0 INFO-VAX	Tue, 23 Jan 2001	Volume 2001 : Issue 45      Contents: Re: $ sh net: no DECnet IV :-( Re: $setimr problem  'Privacy' and technologyP Re: AlphaServer ES40 EV68/833 MHz (was: Re: Compaq continues the Digital TraditiP Re: AlphaServer ES40 EV68/833 MHz (was: Re: Compaq continues the DigitalTraditio4 Re: Compaq: A simple, affordable clustering solution4 Re: Compaq: A simple, affordable clustering solution4 Re: Compaq: A simple, affordable clustering solution4 Re: Compaq: A simple, affordable clustering solution4 Re: Compaq: A simple, affordable clustering solution Re: dec 3000 300 crashing ! Re: digital PrintServer 17 600 ps  Re: DS20 - Slow I/O  Re: DS20 - Slow I/O  Re: DS20 - Slow I/O  Re: DS20 - Slow I/O  Re: DS20 - Slow I/O & Re: Dual ethernet config for failover?& Re: Dual ethernet config for failover? ftp  Re: ftp  Re: GS160 hardware question ( How to delete directory with bad blocks?, Re: How to delete directory with bad blocks?, Re: How to delete directory with bad blocks?, Re: How to delete directory with bad blocks?, Re: How to delete directory with bad blocks?, Re: How to delete directory with bad blocks?, Re: How to delete directory with bad blocks?, Re: How to delete directory with bad blocks?( Imagine 30GB in your Laptop or Multia...2 Is Compaq dropping support for vms on vax platform6 Re: Is Compaq dropping support for vms on vax platform	 Re: KZQSA $ Re: Linux worm and RedHat 7.0 broken$ Re: Linux worm and RedHat 7.0 broken$ Re: Linux worm and RedHat 7.0 broken$ Re: Linux worm and RedHat 7.0 broken$ Re: Linux worm and RedHat 7.0 broken1 Re: Looking for Backup Scheduler for OpenVMS 7.21 % Re: Lowercase filenames on VMS V7.2-1  Re: Multiple path devices...! Re: New OpenVMS Times now on line ! Re: New OpenVMS Times now on line ! Re: New OpenVMS Times now on line ! Re: New OpenVMS Times now on line ! Re: New OpenVMS Times now on line ! Re: New OpenVMS Times now on line ! Re: New OpenVMS Times now on line ! Re: New OpenVMS Times now on line ! Re: New OpenVMS Times now on line 4 Re: No technical computing, was: expanding the niche4 Re: No technical computing, was: expanding the niche4 Re: No technical computing, was: expanding the niche4 Re: No technical computing, was: expanding the niche Re: Oldtimer forgot! Printing problems in TCPIP Re: Printing problems in TCPIP! Re: Queue has lowercase attribute ! Re: Queue has lowercase attribute ! Re: Queue has lowercase attribute ! Re: Queue has lowercase attribute ( Re: RA230+ (KZPAC-CA) and 18/36 GB DisksN Re: Rehash of Andrew's Java on OpenVMS FUD.  (was: Re: GS160 hardwarequestion)N Re: Rehash of Andrew's Java on OpenVMS FUD.  (was: Re: GS160 hardwarequestion) Re: Remote boot via DECnet?  Re: samba 2.0.3 : NT client L Re: Script to read directory of all drives, clustered, shared, etc. at once?L Re: Script to read directory of all drives, clustered, shared, etc. at once?K Re: Script to read directory of all drives, clustered, shared,etc. at once? 3 Re: Seeking OpenVMS docs (was Re: Oldtimer forgot!) 3 Re: Seeking OpenVMS docs (was Re: Oldtimer forgot!) ' Re: Source for used StorageWorks disks? 9 Re: To Hoff Hoffman: Any news about the VMS DHCP client ? ) Urgent:Error recovering files from backup - Re: Urgent:Error recovering files from backup  Re: Vax 7830 or 7730 Re: WRQ drops LAT ???  X-windows paradigm question  Re: X-windows paradigm question  Re: X-windows paradigm question  Re: X-windows paradigm question 6 [Q]: Resetting privileged password on a DECserver 200?: Re: [Q]: Resetting privileged password on a DECserver 200?: Re: [Q]: Resetting privileged password on a DECserver 200?  F ----------------------------------------------------------------------   Date: 22 Jan 2001 11:32:17 PSTT From: Fairfield@SLAC.Stanford.EDU (Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515)' Subject: Re: $ sh net: no DECnet IV :-( 3 Message-ID: <93kNIrqFBrKM@mccdev.slac.stanford.edu>   G In article <94cnov$uk4$1@nnrp1.deja.com>, Didier.Morandi@gmx.fr writes:  [...] 