1 INFO-VAX	Wed, 23 Jun 2004	Volume 2004 : Issue 346       Contents:B $SYNCHRONIZE ... why isn't $STATUS the $STATUS from the batch job?F Re: $SYNCHRONIZE ... why isn't $STATUS the $STATUS from the batch job?O Re: $SYNCHRONIZE ... why isn't $STATUS the $STATUS from the batch job? job?job? ! Re: CAPTURE PID OF PARENT SUBMIT? ! Re: CAPTURE PID OF PARENT SUBMIT? ! Re: CAPTURE PID OF PARENT SUBMIT?  DEC Datatrieve Re: delete directory recursive Re: delete directory recursive Re: delete directory recursive Re: delete directory recursive Re: delete directory recursive Re: DS20 from Tru64 to VMS Re: DS20 from Tru64 to VMS@ Re: Is OpenVMS 'Adaptive' according to HP, or just an Albatross?@ Re: Looking for ideas: SUBMIT from DCL and get completion status@ Re: Looking for ideas: SUBMIT from DCL and get completion status' More LAT questions on LAT$SYSTARTUP.COM + Re: More LAT questions on LAT$SYSTARTUP.COM + Re: More LAT questions on LAT$SYSTARTUP.COM + Re: More LAT questions on LAT$SYSTARTUP.COM + Re: More LAT questions on LAT$SYSTARTUP.COM  Mozilla 1.7  Re: Mozilla 1.7  Re: NFS mounting Re: OT: Sun / Sparc  FUD Re: OT: Sun / Sparc  FUD Re: OT: Sun / Sparc  FUD Re: OT: Sun / Sparc  FUD: Re: RSA Secure ID  or other hardware token authentication?' Re: slap in the face again... thanks HP ' Re: slap in the face again... thanks HP ' Re: slap in the face again... thanks HP  Re: Sortwork file query $ Re: Split I/Os to contiguous file???$ Re: Split I/Os to contiguous file??? Re: SSH protocol in VMS 5.5-2  Re: Sun / Sparc  FUD VMS consultant for rent  Re: VMS consultant for rent & Re: VMS Gets Short Shrift at HP World.C Re: [OpenVMS] BACKUP to two tapes CONCURRENTLY (duplicating data) ?   F ----------------------------------------------------------------------  % Date: Tue, 22 Jun 2004 18:42:43 -0700  From: Z <z@no.spam> K Subject: $SYNCHRONIZE ... why isn't $STATUS the $STATUS from the batch job? 0 Message-ID: <10dhnv9td54v249@corp.supernews.com>  > Isn't $SYNCHRONIZE supposed to set $STATUS to the $STATUS fromA the batch job?  See below ...  why am I getting the error message A for exit status 124 (not 123) and why is $STATUS being OR-ed with  %x10000000 ?       $ type 15sec.com $ wait 00:00:15 
 $ exit 123   $ submit 15sec.com; Job 15SEC (queue SYS$BATCH, entry 607) started on SYS$BATCH    $ sync/entry=6070 Job 15SEC (queue SYS$BATCH, entry 607) completed, %SYSTEM-I-DEVNOTMOUNT, device is not mounted   $ write sys$output $STATUS
 %X1000007B   ------------------------------  % Date: Tue, 22 Jun 2004 21:30:27 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>O Subject: Re: $SYNCHRONIZE ... why isn't $STATUS the $STATUS from the batch job? + Message-ID: <40D8EB43.AA7AD77B@comcast.net>    Z wrote: > @ > Isn't $SYNCHRONIZE supposed to set $STATUS to the $STATUS fromC > the batch job?  See below ...  why am I getting the error message C > for exit status 124 (not 123) and why is $STATUS being OR-ed with  > %x10000000 ? >  > $ type 15sec.com > $ wait 00:00:15  > $ exit 123 >  > $ submit 15sec.com= > Job 15SEC (queue SYS$BATCH, entry 607) started on SYS$BATCH  >  > $ sync/entry=6072 > Job 15SEC (queue SYS$BATCH, entry 607) completed. > %SYSTEM-I-DEVNOTMOUNT, device is not mounted >  > $ write sys$output $STATUS > %X1000007B   DJAS01::DDACHTERA$ a = 123 DJAS01::DDACHTERA$ sh sym a /   A = 123   Hex = 0000007B  Octal = 00000000173   
 Get the idea?   G The "1" in the high-order digit just tells the system not to repeat the F error message (it wouldn't anyway for $SEVERITY = 3, "informational").   D.J.D.   ------------------------------  % Date: Tue, 22 Jun 2004 19:58:47 -0700  From: Z <z@no.spam> X Subject: Re: $SYNCHRONIZE ... why isn't $STATUS the $STATUS from the batch job? job?job?0 Message-ID: <10dhsdtf8o8qsa7@corp.supernews.com>   David J Dachtera wrote: @ >>Isn't $SYNCHRONIZE supposed to set $STATUS to the $STATUS fromC >>the batch job?  See below ...  why am I getting the error message C >>for exit status 124 (not 123) and why is $STATUS being OR-ed with  >>%x10000000 ? >> >>$ type 15sec.com >>$ wait 00:00:15  >>$ exit 123 >> >>$ submit 15sec.com= >>Job 15SEC (queue SYS$BATCH, entry 607) started on SYS$BATCH  >> >>$ sync/entry=6072 >>Job 15SEC (queue SYS$BATCH, entry 607) completed. >>%SYSTEM-I-DEVNOTMOUNT, device is not mounted >> >>$ write sys$output $STATUS >>%X1000007B >  >  > DJAS01::DDACHTERA$ a = 123 > DJAS01::DDACHTERA$ sh sym a 1 >   A = 123   Hex = 0000007B  Octal = 00000000173  >  > Get the idea?   9 I'm not sure what you're getting at ... yes, I understand , that 7Bh is 123.  But that's not my problem.     I have 2 problems:   1.8 The batch jobs exits with 123 but the $STATUS after SYNC9 returns isn't 123.  It's 123 OR-ed with %x10000000.  This 9 is not what I expected.  And that contradicts the OpenVMS  DCL Dictionary, N-Z:    From DCL II-490:  ...  Description: ... : The completion status form the SYCNHRONIZE command is that: same as the completion status of the last command executed in the job.      2.8 I'm getting an error message for status 124 or maybe 1227 (DEVNOTMOUNT) even though I exit with 123.  Why is that 
 hapenning?   ------------------------------  # Date: Tue, 22 Jun 2004 18:02:51 GMT - From: JF Mezei <jfmezei.spamnot@teksavvy.com> * Subject: Re: CAPTURE PID OF PARENT SUBMIT?@ Message-ID: <e8c52d077ab7be437bf19b51d3a88eb6@news.teranews.com>   Larry Kilgallen wrote:G > With current generally released versions of VMS, your only hope is to F > audit successful accesses to the queue and read the audit file data.  L Perhaps if the original poster explained why the submitted job needs to knowL the PID of the submittor and what it intends to do with it, we might be able< to find other solutions to achieve the needed functionality.   ------------------------------  # Date: Tue, 22 Jun 2004 19:25:01 GMT % From: "John Vottero" <John@mvpsi.com> * Subject: Re: CAPTURE PID OF PARENT SUBMIT?A Message-ID: <hI%Bc.27379$eH1.12813228@newssvr28.news.prodigy.com>   5 "John Brandon" <brandon@dalsemi.com> wrote in message + news:04062211194948@dscis6-0.dalsemi.com...  > No dice... > + > I am wanting to know this for batch jobs.  > H > Using the $GETJPI to report back the parent process PID only works for SPAWNed  > processes - not batch jobs.  > 	 > ABC.COM  > ! > $ submit nbc.com; /log /noprint  > $ show symbol $entry' > $ write sys$output f$getjpi("","PID")  > $ wait 00:00:10  > 	 > NBC.COM  > ' > $ write sys$output f$getjpi("","PID") . > $ write sys$output f$getjpi("","MASTER_PID") > G > The PID and MASTER_PID are one and the same and reflect that process.  >  >  > L > Passing the PID as a parameters to the command file will work however thisG > implies changes to the existing code.  Therefore this will not work -  since I 6 > have no desire to modify thousands of command files. >  >  > I > Is there anyway to change the SUBMIT CLI to parse out that information?  What > about using LGI? >   E The SUBMIT command returns the entry number in the DCL symbol $ENTRY.   J Use F$GETQUI("DISPLAY_ENTRY", "JOB_PID", $ENTRY) to get the PID of the job you submitted.  F Remember that there can be a delay between the submit and when the job( starts.  (There could even be a reboot).   And now for the shameless plug:   > Whatever it is you're doing might be easier if you used JAMS!! http://www.mvpsi.com   John Vottero   ------------------------------  % Date: Tue, 22 Jun 2004 21:20:03 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>* Subject: Re: CAPTURE PID OF PARENT SUBMIT?