1     2    3      4     5    67 > OS            VMS   VMS   VMS   VMS   VMS  VMS  LINUXrD > disk         hard  hard  hard  hard  hard  RAM  hard (file cached)0 > buffer          0     0   255   255   255    00 > extend          0     0  2000  2000  2000    00 > highwater     Yes    NO   YES    NO    NO   NO0 > pipe?          NO    NO    NO    NO   YES   NO > 7 > MAKETESTF:  26.26 22.09  3.86  2.20 22.07 2.05   1.20o7 > MAKETEST:   25.60 21.28 10.90  9.15 21.17  .77    .30 7 > MYSTART:      .04   .04   .03   .03   .03  .01    .00n7 > MYOPENIN      .02   .05   .10   .03   .03  .01    .00f7 > MYREAD        .46   .49   .52   .57   .45  .38    .10-7 > MYOPEN       7.69  7.78  7.74  7.75  7.68  .67    .10e7 > MYSPLIT     42.83 37.75 54.75 25.25 37.56 1.38    .40  >  > C > So, Larry is right, other languages are not so bad as C.  But I'm G > also right - you have to go to a lot of work to get the IO even closemE > on OpenVMS to what is obtained with no effort on Linux (Unix).  AndtD > while it's nice to know that Fortran is quicker than C, it doesn't@ > help me at all because all the code I receive is written in C,F > something derived from C (like Perl), or even C++ - been a long time3 > since I've seen anything new written in Fortran. 6  > Well I think it _should_ help you know who to talk to at DECUS> (in your region) next fall.  The caching representative may be> the one who is going to provide the next near-term assistance,< but the C RTL representative is the one with whom you should: really spend time.  There have been a lot of compatibility; complaints over the years, and perhaps they don't even knowy# people are looking for performance.    ------------------------------  $ Date: Thu, 6 Jul 2000 12:12:25 -0400' From: "Bill Todd" <billtodd@foo.mv.com>b: Subject: Re: OpenVMS loses big, was: RE: Compaq advertises( Message-ID: <8k2b29$5df$1@pyrite.mv.net>  . <Shane.F.Smith@Healthnet.com> wrote in message= news:OFD316202B.029F77DD-ON88256914.0015CA60@HEALTHNET.COM...  >mJ > Sadly, this is not usually how things happen. In the real world speed is3 > often king because PHMs don't get the advantages:a >tH > PHM: Somebody just told me Unix's file system is faster than VMS's. Is that > true?bK > Tech:     Well, yes, it's about x% faster, but VMS's filesystem is safer.e  L Then again, in the real world sometimes the technical people don't know whatL they're talking about any more than their managers do.  This particular techH needs to ask which Unix is under consideration before being able to makeK such a statement authoritatively, but (as is too often the case) simply has J assumed that his own limited experience (or, worse, something someone told5 him over a beer one night) has general applicability.7   - bill   ------------------------------  % Date: Thu, 06 Jul 2000 11:24:59 -0500 / From: "Stuart, Ed" <Ed.Stuart@austinenergy.com>n5 Subject: RE: Oracle Press Release, incl. OpenVMS info T Message-ID: <CB874B506A79D1118FBC006097306B8903120434@ohms.electric.ci.austin.tx.us>  ? For what its worth, we received Oracle 8.1.6 for OpenVMS Alpha.-  $ Ed Stuart                           ( Manager, Systems and Desktop Services	   Information Technology Services- City of Austin, Austin Energy  Ed.Stuart@austinenergy.com  + "Glittering prizes and endless compromises k. shatter the illusion of integrity" - Neil Pert  B *Please apply a generous amount of all the usual disclaimers here*     -----Original Message-----0 From: Main, Kerry [mailto:Kerry.Main@COMPAQ.COM]# Sent: Friday, June 30, 2000 5:37 AM9 To: Info-VAX@Mvb.Saic.Coma5 Subject: RE: Oracle Press Release, incl. OpenVMS info      David,  C >>> Now, how do we go about getting this out to the mass media? <<<   F Well, for starters, the press release info is on the front page of the Compaq and Oracle home pages.e  J In addition, the Oracle article is on the front page of eWeek (formerly PC	 Week) at:: <http://www.zdnet.com/eweek/>DD <http://www.zdnet.com/eweek/stories/general/0,11011,2596660,00.html>D <http://www.zdnet.com/eweek/stories/general/0,11011,2596485,00.html>    quote from 1st article in eWeek:  L "The centerpiece of the company's Oracle.Now makeover is the Oracle InternetL Application Server 8i, which Oracle officially rolled out at an event at its& Redwood Shores, Calif., headquarters.   