8 > Well, I always thought that to start DECnet 4 a simpleK > @sys$manager:startnet command was inuf in the systartup_vms.com file (no, E > I don't use SYSMAN STARTUP mumblefratz, I don't like this fantasy).  > K > Well (again) no, it's not inuf, there is around a sys$net_services.com in / > sys$startup: which does the registration. ...   H         Yes, a change in going to  VMS  7.0 or 7.1 (I don't recall whichH     since we skipped 7.0).  I _think_ this was documented in the ReleaseH     Notes,  but in any case, the relavent section of code appears in theH     SYSTARTUP_VMS.TEMPLATE  file  laid  down  during  the  installation/H     upgrade.  It pays to do a good deal of reading _before_ upgrading...           -Ken --  M  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:      650-926-3515N  -----------------------------------------------------------------------------B  These opinions are mine, not SLAC's, Stanford's, nor the DOE's...   ------------------------------  % Date: Mon, 22 Jan 2001 20:52:39 +0000 + From: "antonio.carlini" <arcarlini@iee.org>  Subject: Re: $setimr problem' Message-ID: <3A6C9D97.2D9AA7CA@iee.org>    cstranslations@msn.com wrote:   0 > The code sequence in question runs as follows: > B > 3862 result = sys$qio( , chan by value, io$_readvblk by value, &B > 3863                  qiosb, MailBoxAST, chan, claim,          &B > 3864                  len(claim::ClaimOverlay) by value, , , , )% > 3865 if (result <> ss$_normal) then ( > 3866    call lib$stop(result by value)
 > 3867 end if  > 3868! > 3869 FifteenMinutes = space$(8) 6 > 3870 call sys$bintim("0 :15", FifteenMinutes by ref)  4 "0 00:15:00" might be more obvious, the maintenance & engineers will thank you one day ...    4 I cannot remember, but I assume that BASIC passes in4 the "0 :15" by descriptor? If it's by ref then maybe4 things are going badly wrong here and FifteenMinutes is set to random rubbish.   , You check every other result status, why not0 the one from $bintim ... it's always the routine' that cannot possibly go wrong that does  (remember Murphy ...)   E > 3872 result = sys$setimr( , FifteenMinutes by ref, CloseDropsChan,& 1 > 3873                     DropsTimer by value, )   5 No EFN ... I don't know whether BASIC will just leave 0 random crud on the stack here or whether it puts' a zero in nicely for you. If it leaves  5 random stuff you are probably setting who-knows-whose , event flag ... not good. IIRC the version of- OpenVMS you have allows you to use EFN$C_ENF.   2 You also leave the final arg unspecified. Again if$ BASIC fills it in with a zero then    % > 3874 if (result <> ss$_normal) then ( > 3875    call lib$stop(result by value)
 > 3876 end if      > E > This is code for a process that runs as a detached process. The wee D > hours of this morning it decided to fall over and die on line 3875 > (with ss$_bufferovr).  > I > I'm guessing the offending argument was the wake time (FifteenMinutes). D > This kinda sorta doesn't seem to make much sense. It's a read onlyI > argument and it's passed by reference. I could (possibly) see an accvio 6 > if FifteenMinutes was a string of less than 8 bytes.  2 buffer overflow isn't listed as an explicit return5 so I assume that it is coming from some other service 9 that $setimr is calling. (Or someone missed a possibility  when writing the docs).   ) I'd make sure that you explicitly specify ) both the event-flag and flags arguments.    0 I'd also dump the binary value of FifteenMinutes/ and check that it looks right (the debugger can + do this with something like EXAMINE/DATE).    - I would also add a check on the return status  of the call to $bintim.   / If you cannot easily debug the program in-situ, / try putitng the code snippet above into a short . basic program and running it through there ...' if the fault is reproducible you should  catch it easily enough.    --     --------------- - Antonio Carlini             arcarlini@iee.org    ------------------------------    Date: 23 Jan 2001 02:34:06 +0800, From: Paul Repacholi <prep@prep.synonet.com>! Subject: 'Privacy' and technology 0 Message-ID: <87n1cjcu75.fsf_-_@prep.synonet.com>  ) John Stott <jpstott@src.wisc.edu> writes:   - > WILLIAM WEBB <WWEBB1@email.usps.gov> wrote:  > C > >Speaking of stickers, remember the Pentium PSN privacy brouhaha?  > > ? > >I've got a "Big Brother Inside" Intel logo spoof in my cube.  > >In color, no less.  > G > Meanwhile the NYT Patent column shows IBM just being granted a patent G > for having a network readable ID number in a vital component of every 1 > PC as an anti-theft/post-theft recovery device!   C May I recoment everyone goes to slashdot and reads the John Gilmore  article.   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------  # Date: Tue, 23 Jan 2001 00:14:22 GMT 2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)Y Subject: Re: AlphaServer ES40 EV68/833 MHz (was: Re: Compaq continues the Digital Traditi 5 Message-ID: <y14b6.37$cu.444@gazette.loc1.tandem.com>   -   Somebody somewhere (possibly me? :-) wrote:   H :Well the EV68/833 Mhz version of the ES40 is now "officially" available5 :(it's on the DBL price file and in the QuickSpec)...   A   AlphaServer platform update information is available via email.    To sign up, visit:  1     http://www.compaq.com/alphaserver/fb_acu.html   E   The AlphaServer ES40 and AlphaStation EV68 (833 MHz) platforms are  >   listed in the January 17th 2001 AlphaServer news webpage at:  6      http://www.compaq.com/alphaserver/news/index.html  D   I'll pass along a request that the main AlphaServer ES40 pages be &   checked and updated to reflect this.  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Tue, 23 Jan 2001 05:53:09 GMT  From: Dirk Munk <munk@home.nl>Y Subject: Re: AlphaServer ES40 EV68/833 MHz (was: Re: Compaq continues the DigitalTraditio ' Message-ID: <3A6D1C44.2008C990@home.nl>   P The quickspecs don't mention the 15000 rpm disks as an option, although they are
 available.   regards,   Dirk   Hoff Hoffman wrote:   / >   Somebody somewhere (possibly me? :-) wrote:  > J > :Well the EV68/833 Mhz version of the ES40 is now "officially" available7 > :(it's on the DBL price file and in the QuickSpec)...  > C >   AlphaServer platform update information is available via email.  >   To sign up, visit: > 3 >     http://www.compaq.com/alphaserver/fb_acu.html  > F >   The AlphaServer ES40 and AlphaStation EV68 (833 MHz) platforms are@ >   listed in the January 17th 2001 AlphaServer news webpage at: > 8 >      http://www.compaq.com/alphaserver/news/index.html > E >   I'll pass along a request that the main AlphaServer ES40 pages be ( >   checked and updated to reflect this. > P >  --------------------------- pure personal opinion ---------------------------N >    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Mon, 22 Jan 2001 21:47:23 +0000 ) From: Christof Brass <brass@infopuls.com> = Subject: Re: Compaq: A simple, affordable clustering solution , Message-ID: <3A6CAA6B.D0DE1C50@infopuls.com>  * fabio_compaq@ep-bc.petrobras.com.br wrote:L > But to the mainframe concept return to the corporations, I will insist:  ac > new graphical terminal must be developed. The web in the way it exists today (client-server) will Y > not have a long life because of the growing of applications and limitations of speed. I _ > believe in this new graphical terminal like an ICA Citrix client. Just the images come to the Y > costumers and all the high processing must be in the server - if the image comes to the S > costumers means it is safe to the costumer .... If Microsoft have sucess with WNT < > Datacenter and Terminal Server together, it is a  way .... > 	 > Regards  >  > FC >   9 Why do you ask for this proprietary M$ stuff like Windows = Terminal Server? You will get a fat client which is obviously > not the solution of the future. Stated that, what's wrong with; X11? As far as I know the other solutions (Windows Terminal ! Server, ICA Citrix, WinFrame) are 7 re-invent-the-wheel-approaches. Do we really need that?    ------------------------------  % Date: Mon, 22 Jan 2001 17:43:02 -0500 - From: JF Mezei <jfmezei.spamnot@videotron.ca> = Subject: Re: Compaq: A simple, affordable clustering solution , Message-ID: <3A6CB76E.BC760E91@videotron.ca>   "Main, Kerry" wrote:J > Wow - pretty poor fud - even for you Andrew. The Eurex was a NEW win fo= r a < > NEW project with the CBOT. As was the Sydney Exchange win.  J Ok, MR Main... I have heard that EUREX was dropping VMS, and now you say = that it was a new win.   J Does this mean that once you drop VMS, if you re-use your hardware for a = new < smaller less critical project, you consider this a new win ?  J Lets say I am the Cr=E9dit Lyonnais. I quietly drop VMS from my core busi= ness, J and then buy a small DS10 that runs VMS to run the reservation system for=  theJ corporate dining rooms. Would Compaq be so desperate for any wins that it=  J would be willing to use a client which abandonned VMS as its core system = to brag about a small new system ?   < Did EUREX drop or not VMS from some of its core operations ?  D > I view winning as gaining a Customers confidence that what you areJ > delivering is a solid solution that will meet their business requiremen= ts.   J Spinnning a win requires some confidence in the vendor. When customers do=  not6 have confidence in the vendor, then we see it as spin.  J For instance, in Canada, Air Canada regained its near monopoly status tha= nks J to government stupidity. During the olympics, it plastered TVs with const= ant J ads with the slogan "in competition with the world". They almost had to p= ull J the ads when they realised that people distrusted AC so much that they sa= w J these ads as AC trying to brainwash people that there was in fact competi= tion
 in Canada.    D Compaq brags about VMS being a major platform in international fundsJ transfers, even though SWIFT has pulled the plug on VMS and the last syst= ems J running the SWIFT software on VMS  in banks will go away in a year or two= =2E IfJ you're in that business, and see the Compaq claim, doesn't Compaq lose so= me
 credibility ?   J Similarly, if you are involved with EUREX and know that EUREX did pull th= e J plug on VMS, and you see Compaq brtagging about VMS there even though VMS=  now, plays a much lesser role, how do you react ?  J Since the VMS community is small and fairly tighlty knit, I think that Co= mpaq) should be very careful with its spinning.   J I agree that it is good and necessary to provide good news. But I feel th= atJ Compaq should be looking to provide TRUE good news instead of trying to f= ind a   positive way to spin a net loss.  6 > Quote - "Tru64 Unix and OpenVMS are proven winners."  J In all fairness: if one distrusts Gartner because Gartner doesn't like VM= S,J does this mean that Gartner all of a sudden becomes a beleivable source o= f J information because they publish ONE article that agrees with what you li= ke ?  J I think that the Gartner report only matches the changes between Palmer a= ndJ Compaq. During the Palmer years, Gartner was rightly calling for customer= s toJ move away from VMS ASAP. In hindsight, they proedicted correctly that Dig= italJ didn't really have any intentions with VMS and that customers should avoi= d it.   J And now that the Palmer poison has stopped being injected in VMS, you see=  J Gartner stopping the "VMS is dead" and starts to see VMS out of the immed= iate death list.   J If VMS is still alive despite all the poison it was fed, it is because of=  one single reason: 	EXTREMELY LOYAL CUSTOMERS.   J The remaining VMS shops are those who have managers and workers very loya= l toJ VMS and who were able to fight the strong forces to migrate away from VMS= =2E And J I think that Compaq should take a lesson or two from these very loyal peo= plePJ because of all the people in the world, they know what is needed to get V= MS back in shape.  