+ Message-ID: <40D8E8D2.95568051@comcast.net>    John Brandon wrote:  > 6 > Batch job ABC.COM submits a batch job called NBC.COM > M > Is there a way from with in NBC.COM to identify that ABC.COM was the parent  > submit process?   H I suppose there should be a way to setup an Audit event on a SUBMIT (ACEE on a specific queue's ACL?) and filter the audit info later on. Seems  the long way around though.   0 What's the root problem you are trying to solve?   D.J.D.   ------------------------------    Date: 22 Jun 2004 12:31:27 -0700# From: cbell@qch.on.ca (Johnny Hosp)  Subject: DEC Datatrieve = Message-ID: <e2db8f9b.0406221131.23bdaddf@posting.google.com>   F I am new to VMS and DTR. I am trying to change the default dictionary,E when I do a "SHOW DICTIONARY" I am at the defualt of _CDD$TOP. How do A I list the dictionaries available? Is there a command to list all  dictionaries below this? Thanks,  JH   ------------------------------  % Date: Tue, 22 Jun 2004 22:02:51 +0200 0 From: "Jerome" <jeromeatforissierdotorg@no.spam>' Subject: Re: delete directory recursive 4 Message-ID: <cba3a3$qkd$1@apollon.grec.isp.9tel.net>   Z wrote: > Larry Kilgallen wrote:A >> So if you cannot use a freely available DCL command procedure, D >> write your own.  The overhead of using DCL rather than a compiledA >> language is insignificant compared to the overhead of deleting 	 >> files.  > = > Yes, that's an option.  And, now, looking back, we probably  > should have done that. > ? > But that still leaves us with the unpleasant fact that VMS is ; > missing a feature that many consider standard for an o/s.   - Agreed. Hmm..., Guy P., are you listening? ;)    --   Jerome   ------------------------------  % Date: Tue, 22 Jun 2004 20:57:40 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>' Subject: Re: delete directory recursive + Message-ID: <40D8E394.AEB203C0@comcast.net>    "Craig A. Berry" wrote:  > H > Making no claim about what various sites will or won't allow, I merelyC > point out that sites with Perl installed have yet another option:  > A > $ perl -e "use File::Path; rmtree(\@ARGV,1,0);" dev:[dir] file1  > dev:[dir2] > [snip]    1. Perl does not ship with VMS   2. Perl is "freeware"3  3. Perl is not "supported" any more/less than DFU.   E So, the OP likely can't use it. (I know - there's no way to make that   "right", but whaddaya gonna do?)   D.J.D.   ------------------------------  % Date: Tue, 22 Jun 2004 21:01:49 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>' Subject: Re: delete directory recursive + Message-ID: <40D8E48D.E49351CD@comcast.net>   
 Jerome wrote:  > 
 > Z wrote: > > Larry Kilgallen wrote:C > >> So if you cannot use a freely available DCL command procedure, F > >> write your own.  The overhead of using DCL rather than a compiledC > >> language is insignificant compared to the overhead of deleting  > >> files.  > > ? > > Yes, that's an option.  And, now, looking back, we probably  > > should have done that. > > A > > But that still leaves us with the unpleasant fact that VMS is = > > missing a feature that many consider standard for an o/s.  > / > Agreed. Hmm..., Guy P., are you listening? ;)   B He may be, but ask yourself: as "unhackable" as VMS may be, do youC really want that kind of "gold" to lie "at the end of the rainbow", 7 tempting would-be hackers, crackers and script-kiddies?   E The lack of DELETE/RECURSE may be one of VMS's most valuable security 	 features!   A D'ya ever wonder why DFU is the way it is? (Uses SMG by default.)    D.J.D.   ------------------------------  % Date: Tue, 22 Jun 2004 21:06:10 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>' Subject: Re: delete directory recursive + Message-ID: <40D8E592.D2094F85@comcast.net>    Mike Rechtman wrote: > D > Well, if "supported" is more important than "efficient", how about& > $ Backup/delete [.top...]*.*;* NLA0:  G I think you need a saveset name and /SAVE in P2. That's the only way it ; works for me (I just tested BACKUP/LOG, not BACKUP/DELETE).   C Of course, you still need privilege and/or access ("permission") to  delete the directories.    D.J.D.   ------------------------------  # Date: Wed, 23 Jun 2004 05:07:18 GMT L From: winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr")' Subject: Re: delete directory recursive 6 Message-ID: <00A33C2D.BF53F8AE@SSRL.SLAC.STANFORD.EDU>  ` In article <40D8E394.AEB203C0@comcast.net>, David J Dachtera <djesys.nospam@comcast.net> writes: >"Craig A. Berry" wrote: >>  I >> Making no claim about what various sites will or won't allow, I merely D >> point out that sites with Perl installed have yet another option: >>  B >> $ perl -e "use File::Path; rmtree(\@ARGV,1,0);" dev:[dir] file1
 >> dev:[dir2] 	 >> [snip]  >   > 1. Perl does not ship with VMS > 2. Perl is "freeware" 4 > 3. Perl is not "supported" any more/less than DFU. > F >So, the OP likely can't use it. (I know - there's no way to make that! >"right", but whaddaya gonna do?)   J Actually, I think CSWS_PERL would be supported, the same way CSWS is - you can log service calls, etc.    -- Alan    --  O =============================================================================== 0  Alan Winston --- WINSTON@SSRL.SLAC.STANFORD.EDUM  Disclaimer: I speak only for myself, not SLAC or SSRL   Phone:  650/926-3056 M  Paper mail to: SSRL -- SLAC BIN 99, 2575 Sand Hill Rd, Menlo Park CA   94025 O ===============================================================================    ------------------------------  # Date: Wed, 23 Jun 2004 00:56:50 GMT 5 From: rdeininger@mindspringdot.com (Robert Deininger) # Subject: Re: DS20 from Tru64 to VMS L Message-ID: <rdeininger-2206042101500001@user-uinj4bp.dialup.mindspring.com>  2 In article <1087922666.730776@news.drenet.dnd.ca>,& claude.marinier@drdc-rddc.gc.ca wrote:  C >I will try booting again soon; at that time, I will note messages  C >related to SCSI devices. In the mean time, the 'SHOW CONFIG' PROM  F >command tells me that I have three NCR 53C895 SCSI controllers: PKA, G >PKB, and PKC; and two Adaptec AIC 7895 SCSI controllers: PKD and PKE.  A >These last two are on the same board as the Ethernet controller  J >(DE500-AA). The system also has two IDE controllers: DQA and DQB; the CD G >drive is DQA0. All SCSI devices are on PKA: the internal disk (saving  J >the copy of Tru64), the two external disks, and the external tape drive. I >I can boot from DKA200 but VMS only sees that one SCSI device. MultiNet  & >does not see the Ethernet controller.  I What does the SHOW DEVICE command display at the console?  You might post = that output, the disk names might be recognizable to someone.   J One simple way to make drives disappear is to have a SCSI ID conflict.  Or@ a bad cable.  Have you done all the standard SCSI bus debugging?  H Is this a new installation of VMS?  One can tell VMS to ignore devices.  (See SYSMAN IO SET EXCLUDE.)    J Multinet might not see the network controller if it has been configured to  look for a different controller.  D These problems might show up if you move an existing, customized VMS< system disk from one system to another without adjusting the customizations.   