K The app server is filled with new features designed to make it the nexus ofpK an e-business' computing universe. Its main selling point is the ability toTI cache data from the Oracle 8i database. This feature reduces the time Web.J surfers have to wait by putting much of the data closer to the Web server.L It also improves the scalability and performance of the database by reducingI the number of calls it gets from those Web applications, officials said. -  " Analysts liked the caching plans.   H "It's standard practice for developers to put caching schemes into theirI applications," said Thomas Murphy, program director at the Meta Group, ofiI Stamford, Conn. "The advantage for programmers is they will have a way toA0 have seamless high-performance cache built in."   J The Java-based Internet Application Server 8i also incorporates the ApacheI Web server, the query and reporting services of the Oracle Discoverer andoI Oracle Reports business intelligence applications, and Component ServiceshA for building applications from pre-built, Java-based components. o  G [and it will be available on all Compaq platforms - including OpenVMS.]-   :-)r   Regards,  
 Kerry Main Senior Consultant,
 Compaq Canadan Professional Servicesa Voice : 613-592-4660 FAX   : 819-772-7036 Email : kerry.main@compaq.come       -----Original Message-----< From: David J. Dachtera [mailto:djesys.nospam@earthlink.net]& Sent: Thursday, June 29, 2000 11:17 PM To: Info-VAX@Mvb.Saic.Como5 Subject: Re: Oracle Press Release, incl. OpenVMS info-     "Main, Kerry" wrote: > ! > As a followup to Steve's note -- > K > A few pointers for those folks who have beed asking for more applicationso > and marketing of OpenVMS:M > H > <http://www.oracle.com/corporate/press/index.html?228177.html> (Oracle Press@
 > release) > ; > <http://www.oracle.com/tellmemore/?228177> (Tell me more)- >  > L <http://www1.compaq.com/pressrelease/0,1494,wp%7E14583_2%21ob%7E31866_1_1,00 > .html> (Compaq Press release)@ > I > And in case, some folks missed the other posting (and I like the splashi > stuff :-)o/ > <http://www.openvms.digital.com/e-postcard1/>h/ > <http://www.openvms.digital.com/e-postcard2/>kJ > (under each url is a pointer that you can save and run directly on a PC) [snip]  ; Now, how do we go about getting this out to the mass media?*  0 Preaching to the choir = spitting into the wind!  H (Pardon my tone, but I don't know how to drive that point home anymore!)   -- r David J. Dachterav dba DJE Systems " http://home.earthlink.net/~djesys/  : Unofficial Affordable OpenVMS Home Page and Message Board:+ http://home.earthlink.net/~djesys/vms/soho/o   ------------------------------  % Date: Thu, 06 Jul 2000 07:42:13 -0500e" From: Earl Lakia <lakia@ipact.com>' Subject: Problems with DCOM for OpenVMSu) Message-ID: <39647EA5.6F705CEB@ipact.com>i  = I've been trying to get DCOM up under OpenVMS.  Following the*? installation procedure,  I tried to populate the registry using ? dcom$config.  When selecting option #3 as suggested, it says itTA takes a long time, but I left it running over night and it hadn'tm< completed.  Anyway, a show system is attached to verify that5 all processes are running.  The procedure I used was:N  0 executed: reg$config and modified sylogicals.com executed: nta$startup_nt_acmee executed: dce$rcp_startup.coms executed: dcom$setup, option 3  ! Anyhelp from you gerus out there?S   -earl   7 -------------------------------------------------------   - Populate the registry database on node IPCAL3a    =         [ Starting to Populate the COM for OpenVMS Registry ]u  F          Populating the Registry for OpenVMS may take up to 15 minutes"          depending on your system.   Enter Y[ES] to continue: yes  - DCOMREG-I-DONOTABORT, Do NOT abort procedure.c  2 --------------------------------------------------   IPCAL3>Earl$ show sys B OpenVMS V7.2-1  on node IPCAL3   6-JUL-2000 08:25:33.62  Uptime  0 15:57:17F   Pid    Process Name    State  Pri      I/O       CPU       Page flts Pagess< 00000201 SWAPPER         HIB     16        0   0 00:00:00.97 0      0< 00000204 LANACP          HIB     13       55   0 00:00:00.02