J And when the folks who saved VMS from extinction tell Compaq they want to=  seeJ VMS marketed openily as a sign that Compaq is serious about VMS, then Com= paqe should listen very carefully.n  H The customer base for VMS is in a way, far more religious and loyal thanH APPLE's. Nobody at Apple tried to kill the macintosh, they just had someJ incompetent chiefs during a bad period. The loyal customers just had to k= eep J the MAC on life support during that period. The loyal VMS supporters had = toJ fight active attempts by Digital to kill VMS on top of trying to keep it = on
 life support.f    J > Its a multi-platform world and there is no one solution or one OS platf= ormb: > that will address all requirements, so whats your point?  J The point is that if the other platforms are competing against you, you s= houldk( be able/allowed to compete against them.  J > If increasing your footprint is Sun's main measurement of success, how = doJ > you explain all the server consolidation projects that are going on in = manyJ > medium-large companies ie. the concept of replacing many smaller server= s J > with fewer (but larger), centralized and in many cases, clustered serve= rs ??R  H Yep, FIDO/Microcell just replaces a farm of NT web servers with some SunJ servers running apache because they realised that their cheap PC solution=  H couldn't scale. So you should acknowledge that if server cosolidation isJ happening, then the most popular vendor is bound to get the most contract= s foreC consolidation. There is a lot more happening outside the VMS world.i   ------------------------------  % Date: Mon, 22 Jan 2001 17:46:10 -0500 - From: JF Mezei <jfmezei.spamnot@videotron.ca> = Subject: Re: Compaq: A simple, affordable clustering solution , Message-ID: <3A6CB829.E443271F@videotron.ca>  ! steven.reece@quintiles.com wrote:nM > Success (IMNSHO) has to be measured in whether the client is actually happycJ > with the solution provided and whether it fulfils the criteria that were > put forward for the project.  N In my opinion, success is whether a client is actually adding new applicationsL to a box or not. If the client adds new application to a different platform,J it means that the first platform is not so strategic anymore and while theL existing applocations are maintained, the company won't be betting a greater, proportion of its business on that platform.  N Success is when a company bring in VMS for a small project, sees its value andM decides to move applications from other platforms to VMS, even is there is no  sale of additional hardware.   ------------------------------  # Date: Tue, 23 Jan 2001 01:32:03 GMTt4 From: "Terry C. Shannon" <terryshannon@mediaone.net>= Subject: Re: Compaq: A simple, affordable clustering solutione< Message-ID: <na5b6.11665$t3.2230722@typhoon.ne.mediaone.net>  . "Jon Morgan" <jon@cmkrnl.net> wrote in message$ news:3a6514ef.5728577@130.133.1.4... > E > Although the pitching of the advertising against UNIX and NT is not A > prevalant, I was lucky to be present at one of Compaq's DiamondeE > Forum's at New York's World Trade Center. In that not only did RichiG > Marcello give a convincing arguement about the future of OpenVMS, buteF > made encouraging noises about those noises being heard (and felt) by > Michael Capallas.  >sE > (In fact, I was even more lucky to have lunch with Rich at the sameED > event and he seemed to portray that he really cared about what wasF > going on -- not only to customers like myself, but also to ISVs that > were in attendance.) >dG > Was all of this just PR? Maybe, maybe not. I was very impressed about C > the way that he basically asked Terry (Shannon) to discuss thingscB > without (apparantly) any restrictions (I think he gave Terry tenD > minutes notice that he was going to have to speak -- Terry, please( > correct me if I remember incorrectly).  