I >I would like to put the external devices on a different SCSI bus. I may  A >open the box to see if I can attach the external connector to a  ? >different controller. That would eliminate the possibility of  ' >interference between the SCSI devices.   J In general, if a SCSI controller has an internal connector and an externalA connector for the same SCSI bus, you are only supposed to use one G connector.  There may be exceptions, but I can't think of them offhand.      > J >I tried to apply the most recent VMS update patch in the hopes of fixing J >the driver problem. It fails with an undefined symbol (something related  >to post processing).   " You should post the exact message.  2 >I may have to re-install VMS so it discovers the G >actual devices; I do not know how VMS goes about the boot process and  A >device discovery & configuration but expected it to be dynamic.    J VMS discovers disks when it boots, or when you explicitly ask it to (usingF SYSMAN IO AUTOCONFIGURE).  You don't need to re-install VMS to make itB locate disks, assuming you have a good installation to begin with.  I The fact that VMS recognizes one disk on the SCSI controller, but not the E others, suggests the controller is probably not the problem.  I guess D there's something wrong with the disks (at least from VMS's point of< view), or the cabling is bad, or VMS has been misconfigured.     >I will D >start asking about purchasing the latest VMS (I think it is 7.3-2). > ! >Peter 'EPLAN' LANGSTOEGER wrote: A >> In article <40D7350A.5060706@drdc-rddc.gc.ca>, Claude Marinier ) <claude.marinier@drdc-rddc.gc.ca> writes:  >>  L >>>Could the SCSI controler be unsupported with VMS? All three disks are on 1 >>>the same controller (PKA) and I do see DKA200.  >>   >>  6 >> What does a SHOW CONFIG tell about the controller ? >>  ; >> Could it be that the disks aren't compatible (non-ultra) % >> with the controller (ultra SCSI) ?  >  >-- ( >Claude Marinier, Information Technology/ >Defence Research & Development Canada (Ottawa)   >claude.marinier@drdc-rddc.gc.ca" >http://www.ottawa.drdc-rddc.gc.ca   ------------------------------  # Date: Wed, 23 Jun 2004 03:07:55 GMT - From: JF Mezei <jfmezei.spamnot@teksavvy.com> # Subject: Re: DS20 from Tru64 to VMS @ Message-ID: <a2f904a978968004d3eacf91af10b9ea@news.teranews.com>   Robert Deininger wrote: L > One simple way to make drives disappear is to have a SCSI ID conflict.  OrB > a bad cable.  Have you done all the standard SCSI bus debugging?  M I believe that in the original post, the writer stated that the system worked R fine with TRu64, and it is only when he tried to load VMS that it stopped working.  M Assuming the above paragraph is correct, wouldn't the only possible answer be H a problem with VMS drivers not being compatible with whatever device was' loaded whereas the Tru64 drivers were ?   M Could it be a problem with the hardware not having the ROM software needed to N boot VMS (or would a DS20 have all pre-requisite ROM based console software to boot VMS ?)   K Also, not sure if I saw this, but starting at what version of VMS would the  DS20 be supported ?    ------------------------------  % Date: Tue, 22 Jun 2004 22:32:21 -0400 # From: "John Smith" <a@nonymous.com> I Subject: Re: Is OpenVMS 'Adaptive' according to HP, or just an Albatross? , Message-ID: <lbWdnTY8OLcqdkXdRVn-gg@igs.net>  > "Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message7 news:cf15391e.0406220752.7d4d0e04@posting.google.com... 0 > "John Smith" <a@nonymous.com> wrote in message( news:<jKudnZL2GbSIa03dRVn-ig@igs.net>...J > > I wonder how many times OpenVMS is mentioned and discussed extensively and $ > > favorably at this conference.... > F > Despite having the name "HP Software Forum", the focus at this forumE > is on HP OpenView and the software that interoperates with OpenView F > (one example of which is the OpenView Operations Agent for OpenVMS), > not software in general. > F > OpenVMS has many favorable characteristics that allow one to build aG > readily adaptive enterprise today, as many here can attest. (And Dave @ > Holt did a presentation at the joint ENSA@work/HP-Interex EMEA@ > conference in Munich last month entitled "OpenVMS delivers the > Adaptive Enterprise now.") > E > There are significant technical challenges being worked on today in D > the AE world to try to make Windows, Linux, etc. easier to manage,G > continuously available, secure, and able to readily scale up and down G > with changing workload demands. These are very hard problems to solve  > for platforms other than VMS.   H Since these problems are very hard to solve (read: expensive) on non-VMSJ platforms, wouldn't it make sense to take a lower cost approach and simply- advertise VMS as the solution to the problem?     G > Consider software installation: The shared system disk concept in VMS ? > allows one to avoid having to install an operating system and H > applications once for each and every server -- you can simply boot anyE > number of servers from a system disk which already has the required G > set of applications installed. And one can easily shift a server from H > one cluster to another as workload shifts, by simply rebooting it from > a different system disk. > = > Also consider data sharing: Because of the crude clustering < > capabilities on other platforms, there is only a subset ofH > applications which can readily scale up or down by addition or removalG > of nodes, and those are ones which deal basically with read-only data E > and don't have shared read/write data. (About the only exception to D > this is Oracle RAC, and we know where that technology comes from.)G > OpenVMS clusters allow full read/write access to data from any number " > of servers sharing the workload. > ! > I loved this article from Cnet: . > http://news.com.com/2102-1071_3-5070877.html > --- # > The futility of utility computing  > By Jon Oltsik  > H > Back in the 1980s, the folks at Digital Equipment had a problem. WhileA > their VAX systems were selling like hotcakes, the systems still 9 > couldn't deliver the kind of scale found on mainframes.  > F > Digital, which had no plans to build big iron, set out to find a wayF > to use existing equipment to increase capacity--a search that led toB > the development of the VAX cluster and the concept of horizontal
 > scaling. > G > Rather than roll in bigger and bigger boxes, Digital discovered a way E > to balance compute loads across multiple machines. What's more, the E > systems behaved as a single system image so IT administrators could E > manage the cluster as a distinct entity. This masked the underlying A > complexity, lowered cost, and simplified day-to-day activities.  > B > This history review provides a good analog for today's computingC > utility vision. HP, IBM,  Microsoft and Sun Microsystems all have H > major initiatives underway to ease complexity with their own computingH > utilities. (It is ironic that these same companies contributed greatlyF > toward building this complexity, but that's another story.) The factF > is that in spite of good intentions and armies of engineers, none ofB > them will achieve anywhere near the success of the venerable VAX