 107    144< 00000206 IPCACP          HIB     10       10   0 00:00:00.00	 32     45 < 00000207 ERRFMT          HIB      8     1851   0 00:00:00.55	 88    103D< 00000209 OPCOM           HIB      7      579   0 00:00:00.07
 123     45< 0000020A AUDIT_SERVER    HIB     10       89   0 00:00:00.02
 152    163< 0000020B JOB_CONTROL     HIB     10      204   0 00:00:00.02	 59     83e< 0000020D SECURITY_SERVER HIB     10       42   0 00:00:00.06
 707    192< 0000020E DNS$ADVER       LEF      5    11894   0 00:00:02.60 2155    733t< 0000020F LES$ACP_V30     HIB      9      129   0 00:00:00.08	 75     88 < 00000210 NET$ACP         HIB      4       77   0 00:00:00.03
 139    161< 00000211 REMACP          HIB     12       18   0 00:00:00.00	 36     24x< 00000212 NET$EVD         HIB      6       20   0 00:00:00.24
 794    171< 00000213 DTSS$CLERK      LEF     12     1003   0 00:00:00.25
 163    178< 00000214 TP_SERVER       HIB     10     3835   0 00:00:03.15	 81    104 < 00000215 TCPIP$INET_ACP  HIB     10      215   0 00:00:00.05
 227    162< 0000021C LATACP          HIB     14        9   0 00:00:00.16	 74     60m< 0000021D CCAGENTRA200    HIB      6   243470   0 00:00:11.26
 698    305< 0000021F DECW$SERVER_0   HIB      8      765   0 00:00:01.11
 733    881< 00000220 DTLOGIN         LEF      5      143   0 00:00:00.09
 327    279< 00000222 DTGREET         LEF      4      470   0 00:00:00.93
 536    440  S < 00000224 DECW$TE_0224    COM      4   135287   0 00:00:06.41
 903    857< 00000227 LAKIA_2         LEF      6    22046   0 00:00:13.52 31505    505< 00000228 ACME_SERVER     HIB     10      134   0 00:00:00.15 503    423 M< 0000022A DCE$RPCD        HIB      8      103   0 00:00:00.09
 585    297< 0000022B TCPIP$PWIP_ACP  HIB     10       79   0 00:00:00.03
 127    147< 0000024A NETBIOS         HIB      4     4776   0 00:00:00.24	 63     79o< 0000024C PWRK$KNBDAEMON  HIB     12       69   0 00:00:00.05
 243    297< 0000024E PWRK$LICENSE_R  HIB     11      178   0 00:00:00.17
 496    355< 00000250 PWRK$LICENSE_S  HIB      6      248   0 00:00:00.22
 774    313< 00000253 PWRK$MASTER     HIB      6      229   0 00:00:00.13
 415    151< 00000255 PWRK$LMMCP      HIB     11     6043   0 00:00:01.49 1027    411-< 00000257 PWRK$MONITOR    HIB      6       73   0 00:00:00.19
 588    214< 00000259 PWRK$ADMIN_0    LEF      6       58   0 00:00:00.05
 144     94F 0000025B PWRK$LMSRV      HIB     11    10837   0 00:00:02.12      1576 1178< 0000025D PWRK$LMDMN      HIB     11     3026   0 00:00:00.68 1161    437s< 0000025F PWRK$LMBROWSER  HIB      9     3260   0 00:00:00.50
 726    628< 00000272 REGISTRY_SERVER HIB     10      732   0 00:00:20.50 487    332 M< 00000274 LAKIA_1         CUR      7      307   0 00:00:00.17
 839    123< 00000277 DECW$TE_0277    LEF      6     4122   0 00:00:01.47 1379    828 < 0000027A LoganHM_1       LEF      4     4437   0 00:00:02.53 2089     70r --
 Earl D. Lakia 0 Senior Staff Engineer         Web: www.ipact.com4 Snail Mail:                   Email: lakia@ipact.com
 IPACT Inc.1 260 S. Campbell St.           Phone: 219-464-7212  Valparaiso, IN 46383   ------------------------------  $ Date: Thu, 6 Jul 2000 12:44:30 -04003 From: "Gaitan D'Antoni" <gaitan.dantoni@compaq.com>n+ Subject: Re: Problems with DCOM for OpenVMSa* Message-ID: <8k2d42$rkb@usenet.pa.dec.com>  < Earl Lakia wrote in message <39647EA5.6F705CEB@ipact.com>...> >I've been trying to get DCOM up under OpenVMS.  Following the@ >installation procedure,  I tried to populate the registry using@ >dcom$config.  When selecting option #3 as suggested, it says itB >takes a long time, but I left it running over night and it hadn't= >completed.  Anyway, a show system is attached to verify that 6 >all processes are running.  The procedure I used was: >'1 >executed: reg$config and modified sylogicals.com  >executed: nta$startup_nt_acme >executed: dce$rcp_startup.com >executed: dcom$setup, option 3   E It appears that the REGISTRY_SERVER is not responding to the populatemL requests. This may be caused by missing components or components having beenI started out of order. I suggest rebooting the system and making sure thatvK Advanced Server has been started. Then, execute @SYS$STARTUP:DCOM$SETUP andoI select option 3 to populate the registry. DCOM$SETUP will start all othergI components that are required (ACME, RPC, ETC...)  so there is no need fori you to explicitly start them.0   ------------------------------   Date: 6 Jul 2000 15:09:30 GMTi0 From: John Quintini <john.quintini@maf.nasa.gov> Subject: Test!! Ingnore!!), Message-ID: <3964A12A.8D13D513@maf.nasa.gov>   Testinga   ------------------------------  % Date: Thu, 06 Jul 2000 13:30:02 +0100r- From: Tim Llewellyn <tim.llewellyn@bbc.co.uk>  Subject: Re: Trapping a STOP/IDt) Message-ID: <39647BCA.AD90C6AE@bbc.co.uk>O   Earl Lakia wrote:P  ? > Tim's Solution are good.  I used  a simple foriegn command toe= > to this with the lock manager and a DCL script (used it forhA > automatically switching applications on a cluster when a systemuE > faulted).  I've also used the mailbox method that Tim stated, againsA > a good solution.  However, with both, you need a second process : > to do the cleanup.  This can sometimes can be a problem.  ; I'd actually assumed a second process was needed. Otherwise C you're talking kernel mode exit handlers or something like that. OfaG course, with the second program watching the first, do you need a thirdtH watching the second? Not if the second is well designed and implemented.   --6 Tim Llewellyn, OpenVMS Infrastructure, Remarcs Project0 MedAS at the BBC, Whiteladies Road, Bristol, UK.A Email tim.llewellyn@bbc.co.uk. Home tim.llewellyn@cableinet.co.ukn  A I speak for myself only and my views in no way represent those of  MedAS or the BBC.(   ------------------------------  % Date: Thu, 06 Jul 2000 13:32:18 +0100d- From: Tim Llewellyn <tim.llewellyn@bbc.co.uk>  Subject: Re: Trapping a STOP/IDS) Message-ID: <39647C52.CBCA5637@bbc.co.uk>S   Hoff Hoffman wrote:h   >hD >   And please note that no matter what you do, a hard-stop is stillD >   possible as a result of a system (eg: power) failure -- you willG >   likely want to survive this case, so you thus have the same problemNC >   of handling (recovering from) $delprc.  Thus the correct way to E >   "block" a $delprc is to simply remove the privilege(s) of anybodyYH >   that is unauthorized to or that proves otherwise unable to correctly< >  use it.  (Administrative retribution, in other words. :-)  D If you use the lock manager approach then the monitoring process can be on a different cluster node.7  --46 Tim Llewellyn, OpenVMS Infrastructure, Remarcs Project0 MedAS at the BBC, Whiteladies Road, Bristol, UK.A Email tim.llewellyn@bbc.co.uk. Home tim.llewellyn@cableinet.co.uk   A I speak for myself only and my views in no way represent those of  MedAS or the BBC.k   ------------------------------  % Date: Thu, 06 Jul 2000 13:36:03 +0100t- From: Tim Llewellyn <tim.llewellyn@bbc.co.uk>e Subject: Re: Trapping a STOP/ID ) Message-ID: <39647D33.87E7C329@bbc.co.uk>t   "John E. Malmberg" wrote:s   >e >d> >     Using termination mailboxes for process termination does> >     not work if the process does not have suffient resources0 >     available to send the termination message.  )  Well, you learn something every day :-).    --6 Tim Llewellyn, OpenVMS Infrastructure, Remarcs Project0 MedAS at the BBC, Whiteladies Road, Bristol, UK.A Email tim.llewellyn@bbc.co.uk. Home tim.llewellyn@cableinet.co.uk7  A I speak for myself only and my views in no way represent those ofe MedAS or the BBC.V   ------------------------------  % Date: Thu, 06 Jul 2000 07:44:50 -0700r5 From: Richard  <maher_rjNOmaSPAM@hotmail.com.invalid>e Subject: Re: Trapping a STOP/IDu9 Message-ID: <01f8d464.612707d9@usw-ex0110-076.remarq.com>    Hi,   - How long do messages live in this news group?   : I finally clicked on remarq and found a very useful search2 engine but can't find the macro I was looking for.  . Does anyone have a copy that they can mail me?  8 Someone was complaining about a missing "?" in the macro: but the problem dealt with validating previous mode access to a homed argument list.    Regards Richard Maher.     * Sent from AltaVista http://www.altavista.com Where you can also find related Web Pages, Images, Audios, Videos, News, and Shopping.  Smart is Beautiful    ------------------------------   Date: 6 Jul 2000 12:27:09 -0500 9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen). Subject: Re: Trapping a STOP/ID7+ Message-ID: <jKiTa9f6UWO7@eisner.decus.org>   Y In article <39647C52.CBCA5637@bbc.co.uk>, Tim Llewellyn <tim.llewellyn@bbc.co.uk> writes:o >  >  > Hoff Hoffman wrote:  >  >>E >>   And please note that no matter what you do, a hard-stop is still.E >>   possible as a result of a system (eg: power) failure -- you willeH >>   likely want to survive this case, so you thus have the same problemD >>   of handling (recovering from) $delprc.  Thus the correct way toF >>   "block" a $delprc is to simply remove the privilege(s) of anybodyI >>   that is unauthorized to or that proves otherwise unable to correctly = >>  use it.  (Administrative retribution, in other words. :-)v > F > If you use the lock manager approach then the monitoring process can! > be on a different cluster node.t  > I am not sure how your comment relates to the part you quoted.- Doesn't $DELPRC work across the cluster too ?e   ------------------------------  % Date: Thu, 06 Jul 2000 13:07:57 -0400h- From: JF Mezei <jfmezei.spamnot@videotron.ca>  Subject: Re: Trapping a STOP/IDR, Message-ID: <3964BCEB.AAD2160F@videotron.ca>   Tim Llewellyn wrote:F > If you use the lock manager approach then the monitoring process can! > be on a different cluster node..  E But don't you have to grant the application SYSLCK privilege for thataD functionality ? I have had customers deny request to have SYSLCK forN applications because they felt it was too dangerous. How common is the fear of granting SYSLCK ?t   ------------------------------   Date: 6 Jul 2000 14:22:05 -0500l9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)e Subject: Re: Trapping a STOP/ID + Message-ID: <fJ+dfwHh9nPG@eisner.decus.org>i  \ In article <3964BCEB.AAD2160F@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes: > Tim Llewellyn wrote:G >> If you use the lock manager approach then the monitoring process canu" >> be on a different cluster node. > G > But don't you have to grant the application SYSLCK privilege for thatpF > functionality ? I have had customers deny request to have SYSLCK forP > applications because they felt it was too dangerous. How common is the fear of > granting SYSLCK ?   B Like all privileges, SYSLCK should never be granted to someone who does not need it.e  I Without SYSLCK a user is restricted to communicating between applicationsOH in the same UIC group.  I cannot envision a Watcher-Watchee relationshipF that would require the two be in separate UIC groups, so SYSLCK should not be a problem.d   ------------------------------  $ Date: Thu, 6 Jul 2000 10:53:08 -0400- From: "Peter Weaver" <peter.weaver@stelco.ca> : Subject: Re: Upgrade to VMS 7.2 fails on Help Message File. Message-ID: <sm97ajargas60@corp.supernews.com>  ? At this point in the install booting from root 0 (where 6.2 was2F installed) brings up Standalone backup since all of the 6.2 files haveD been deleted two steps earlier. Booting from root F runs the commandE procedure that was failing. In the end I had to boot conversationally F in root F and edit [SYSF.SYSEXE]UPDATE.KIT so that RESTORE_VMS_MSG wasB set to 0. When I was done the help file was in DKA1:[HELP_MESSAGE] where I wanted it.  