 > cluster.H ....... except the venerable Alpha-cluster, and the mixed mode VAX-AlphaI cluster, and the IA64/VMS cluster, the mixed-mode Alpha-IA64/VMS cluster, H and the working but not officially supported VAX-Alpha-IA64/VMS cluster.  " Time for the next OpenVMS ad......  G See if HP had a brain (or even a heart), it would have known about this C article in advance and would have had pop-up and OpenVMS banner ads . associated with this article on the Cnet site.   ------------------------------  # Date: Tue, 22 Jun 2004 19:27:37 GMT 1 From: "konabear" <konabearg-newsgroups@yahoo.com> I Subject: Re: Looking for ideas: SUBMIT from DCL and get completion status A Message-ID: <JK%Bc.27380$eH1.12813908@newssvr28.news.prodigy.com>   B Wait for the job(s) to complete and go after the Accountng process termination record.    Todd; "Alan E. Feldman" <spamsink2001@yahoo.com> wrote in message 6 news:b096a4ee.0406220636.c3fd4de@posting.google.com...  > Z <z@no.spam> wrote in message, news:<10df24ntm5rgu07@corp.supernews.com>...B > > I want to submit N batch jobs (all at once) from a DCL commandC > > procedure (the main program).  From the main program, I need to 0 > > know when each job finishes and its $STATUS. > > C > > $SYNCHRONIZE is unsuitable since I'll lose the $STATUS of job 2 2 > > if I'm waiting for job 1 when job 2 completes. > > A > > I've looked through f$getqui() and f$getjpi() and I don't see % > > any way to use either to do this.  > > @ > > The only solution I've come up with is to write a dispatcher? > > in C and do $SNDJBCs with completion ASTs.  That's a LOT of 	 > > work.  > >  > > Any better ideas?  >  > 6 > Uh, in my previous post, the SPAWN command is wrong! >  > Make that  >  > $ SPAWN/NOWAIT @WATCH 'ENTRY'    ------------------------------  % Date: Tue, 22 Jun 2004 21:16:42 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>I Subject: Re: Looking for ideas: SUBMIT from DCL and get completion status + Message-ID: <40D8E809.64EF9A7A@comcast.net>    Z wrote: > @ > I want to submit N batch jobs (all at once) from a DCL commandA > procedure (the main program).  From the main program, I need to . > know when each job finishes and its $STATUS. > A > $SYNCHRONIZE is unsuitable since I'll lose the $STATUS of job 2 0 > if I'm waiting for job 1 when job 2 completes. > ? > I've looked through f$getqui() and f$getjpi() and I don't see # > any way to use either to do this.  > > > The only solution I've come up with is to write a dispatcher= > in C and do $SNDJBCs with completion ASTs.  That's a LOT of  > work.  >  > Any better ideas?   H When I wrote backup procedures for a former site, I employed the list ofE entries idea followed by a "watcher" job that waits for each entry to : complete, then samples the completion status via F$GETQUI(G "DISPLAY_ENTRY", ...). The jobs were SUBMITted /RETAIN=ALWAYS. Then, it F would DELETE/ENTRY the jobs after collecting all the completion stati.   Just a suggestion...   D.J.D.   ------------------------------  % Date: Tue, 22 Jun 2004 16:25:28 -0500 % From: Neil Cherry <njc@wolfgang.uucp> 0 Subject: More LAT questions on LAT$SYSTARTUP.COM. Message-ID: <slrncdh8pb.gj0.njc@wolfgang.uucp>  D I'm doing some more poking around and I'd like to understand some of? the stuff I'm reading in the LAT startup file better. I see the  following 2 line:   $ $ LCP CREATE PORT LTA1: /APPLICATION0 $ LCP SET PORT LTA1: /NODE=server_1 /port=port_1  F The machine I'm editing this on is VAX (wow, original thinking ;-) andE the other device is a DEC Server called server_1. Does this mean that C I now can connect to port_1 of server? Does this get advertised via 8 LAT advertisements? Does this mean anything else? Thanks   --  D Linux Home Automation         Neil Cherry        ncherry@comcast.net; http://home.comcast.net/~ncherry/               (Text only) = http://linuxha.sourceforge.net/                 (SourceForge) 8 http://hcs.sourceforge.net/                     (HCS II)   ------------------------------  % Date: Tue, 22 Jun 2004 14:44:09 -0700 + From: "Barry Treahy, Jr." <Treahy@MMaz.com> 4 Subject: Re: More LAT questions on LAT$SYSTARTUP.COM' Message-ID: <40D8A829.2040907@MMaz.com>    Neil Cherry wrote:  E >I'm doing some more poking around and I'd like to understand some of @ >the stuff I'm reading in the LAT startup file better. I see the >following 2 line: > % >$ LCP CREATE PORT LTA1: /APPLICATION 1 >$ LCP SET PORT LTA1: /NODE=server_1 /port=port_1  > G >The machine I'm editing this on is VAX (wow, original thinking ;-) and F >the other device is a DEC Server called server_1. Does this mean thatD >I now can connect to port_1 of server? Does this get advertised via9 >LAT advertisements? Does this mean anything else? Thanks  >    > G Yes, the DECserver advertises the ports available so that from the VAX  I you can establish this application port.  These names are arbitrary, and  K can be changed.  You can also connect to the DECserver using NCP by typing:    NCP CONNECT NODE SERVER_1   E which will connect as a console to the DECserver, allowing you to do  B configuration there, but you must know the access password.  Just J remember you use CTRL-D to instruct NCP to disconnect when you are done...   Barry    --    > Barry Treahy, Jr                       E-mail: Treahy@MMaz.com> Midwest Microwave                          Phone: 480/314-1320> Vice President & CIO                         FAX: 480/661-7028                            ------------------------------  % Date: Tue, 22 Jun 2004 20:33:03 -0400 % From: "Chris" <mc.moore@sympatico.ca> 4 Subject: Re: More LAT questions on LAT$SYSTARTUP.COM: Message-ID: <3d4Cc.25957$Nz.1084377@news20.bellglobal.com>   Just an aside;  L I'm fairly sure that trying to connect from NCP requires that you define theG hardware (ethernet) address of the DECserver in the NCP database first.     6 "Barry Treahy, Jr." <Treahy@MMaz.com> wrote in message! news:40D8A829.2040907@MMaz.com...  > Neil Cherry wrote: > G > >I'm doing some more poking around and I'd like to understand some of B > >the stuff I'm reading in the LAT startup file better. I see the > >following 2 line: > > ' > >$ LCP CREATE PORT LTA1: /APPLICATION 3 > >$ LCP SET PORT LTA1: /NODE=server_1 /port=port_1  > > I > >The machine I'm editing this on is VAX (wow, original thinking ;-) and H > >the other device is a DEC Server called server_1. Does this mean thatF > >I now can connect to port_1 of server? Does this get advertised via; > >LAT advertisements? Does this mean anything else? Thanks  > >  > > H > Yes, the DECserver advertises the ports available so that from the VAXJ > you can establish this application port.  These names are arbitrary, andE > can be changed.  You can also connect to the DECserver using NCP by  typing:  >  > NCP CONNECT NODE SERVER_1  > F > which will connect as a console to the DECserver, allowing you to doC > configuration there, but you must know the access password.  Just L > remember you use CTRL-D to instruct NCP to disconnect when you are done... >  > Barry  >  > --   > @ > Barry Treahy, Jr                       E-mail: Treahy@MMaz.com@ > Midwest Microwave                          Phone: 480/314-1320@ > Vice President & CIO                         FAX: 480/661-7028 >  >  >    ------------------------------  % Date: Tue, 22 Jun 2004 21:23:52 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>4 Subject: Re: More LAT questions on LAT$SYSTARTUP.COM+ Message-ID: <40D8E9B8.260B18B8@comcast.net>    Chris wrote: >  > Just an aside; > N > I'm fairly sure that trying to connect from NCP requires that you define theI > hardware (ethernet) address of the DECserver in the NCP database first.     