E If I was doing this on a real system I would log a support call sinceeE this part of the VMS upgrade seems to be broken, but since this is one, my hobbyist Charon-VAX I can not log a call.   -- Peter Weaver  " Mike Drabicky wrote in message ...2 >On Wed, 5 Jul 2000 11:40:29 -0400, "Peter Weaver"  ><peter.weaver@stelco.ca> wrote: >rB >>I am upgrading my Hobbyist Charon-VAX from V6.2 to V7.2 and haveE >>placed the Help message on a different disk. After answering all ofnF >>the questions and rebooting the node I see the following (the system: >>drive is DKA0, the drive for the help messages is DKA1); >>% >>Any idea on what is happening here?e > C >So move your help file back to the common part of SYS$HELP for themD >install then copy it back to your normal location afterwards. Maybe@ >I'm missing something here but that certainly seems the easiest >solution to this issue. >c >Mike Drabicky   ------------------------------  $ Date: Thu, 6 Jul 2000 09:03:49 -0400  From: "Jim Kim" <sjkim@istar.ca> Subject: Vax in Canada ?8 Message-ID: <1h%85.22850$Ip.558681@cac1.rdr.news.psi.ca>  / Is anybody using VAX system in Toronto Canada ?iG Though I've been looking for a company using VAX in the Toronto area, Ie couldn't find any.9 I'm interested in pursuing a career back into VAX in VMS.c/ Please let me know if you have any suggestions.b It will be great thanks.   Jim,   ------------------------------  % Date: Thu, 06 Jul 2000 10:06:20 -0400s+ From: "Main, Kerry" <Kerry.Main@compaq.com>  Subject: RE: Vax in Canada ?J Message-ID: <910612C07BCAD1119AF40000F86AF0D8052844F3@kaoexc4.kao.dec.com>   Jim,  I Just curious, but do you mean you want to get back into VMS in general or  for some reason VAX VMS?  K Reason for asking is that VAX systems are now end of life (support for next@K 10 years though) and most folks starting new environments with VMS do so on. Alpha systems. :  J If you want used equipment, older Alpha systems that easily run the latestH versions of VMS are kicking around the grey market for fairly reasonable prices.    Regards,  
 Kerry Main Senior Consultant,
 Compaq Canada  Professional Servicesc Voice : 613-592-4660 FAX   : 819-772-7036 Email : kerry.main@compaq.com    -----Original Message-----% From: Jim Kim [mailto:sjkim@istar.ca] % Sent: Thursday, July 06, 2000 9:04 AM1 To: Info-VAX@Mvb.Saic.Com6 Subject: Vax in Canada ?    / Is anybody using VAX system in Toronto Canada ?dG Though I've been looking for a company using VAX in the Toronto area, Ip couldn't find any.9 I'm interested in pursuing a career back into VAX in VMS.// Please let me know if you have any suggestions.a It will be great thanks.   Jimi   ------------------------------  $ Date: Thu, 6 Jul 2000 11:31:11 -0400- From: "Peter Weaver" <peter.weaver@stelco.ca>  Subject: Re: Vax in Canada ?. Message-ID: <sm99hu7ogas13@corp.supernews.com>  A Another thread today mentioned www.dice.com. I checked it out anddF found two hits in the Toronto area for VMS. You can go to the advancedA search page, there you put in VMS and under the "Restrict by Area- Code" put 416 and 905.  8 There are other job search sites out there, (MONSTER.CA,D THINKPATH.COM, GLOBECAREERS.COM...) that I have not tried lately but$ they may have something in the area.   -- Peter Weaver   Jim Kim wrote in message/ <1h%85.22850$Ip.558681@cac1.rdr.news.psi.ca>...r0 >Is anybody using VAX system in Toronto Canada ?F >Though I've been looking for a company using VAX in the Toronto area, Ie >couldn't find any.m: >I'm interested in pursuing a career back into VAX in VMS.0 >Please let me know if you have any suggestions. >It will be great thanks.o >g >Jim >f >  >o >a   ------------------------------  % Date: Thu, 06 Jul 2000 11:12:15 -0700c/ From: twede <lnelsonNOlnSPAM@icius.com.invalid>r Subject: Re: VMS employment-9 Message-ID: <164c000c.dd384812@usw-ex0107-049.remarq.com>o  G As a matter of fact, my company is looking for a VMS person in SouthernuB California.  They need a system manager.  Anyone interested should6 contact Steve Michels at (626)305-7200, extension 124.    L * Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *G The fastest and easiest way to search and participate in Usenet - Free!-   ------------------------------  $ Date: Thu, 6 Jul 2000 08:19:45 +01005 From: "Chris Casey" <Chris.Casey@HendersonNOSPAM.com>  Subject: Re: We need VMS people ) Message-ID: <8k1bug$fum$1@soap.pipex.net>    Main, Kerry wrote in messageG >Current Customer I am working with is considering an Oracle on UNIX to0F >Oracle on VMS database migration as well. They are centralizing their OracleC >on OpenVMS applications and are planning a disaster tolerant, load0	 balanced,  >multi-site SAN config.  > F >Single site solutions are very hard to justify when doing large scale serverL >consolidation projects. Regional, district types typically say "centralize?I >But if we lose that datacenter, then I am out of nbusiness, so forget it  !". C >By implementing a multi-site, load balanced config, you remove the = >"yeahbuts" from those Regional and district folks arguments.H    G It would be very useful if some of the paperwork behind justifying that  decision could be published.> VMS just isn't even being considered for Oracle projects here.   ------------------------------  % Date: Thu, 06 Jul 2000 07:46:02 -0400:+ From: "Main, Kerry" <Kerry.Main@compaq.com>  Subject: RE: We need VMS people.J Message-ID: <910612C07BCAD1119AF40000F86AF0D8052844F2@kaoexc4.kao.dec.com>   Chris,  K >>> It would be very useful if some of the paperwork behind justifying that0  decision could be published. >>>  L Essestially, this Customer wants a 365x24x7 Oracle database environment thatG has high availability that INCLUDES scheduled downtime. It is currently7B running on a FC SAN with VMS V7.2-1 for both their development andF production environments.  They have a number of different Oracle basedL applications - some are home grown host based applications, but the majority are via SQL*net client-server.  J They want to acheive a state whereby the end users do not know what systemH in what datacenter is actually processing their requests. Load balanced,G taking advantage of all available systems. The HW is planned and scoped  appropriately.  H They want to acheive a state whereby there are no fail-overs for plannedI maint like HW upgrades, tuning reboots, OS upgrades etc. With the plannedRJ environment we are working towards, end users will not even be notified ofG occasional system scheduled downtimes, since the IT group will be using$H VMS's ability to migrate new users to an alternate system in the clusterL while maintaining current connections until they finish. When no users or DBG connections are are left on that system, they can simply shut it down. O  ? The concept is to separate SYSTEM availability from APPLICATIONPK availability. Business groups don't care if systems are rebooted as long as$K their application remains 100% available and no end users get disconnected.$  J The problem with fail-overs is that no matter how quickly they occur, theyF do impact the end users, therefore these outages need to be scheduled,H approved, messages sent to end users (or posted on system messages). ForJ many environments, the end users do not read these messages and simply getK upset with the IT group for disconnecting them at 02:00 AM and causing themM* to lose what they were doing at the time.   