CONNECT NODE requires that, yes.  * CONNECT VIA mumble PHYS ADDR ... does not.   D.J.D.   ------------------------------  % Date: Tue, 22 Jun 2004 22:02:32 -0700 + From: "Barry Treahy, Jr." <Treahy@MMaz.com> 4 Subject: Re: More LAT questions on LAT$SYSTARTUP.COM' Message-ID: <40D90EE8.3020207@MMaz.com>    Chris wrote:   >Just an aside;  > M >I'm fairly sure that trying to connect from NCP requires that you define the H >hardware (ethernet) address of the DECserver in the NCP database first. >    > G Yes, I believe you are correct but I was also presuming that this same  F VAX is the load host for MOM, which if it is, then the MAC's would be 	 loaded...    Barry    --    > Barry Treahy, Jr                       E-mail: Treahy@MMaz.com> Midwest Microwave                          Phone: 480/314-1320> Vice President & CIO                         FAX: 480/661-7028                            ------------------------------  % Date: Wed, 23 Jun 2004 06:54:08 +0930 * From: Mark Daniel <mark.daniel@vsm.com.au> Subject: Mozilla 1.7/ Message-ID: <40d8a382$1@duster.adelaide.on.net>   G Thanks to VMS Engineering for making Mozilla 1.7 available so promptly.   '    http://www.mozilla.org/releases/#1.7   F +--------------------------------------------------------------------+E   Mark Daniel                         http://wasd.vsm.com.au/adelaide F   mailto:Mark.Daniel@wasd.vsm.com.au (Mark.Daniel@dsto.defence.gov.au);   A pox on the houses of all SPAMers.  Make that two poxes. F +--------------------------------------------------------------------+   ------------------------------  % Date: Wed, 23 Jun 2004 06:52:11 +0200  From: Dirk Munk <munk@home.nl> Subject: Re: Mozilla 1.72 Message-ID: <cbb29r$thr$1@news2.tilbu1.nb.home.nl>   Mark Daniel wrote:I > Thanks to VMS Engineering for making Mozilla 1.7 available so promptly.  > ( >   http://www.mozilla.org/releases/#1.7 > H > +--------------------------------------------------------------------+F >  Mark Daniel                         http://wasd.vsm.com.au/adelaideG >  mailto:Mark.Daniel@wasd.vsm.com.au (Mark.Daniel@dsto.defence.gov.au) < >  A pox on the houses of all SPAMers.  Make that two poxes.H > +--------------------------------------------------------------------+N I second that. I'm very pleasantly surprised by this very fast release of 1.7! Thanks very much guys!   ------------------------------  % Date: Tue, 22 Jun 2004 14:53:00 -0400 2 From: "Stanley F. Quayle" <squayle@insight.rr.com> Subject: Re: NFS mounting . Message-ID: <40D847CC.2254.104FCD64@localhost>   > In my OpenVMS nfs client: 7 > $ tcpip mount dnfs98: /host=servernfs / path="/mynfs"   B You might want to add /ADF=CREATE.  This allows the VMS system to ? create special files on the NFS server which hold the VMS file  ; characteristics.  I do this when mounting Linux NFS shares.   
 --Stan Quayle  Quayle Consulting Inc.  
 ----------- Stanley F. Quayle, P.E. N8SQ  +1 614-868-136343 8572 North Spring Ct., Pickerington, OH  43147  USAS0 stan-at-stanq-dot-com       http://www.stanq.com   ------------------------------    Date: 22 Jun 2004 21:05:43 -0600+ From: young_r@encompasserve.org (Rob Young) ! Subject: Re: OT: Sun / Sparc  FUDI3 Message-ID: <A4hKOteOZ+7G@eisner.encompasserve.org>T  p In article <d5abdc5c04db282b037f56ec73425413@news.teranews.com>, JF Mezei <jfmezei.spamnot@teksavvy.com> writes:\ >> http://news.com.com/Fujitsu+revs+faster+server+chip/2100-1006_3-5242884.html?tag=nefd.top > H > Fujitsu unveils new SPARC chip at 1.89GHz with 3MB on-chip cache. 90nmP > process. Will appear on servers later this year, sold by both Sun and Fujitsu.L > Sun continues to have some Sparcs built by Texas Instruments (making Sparc? > available from multiple sources, something IA64 can't claim).e > K > Seems to me that such news kill any/all FUD that some have been spreadingU5 > about the lack of future of the Sparc architecture.e  G 	A year ago, there were 88 SPARC machines on the SuperComputer Top 500. 7 	Today there are 3.  SPARC has slid off to irrelevance.    				Rob    ------------------------------  % Date: Tue, 22 Jun 2004 22:19:13 -0400t# From: "John Smith" <a@nonymous.com>e! Subject: Re: OT: Sun / Sparc  FUDH, Message-ID: <6oednWtuVL0-dUXd4p2dnA@igs.net>  8 "Rob Young" <young_r@encompasserve.org> wrote in message- news:A4hKOteOZ+7G@eisner.encompasserve.org...eK > In article <d5abdc5c04db282b037f56ec73425413@news.teranews.com>, JF Mezei & <jfmezei.spamnot@teksavvy.com> writes: > >>L http://news.com.com/Fujitsu+revs+faster+server+chip/2100-1006_3-5242884.html
 ?tag=nefd.tope > >CJ > > Fujitsu unveils new SPARC chip at 1.89GHz with 3MB on-chip cache. 90nmI > > process. Will appear on servers later this year, sold by both Sun andm Fujitsu.H > > Sun continues to have some Sparcs built by Texas Instruments (making SparcbA > > available from multiple sources, something IA64 can't claim).H > >EC > > Seems to me that such news kill any/all FUD that some have beeno	 spreadinga7 > > about the lack of future of the Sparc architecture.1 > H > A year ago, there were 88 SPARC machines on the SuperComputer Top 500.8 > Today there are 3.  SPARC has slid off to irrelevance.    J Compared to the number and breadth of 3rd party applications that are soldH and run each day on Sparc vs. those sold & run each day on VMS-supportedH platforms, I'd say you were full of shit Rob if I was so inclined to use$ language other than the Queen's own.   ------------------------------    Date: 22 Jun 2004 23:29:42 -0600+ From: young_r@encompasserve.org (Rob Young)'! Subject: Re: OT: Sun / Sparc  FUDh3 Message-ID: <qIWLn4di8f0a@eisner.encompasserve.org>   R In article <6oednWtuVL0-dUXd4p2dnA@igs.net>, "John Smith" <a@nonymous.com> writes: > : > "Rob Young" <young_r@encompasserve.org> wrote in message/ > news:A4hKOteOZ+7G@eisner.encompasserve.org...TL >> In article <d5abdc5c04db282b037f56ec73425413@news.teranews.com>, JF Mezei( > <jfmezei.spamnot@teksavvy.com> writes: >> >> N > http://news.com.com/Fujitsu+revs+faster+server+chip/2100-1006_3-5242884.html > ?tag=nefd.tope >> >K >> > Fujitsu unveils new SPARC chip at 1.89GHz with 3MB on-chip cache. 90nm J >> > process. Will appear on servers later this year, sold by both Sun and
 > Fujitsu.I >> > Sun continues to have some Sparcs built by Texas Instruments (making  > Sparc B >> > available from multiple sources, something IA64 can't claim). >> >D >> > Seems to me that such news kill any/all FUD that some have been > spreading.8 >> > about the lack of future of the Sparc architecture. >>I >> A year ago, there were 88 SPARC machines on the SuperComputer Top 500.'9 >> Today there are 3.  SPARC has slid off to irrelevance.- >  > L > Compared to the number and breadth of 3rd party applications that are soldJ > and run each day on Sparc vs. those sold & run each day on VMS-supportedJ > platforms, I'd say you were full of shit Rob if I was so inclined to use& > language other than the Queen's own. >   G 	But that is just one data point.  Sure sun is still alive and doing a  ? 	few billion but have been trending downward badly for the lastd? 	3-4 years while everyone else is going in the other direction.T   May 26, 2004  L In figures released by research firm Gartner (IT: news, chart, profile), IBMM (IBM: news, chart, profile) topped the worldwide first-quarter server revenueTL figures with $3.6 billion, a 16.7 percent rise from $3.1 billion a year ago.O H-P (HPQ: news, chart, profile) was second with $3 billion, up 6.2 percent fromh $2.9 billion last year.m  L Rounding out the top five revenue leaders were Sun Microsystems (SUNW: news,K chart, profile), with $1.2 billion, down 12.5 percent from last year's $1.4 L billion; Dell (DELL: news, chart, profile), whose sales jumped 25 percent toK $1.2 billion from $938 million; and Fujitsu/FSC, with $842 million, up fromo