K Another Customer (different database vendor) I know of uses this capabilitytH to do planned maint on systems at 07:30AM (start of prime time, new userH migration started the evening before) to reduce their overtime costs andE provide a better lifestyle for their operations staff.. ok, ok, the $ # savings is the prime motivator :-)]n  A [main issue to be resolved is how to do database upgrades with no-F application availability impact, but that is a bridge to cross at some future point :-)]c  C With this planned environment, they will even be able to do planned H datacenter shutdowns with zero impact on application availability. SinceH there is one common system disk per site [HW RAID 0+1, and shared commonH sysuaf, queue files, startup files etc], rolling OS upgrades can be done+ with zero application availability impact. l  J [When a datacenter comes back online, they do realize their is some impactL on performance while the drives get re-synched, but they are also looking atK dark fibre multi-site connections between their FC switches for much higherT throughput on this aspect.]o  H This Customer is also the type that does not upgrade their back end moreL than once per year (12-18 months is their target) because of the huge amountJ of application testing it requires (not to mention the $'s involved). TheyL plan to leave the current App servers on UNIX/NT since that is where most of' the version upgrades etc usually occur.r  J Imho, the days of vendors getting away with stating high availability, butK NOT including all of their scheduled downtime is rapidly coming to a close. L Best exanmple is a typical mainframe that goes down multiple times per year,I but they do not include scheduled downtime in their availability numbers.-  F [Theoretical question - how do you schedule downtime in an Internet orH complex multi-vendor B2B environment and not disconnect any users at any time for planned maint?]  I This Customer is currently evaluating Oracle V8.1.6 on OpenVMS V7.2-1 fore# their environment later this year. c  J I suspect the recent Oracle announcements to include OpenVMS as a platformE for their new IAS strategy will also beneft this Customer. Reference: / <http://platforms.oracle.com/com/index_com.htm>.D <http://www.zdnet.com/eweek/stories/general/0,11011,2596660,00.html>D <http://www.compaq.com/inform/issues/issue31/features-184-b1-3.html>   Regards,  
 Kerry Main Senior Consultant,
 Compaq Canadae Professional Servicesn Voice : 613-592-4660 FAX   : 819-772-7036 Email : kerry.main@compaq.comu       -----Original Message-----: From: Chris Casey [mailto:Chris.Casey@HendersonNOSPAM.com]% Sent: Thursday, July 06, 2000 3:20 AM, To: Info-VAX@Mvb.Saic.Comt Subject: Re: We need VMS people,       Main, Kerry wrote in messageG >Current Customer I am working with is considering an Oracle on UNIX toeF >Oracle on VMS database migration as well. They are centralizing their OracleC >on OpenVMS applications and are planning a disaster tolerant, load 	 balanced,l >multi-site SAN config.R >TF >Single site solutions are very hard to justify when doing large scale serverL >consolidation projects. Regional, district types typically say "centralize?I >But if we lose that datacenter, then I am out of nbusiness, so forget it  !".iC >By implementing a multi-site, load balanced config, you remove the = >"yeahbuts" from those Regional and district folks arguments.t    G It would be very useful if some of the paperwork behind justifying that  decision could be published.> VMS just isn't even being considered for Oracle projects here.   ------------------------------   End of INFO-VAX 2000.375 ************************nitoring process can be on a different cluster node.7  --46 Tigʘ1ų*8^iz3
`&sոjv@ݟF w}dg8W,5	7&LI!r	qup9ΞB!:Kei*E6͌K ,3'`Ǐ#B"0oF,y7^ Ü/!Ӛ3O9~-mBʪ.#JQC-s{;pAq*W򸸊
>JTSt%O<9^0Վ~ۨrJE5S#r=yܾ낂Qݓɖ5eͼpxƦΙ=Sg҉T-]n4PcK"YfKVZXi`'Pfz>Tp}E1(C6F]/N?H[lT6Uc|n{칬<nw<3)84g=;Jyp}h^%d>]j;K<kr9d϶N@@5`Ywj)kGՁ.'pҢ9ٵb>^U22KeDlן'Ϣ89C6p|,8SXfv+[Z@
#>\}_
.KAY(}jvD3Pāﵭ}$B| ɽvbvM.uqK8t%'7ymgj-^2kOQGSSCX]ܟȡLmèxZ[Rx(\d" @]jPu8e 6ɾ[Βr%&P)Bj(*FiHDl6UzˡgWS&3PYyLum
Ы `IGl_%k

`uW-j@=iue
/UYp&-vjw9y߯Dxy-0"Mp'̫_<:#ZICI(.y 3u}