 $786 million.I    ; 	There are signs all around that the good ship SPARC struckm! 	an iceberg and is listing badly.   > 	SPARC is a weak CPU compared to the competition resulting in:    @ 		o  3 machines left in the Top500  (That has to be embarassing)3 		o  abandonment of TPC-C benches, smashed in TPC-H:$ 		o  abandonment of SPEC CPU benches1 		o  Chased away by x86 in the workstation marketA5 		o  Quarter after quarter (11 or 12 straight now) of8 		   declining revenue    A 	BUT as you point out "has an installed base of software."  Yep - F 	same story crowed by Digital in the mid to late 80s when faced by SunC 	Sparcstations that were faster/cheaper and eroding Digital market.d  D 	Face it - SPARC is the VAX CPU of the 2000s - mostly irrelevant at  	this point in time.  @ 	Must be terribly annoying for Scotty to realize he is following9 	in Ken Olsen's footsteps, watching his empire crumble touB 	faster and/or cheaper machines.  And I'm sure it gets worse when D 	Power5 gets cranking and Itanium cranks and approaches Xeon pricing 	parity.  You betcha!.     				Robc    B Men with walkie-talkie                  I'm home again to you babeC Men with flashlights waving             You know it makes me wondergG Up upon the tower                       Sittin' in the quiet slipstreame> The clock reads daylight savin'         Rollin' in the thunder  -                                 -- Neil Youngc   ------------------------------  # Date: Wed, 23 Jun 2004 05:20:42 GMT(- From: JF Mezei <jfmezei.spamnot@teksavvy.com> ! Subject: Re: OT: Sun / Sparc  FUD"@ Message-ID: <b695cacb9b15dde5e343e06de7161a9e@news.teranews.com>   Rob Young wrote:O >         But that is just one data point.  Sure sun is still alive and doing asH >         few billion but have been trending downward badly for the lastH >         3-4 years while everyone else is going in the other direction.  L HP has also gone down in many metrics, especially when you consider what the> combined HP/Compa/Digital/Tandem markets that could have been.  L But it is true that Sun has fallen hard from the high of the .COM. But Sun's1 highs were much higher than HP/Compaq/Digital's. J  H Compaq benefited from Y2K since so many wintel boxes had to be replaced.H Digital had zero benefit from the whole .COM boom since it was busy self> destructing itself so it could be purchased by smaller Compaq.  N Sun may have gone too high for its size, but that .COM excess has also allowedM Sun to make major inroads in banks where it had no presence before. It is now2> a respected name in banks that used to be big blue-only shops.  J One should be looking at installed base of Solaris. Is it growing of going down ? nM Sun is also transitioning from a commodity producer to a services company nowdA that it has "serious" customers such as banks who demand support.t  G Compare this to VMS which has been in a long decline for over a decade.   H Linux will hurt Solaris just as much as HP-UX and AIX. But out of the 3,L Solaris has the biggestr reputation as "the" standard Unix.  And since HP isM not interested in pushing VMS, and since its Wintel business doesn't generate-K profits, HP is just as dependant on HP-UX as Sun is depandant on Solaris inoL terms of enterprise systems. So HP is just as endangered by Linux as Sun is.  G >         SPARC is a weak CPU compared to the competition resulting in:uP >                 o  3 machines left in the Top500  (That has to be embarassing)  M This can be misleading. Consider the big Apple cluster that had ranked #2 (oraK was it #3?). It is not listed because it is in the process of being rebuilt,N with new faster POWER G5 machines, and some military department is building anK even bigger APPLE cluster which will rank way up there. So next year, APPLErS should have "infinite growth" by going from 0 to 2 machines listed at the very top.e  L Sparc is getting faster wity the new Fujitsu chip. So next year, it may have1 some new supercomputers and increase in rankings.i  M However, not placing in the list isn't necessarily a problem on the financialoG aspect. It is nice publicity, but if you sell a large numebr of smaller J machines that are not even close to ranking, it might still be better than0 selling a small number of machines that do rank.  I And another question: when was the last time a VMS system was included in- those rankings ?  J >         BUT as you point out "has an installed base of software."  Yep -O >         same story crowed by Digital in the mid to late 80s when faced by Sun9L >         Sparcstations that were faster/cheaper and eroding Digital market.  L Difference being that Sun is not busy actively destroying its installed base, and software as Palmer was so keen on doing.    J >         faster and/or cheaper machines.  And I'm sure it gets worse whenM >         Power5 gets cranking and Itanium cranks and approaches Xeon pricinge >         parity.  You betcha!  K I am not sure that Power or Sparc are really affraid of IA64's performance.aL They may fear Intel's business tactics that may force customers to go to theN inferior IA64, but I doubt anyone really fear IA64 and its EPIC compilers willI be able to keep up with the other chips. The 8086 may in fact be a biggerD+ threath due to the installed software base.w   ------------------------------  % Date: Tue, 22 Jun 2004 15:53:03 -0700s0 From: Mark Berryman <Mark.Berryman@Mvb.Saic.Com>C Subject: Re: RSA Secure ID  or other hardware token authentication? % Message-ID: <40d855e0@cpns1.saic.com>o   Hal Kuff wrote:lM > Anyone out there using Secure ID or another vendor's key FOB hardeare token ! > type authentication on OpenVMS?t >  > H Yes, I use SecurID tokens to log into my VMS systems.  Support for this  is built into Multinet.m  
 Mark Berrymand   ------------------------------    Date: 22 Jun 2004 18:22:13 -0700( From: bob@instantwhip.com (Bob Ceculski)0 Subject: Re: slap in the face again... thanks HP= Message-ID: <d7791aa1.0406221722.193e6ad9@posting.google.com>d  X VAXman-  @SendSpamHere.ORG wrote in message news:<00A33BCA.6247AC22@SendSpamHere.ORG>... > 9 > Still, there's no reason for making it WEENDOZE only.  U > --  D > http://www.legacy-2000.com  for the *best* OpenVMS system securityE >                             solutions that others only claim to be.e    @ why not windoze only ... it is fine as a client, but that is it!   ------------------------------  # Date: Wed, 23 Jun 2004 02:25:12 GMT " From:   VAXman-  @SendSpamHere.ORG0 Subject: Re: slap in the face again... thanks HP0 Message-ID: <00A33C30.40285846@SendSpamHere.ORG>  h In article <d7791aa1.0406221722.193e6ad9@posting.google.com>, bob@instantwhip.com (Bob Ceculski) writes:Y >VAXman-  @SendSpamHere.ORG wrote in message news:<00A33BCA.6247AC22@SendSpamHere.ORG>...d >> o: >> Still, there's no reason for making it WEENDOZE only.   >> -- E >> http://www.legacy-2000.com  for the *best* OpenVMS system securitysF >>                             solutions that others only claim to be. >s >-A >why not windoze only ... it is fine as a client, but that is it!3  C If you want people to take you seriously about your pro-VMS stance,wC you'd not make such stupid statements in public.  Weendoze *sucks*!  -- uB http://www.legacy-2000.com  for the *best* OpenVMS system securityC                             solutions that others only claim to be.e -- hK VAXman- A Bored Certified VMS Kernel Mode Hacker   VAXman(at)TMESIS(dot)COMP            s5   "Well my son, life is like a beanstalk, isn't it?" t   ------------------------------  % Date: Tue, 22 Jun 2004 23:36:35 -0600(" From: GreyCloud <mist@cumulus.com>0 Subject: Re: slap in the face again... thanks HP0 Message-ID: <p_6dnWKLpMIOi0TdRVn-tw@bresnan.com>   Bob Ceculski wrote:   Z > VAXman-  @SendSpamHere.ORG wrote in message news:<00A33BCA.6247AC22@SendSpamHere.ORG>... > 9 >>Still, there's no reason for making it WEENDOZE only.  A >>-- OD >>http://www.legacy-2000.com  for the *best* OpenVMS system securityE >>                            solutions that others only claim to be.d >  >  > B > why not windoze only ... it is fine as a client, but that is it!   You're joking... right??   ------------------------------  # Date: Tue, 22 Jun 2004 20:56:55 GMT / From: "Duane Smith" <duane.smith@hp-nospam.com>g  Subject: Re: Sortwork file query2 Message-ID: <r21Cc.4507$M01.1392@news.cpqcorp.net>  K Prior to the DEC C Runtime Library using a LRL value of 32767, it will look.K for a logical name of DECC$DEFAULT_LRL.  If the logical name is found *and* J the translation is a numeric value between 0 and 32767, that value is usedL as the LRL value on all stream files whose LRL value previously defaulted toH 32767.  If the logical name is not found or is defined outside the valid$ range, then the value 32767 is used.   Duane Smithn# Former member of the DEC C RTL teamo  7 "Paul Sture" <nospam@sture.homeip.net> wrote in messages$ news:2hvthvFgtrvhU1@uni-berlin.de... > Dirk Munk wrote: > > Paul Sture wrote:  > >( > >> Dirk Munk wrote:T > >> > >>> Adrian Birkett wrote:t > >>>r
 > >>>> Hi, > >>>>E > >>>> One of out applications sorts a file. The logical names of the J > >>>> sortwork files are defined as DSA1:[NODIR]. This directory does notJ > >>>> exist on the device and the files are showing in SDA as DSA1:[]. IsJ > >>>> there any way of viewing these files to get an idea of how big theyJ > >>>> are as the seem to be eating disk space at  quite an alarming rate? > >>>>J > >>> You don't tell us what you are sorting. If these are files that wereD > >>> made with a C program, the please keep in mind that C producesJ > >>> Stream-lf files with a fake record size that is the maximum that RMS > >>> supports (32 kB).b > >>>gI > >>> You may have to add a record-size switch with the sort command, whes$ > >>> you know the true record size. > >>>z > >>
 > >> Dirk, > >>H > >> I had the same idea, and in case you missed it, Adrian replied with this:s > >>C > >> "No, the program is written in COBOL and is run under the CODAaJ > >> environment. The file in question has 4 keys in 5 areas which doesn't > >> really help"  > >> > > Hi Paul, > >a > > I indeed missed his reply. > > J > > But here is something I don't understand. Am I wrong in the assumptionH > > that he is sorting a indexed file ?? The output may be sequential of > > course.p > >h >o% > I think your assumption is correct.  >fK > > As an ex-COBOL programmer I know that there are two ways to use sort inaJ > > COBOL. For input you can tell COBOL to sort a file, or you can write aF > > procedure that outputs records to sort. The same system applies toK > > output as well. The sort output can be returned as a file, or it can betK > > read record by record in a output procedure for further processing. I'm J > > sure choosing the right method can mean a lot for the size of the sort > > workfiles. > >  >mG > The COBOL input and output sort procedures are a very useful feature,.J > and your mentioning them reminds me that the output file could be a fullD > blown report with sub-totals, grand totals etc, rather than just aE > sequential file. Changing such a program could be a major task, and E > since CODA is a large financial system, we don't know if Adrian hasI: > access to the sources to see what processing is involved >i   ------------------------------    Date: 22 Jun 2004 14:36:57 -0600 From: gleason@encompasserve.org - Subject: Re: Split I/Os to contiguous file???.3 Message-ID: <Mw1K8qTNJz0U@eisner.encompasserve.org>s  c In article <lfi9ilBcHVdw@eisner.encompasserve.org>, Kilgallen@SpamCop.net (Larry Kilgallen) writes:>W > In article <A5Fui+qw0Fu6@eisner.encompasserve.org>, gleason@encompasserve.org writes: f >> In article <CPJwD4mW0Yds@eisner.encompasserve.org>, Kilgallen@SpamCop.net (Larry Kilgallen) writes:h >>> In article <bdc65a53.0406210326.1de2666c@posting.google.com>, gspamtackett@yahoo.com (Galen) writes: >>> J >>>> The UCB$L_MAXBCNT for this and the other SCSI disks that I checked isA >>>> indeed 20000 hex which corresponds to a limit of 256 blocks.s >>> J >>>> Now the problem is to determine whether the split I/Os are really the- >>>> big limiting factor on write throughput.i >>> F >>> Try cutting the UCB$L_MAXBCNT value in half and seeing if there is >>> an effect. >>> 8 >>> CMKRNL is your friend (at least on test systems :-). >>  A >>   Heck, I was thinking he oughtta double the value, and see ifmA >> it's just pro forma, or if it is a real limitation...on a test  >> system, of course.. > G > But his root goal is to affect performance, not the academic exerciser& > of learning if this is a real limit.      H   Well...if it's an artificial limitation, increasing it has a middlin'  chance of making it faster...t   Lee K. Gleason N5ZMR Control-G Consutlantst   ------------------------------    Date: 22 Jun 2004 16:09:09 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen) - Subject: Re: Split I/Os to contiguous file??? 3 Message-ID: <+oFkE3FUVjmP@eisner.encompasserve.org>-  U In article <Mw1K8qTNJz0U@eisner.encompasserve.org>, gleason@encompasserve.org writes:me > In article <lfi9ilBcHVdw@eisner.encompasserve.org>, Kilgallen@SpamCop.net (Larry Kilgallen) writes:cX >> In article <A5Fui+qw0Fu6@eisner.encompasserve.org>, gleason@encompasserve.org writes:g >>> In article <CPJwD4mW0Yds@eisner.encompasserve.org>, Kilgallen@SpamCop.net (Larry Kilgallen) writes: i >>>> In article <bdc65a53.0406210326.1de2666c@posting.google.com>, gspamtackett@yahoo.com (Galen) writes:t >>>> yK >>>>> The UCB$L_MAXBCNT for this and the other SCSI disks that I checked isHB >>>>> indeed 20000 hex which corresponds to a limit of 256 blocks. >>>> tK >>>>> Now the problem is to determine whether the split I/Os are really thee. >>>>> big limiting factor on write throughput. >>>> eG >>>> Try cutting the UCB$L_MAXBCNT value in half and seeing if there isb >>>> an effect.  >>>> e9 >>>> CMKRNL is your friend (at least on test systems :-).o >>> B >>>   Heck, I was thinking he oughtta double the value, and see ifB >>> it's just pro forma, or if it is a real limitation...on a test >>> system, of course. >> rH >> But his root goal is to affect performance, not the academic exercise' >> of learning if this is a real limit.s >  >  > J >   Well...if it's an artificial limitation, increasing it has a middlin'  > chance of making it faster...s  D But if it has no effect, it would be more difficult to determine theC real situation, compared to decreasing it (which is unlikely to hitn other limits).   ------------------------------  # Date: Tue, 22 Jun 2004 19:05:13 GMTe# From: hoff@hp.nospam (Hoff Hoffman)l& Subject: Re: SSH protocol in VMS 5.5-21 Message-ID: <Jp%Bc.4496$yO.2978@news.cpqcorp.net>s  e In article <slrncdgnd2.t9.thierry@MARS.Family>, Thierry Dussuet <thierry@squeeeez.dyndns.org> writes:-> :On 2004-06-22, Shiva MahaDeva <contracer11@uol.com.br> wrote:0 :> Is there any free SSH protocol to VMS 5.5-2 ?2 :> I'd like connect to Solaris box from VMS box... : N :http://www.free.lp.se/fish/ seems to have something - Don't know about BAMSE, :but FISH might work for you.l  E   The OpenVMS Frequently Asked Questions has pointers to some of the sB   available SSH implementations.  (The SSH V2 available in TCP/IP D   Services V5.4 and later is not yet mentioned in the FAQ, but it isA   also roughly a decade too recent for your OpenVMS VAX version.)'    N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq N  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com_   ------------------------------  % Date: Tue, 22 Jun 2004 17:31:05 -0400i* From: "Bill Todd" <billtodd@metrocast.net> Subject: Re: Sun / Sparc  FUDW2 Message-ID: <kcidnf6kPMx6OUXd4p2dnA@metrocast.net>  : "JF Mezei" <jfmezei.spamnot@teksavvy.com> wrote in message: news:d5abdc5c04db282b037f56ec73425413@news.teranews.com... > >gL http://news.com.com/Fujitsu+revs+faster+server+chip/2100-1006_3-5242884.html
 ?tag=nefd.top  > H > Fujitsu unveils new SPARC chip at 1.89GHz with 3MB on-chip cache. 90nmG > process. Will appear on servers later this year, sold by both Sun ando Fujitsu.L > Sun continues to have some Sparcs built by Texas Instruments (making Sparc? > available from multiple sources, something IA64 can't claim).l > K > Seems to me that such news kill any/all FUD that some have been spreadingp5 > about the lack of future of the Sparc architecture.i  ; Well, there are some pretty dedicated FUDsters out there...y  H Still, the 1174/1345 base/peak SPECint scores Fujutsu just posted exceedL those of the fastest Itanic sporting an equal-sized 3 MB cache (1140/1178 atK 1.6 GHz) - and indeed the 6 MB Itanic SPECint scores can't match it on peak F and only beat it slightly on base, save for those running the trick HPK compiler under HP-UX.  Its large (up to 32 processor) systems scale SPECintmK similarly competitively with Itanic systems.  And even its SPECfp scores of.J 1510/1803 are quite good (comparable to the best Opterons, better than the4 best P4/Xeons, but Itanic still leads in this area).  J Moving to possibly more commercially-relevant areas, Fujitsu's new SPARC64L in its 32-processor configuration handily beats the best 32-processor ItanicK and POWER4+ 32-processor submissions in jbb2000 (the Itanic system uses thehL top-of-the-line 1.5 GHz 6 MB chips; the POWER4+ the 1.7 GHz chips which have@ recently been superseded by 1.9 GHz chips in IBM's TPC-C entry).  L All in all, pretty damn respectable - considering the amount of abuse peopleD have been inclined to dump on SPARC lately.  They don't seem to haveJ submitted a system for SPECweb99_SSL, but I just can't resist pointing outJ that the new 4-processor Opteron systems have opened up a significant leadK over the top-of-the-line Itanic and POWER4+ competition since the last times
 I checked.   - bill   ------------------------------  % Date: Tue, 22 Jun 2004 14:41:09 -070043 From: Alan Frisbie <Usenet01REMOVE@Flying-Disk.com>t  Subject: VMS consultant for rent. Message-ID: <40D8A775.8080606@Flying-Disk.com>  ? I check ebay daily for VMS-related stuff, but this is the first B time I have seen someone selling themself.   (Actually, it is just? their services.)   I hope it works for them.   I would probably & do the same if I were in need of work.      item 5706036555"    VAX ALPHA VMS Consulting 1 Hour	    $60.00e   It is in category:    Computers & Networkingc    > Networking "      > Mainframe, DEC, VAX, AS/400$        > DEC, Digital Equipment Corp   Alan   ------------------------------  % Date: Tue, 22 Jun 2004 21:25:22 -0500p2 From: David J Dachtera <djesys.nospam@comcast.net>$ Subject: Re: VMS consultant for rent+ Message-ID: <40D8EA12.56EA0C26@comcast.net>,   Alan Frisbie wrote:0 > A > I check ebay daily for VMS-related stuff, but this is the firstLD > time I have seen someone selling themself.   (Actually, it is justA > their services.)   I hope it works for them.   I would probablyu( > do the same if I were in need of work. >  >    item 5706036555$ >    VAX ALPHA VMS Consulting 1 Hour >    $60.00  >  > It is in category: >    Computers & NetworkingS >    > Networkingt$ >      > Mainframe, DEC, VAX, AS/400& >        > DEC, Digital Equipment Corp  E Well, I guess that's one way to determine the value of a VMS person's  time by the hour.a  > Perhaps it could be a useful bargaining chip in some contexts.   D.J.D.   ------------------------------  % Date: Tue, 22 Jun 2004 20:55:31 -0500a2 From: David J Dachtera <djesys.nospam@comcast.net>/ Subject: Re: VMS Gets Short Shrift at HP World.m+ Message-ID: <40D8E313.ADC1A56A@comcast.net>a   Andy Bustamante wrote: > K > David, I believe most of the regulars here would have no problem with theDH > certification exams.  I took the ASE exams (OpenVMS/Alpha hardware) toL > comply with Compaq's re-sellers standards.  After making time to look overJ > the self study exams over a few evenings, I walked in and passed the VMSL > exam.  The Alpha hardware exam I took cold.  Talking with some of the examH > writers at HP, the goal is that anyone with the experience behind themH > should be able to pass these certifcations.  One of them claims he can: > predict your score within 5 points based on 5 questions. > M > The deep discounts for Encompass/HPWorld/CETS have mad the difference in myE$ > attending over the last few years.  F I passed the OpenVMS ASE exam once a few years back. The certificationF itself is of little or no value in the job market due to VMS's lack ofD mind-share - you're certified on something no one knows/cares about; what's the point?n  H The fee for the test is usually $100 or so. Not a bank-breaker, but thatE still leaves me negotiating with "the administration" to blow a wholehF week's pay on a symposium that has next to no VMS content. Tough sell, to say the very least.  H The good news is that we are an "empty nest" couple. We just married off; her daughter and I have no children of my own aside from myrD step-daughter and "step-son-in-law". If we had a young family, well,@ guess what the outcome of that negotiation would most likely be?  H ...and before anyone mentions it, no - the pay-back on the tax deduction1 for professional expenses doesn't balance it out.h   D.J.D.   ------------------------------  # Date: Tue, 22 Jun 2004 22:43:38 GMTA1 From: "Ken Randell" <kenneth.randell@verizon.net>rL Subject: Re: [OpenVMS] BACKUP to two tapes CONCURRENTLY (duplicating data) ?8 Message-ID: <uC2Cc.22024$Yb1.21912@nwrddc02.gnilink.net>  4 > But OTOH it is not supported by vms engineering... >1 > Thanks anywayt >  >   G I think he's after a 'bundled' solution (i.e., no extra cost).  I don't C think he meant to insult 3rd-party ISVs of which there are too few.r   ------------------------------   End of INFO-VAX 2004.346 ************************