K Your memory is flawless. I innocently wandered into the Forum to see what's K new in VMS-land and the next thing I knew I was asked to say a few words. I I did in fact have less than ten minutes notice, which of course mercifully A spared the audience from yet another PowerPoint presentation. ;-}l  F And there were no restrictions. I really don't think CPQ PR would haveJ approved my digs at "a clueless prior regime that tried to force VMS usersI to take up the Windows burden." But the audience seemed to appreciate theo observation!  L In any event, the Forum was well worth attending. I took about a dozen pagesJ of notes, and heard some mighty good things about VMS, Rdb, Alpha futures, etc.   cheers,2   terry sl   ------------------------------  # Date: Tue, 23 Jan 2001 01:35:50 GMT|4 From: "Terry C. Shannon" <terryshannon@mediaone.net>= Subject: Re: Compaq: A simple, affordable clustering solutioni< Message-ID: <Wd5b6.11669$t3.2232987@typhoon.ne.mediaone.net>  - "Paul Sture" <paul@sture.ch> wrote in message0% news:VA.0000023a.1e5a3a19@sture.ch...iC > In article <3A681118.C718367E@uk.sun.com>, Andrew harrison wrote:3 <snip> >AH > What's wrong Andrew? Not enough Enterprise IT Architecting to keep you > busy?  >s  L Or perhaps the Cachegate issue has been resolved! Seriously, though, I doubtJ there's any lack of architecting in Sun Country. I note with interest thatK Sun in the last quarter sold ~743 StarFires. Which is pretty close to CPQ'sr" WildFire shipments for all of Y2K.  J Goes to show you what a huge apps portfolio and superior marketing can do!   ------------------------------  % Date: Tue, 23 Jan 2001 01:35:48 -050002 From: rdeininger@mindspring.com (Robert Deininger)" Subject: Re: dec 3000 300 crashingL Message-ID: <rdeininger-2301010135490001@user-2ivea1s.dialup.mindspring.com>  l In article <6Rx86.1771$FC1.73637@ozemail.com.au>, "Antony Wardle" <antony.wardle@noospammm.met.co.nz> wrote:   > test mem addrv > STS-MEM-OK > test mem eccA > T_ERR_MEM       - ldl_l bcache miss no victim ad Addr: 00200264o > Uncorrected error count = 1m	 > 84 FAILy >  > T mem eccr > T-ERR-MEM TNF ECC  >  > test mem init, > T_STS_MEM ok >  > test mem cell  >  > T-STS-MEM ok >  > test mem bctpN >T-ERR-MEM - TNF BCTPe >   ) Well, some more about your memory errors.k   We decided to rejuvenate a DEC 3000 300 that's been gathering dust.  It only had 48 MB of RAM, so we obtained an extra 64 MB.  Today the RAM arrived and I went to install it.   "Just to make sure", I typed TEST MEM at the console before adding the new RAM.  It failed, sort of like your machine is failing.  Not quite reproducible, but not quite random.   The machine passes the power-up tests fine.  But the console starts with DIAG_SECTION set to 0, and after the initial test, it changes DIAG_SECTION to 1. So when you type TEST or TEST MEM, you get a more extensive test.)   Despite what the manual says, and what I posted previously, T MEM BCTP and T MEM ECC don't appear to exist on the model 300.  I get "T-ERR-MEM - TNF BCTP" just like you.  I think TNF really does mean "test not found".a   If you change DIAG_SECTION to 2, you get still more extensive tests, and a little more information.  TEST MEM ? shows you how to set the starting and ending address of a memory test.  I found that T MEM LLSC always fails on this machine, though not always in the same way, and at a different address each time.  Since the test stops after the first error, I had to try different starting addresses to test all the memory SIMMs. (There's supposed to be a flag to make the test continue past an error, but I didn't  a I didn't have a way to log the console on this machine, so I just jotted down some of the errors:-$ stl_c bcache miss with victim at ... stl_c bcache hit at addr .../ Ldl_l/stl_c atomic (can read my writing) at ...t
 and always' ? T-ERR-MEM - FAILED status = 20 at ...    At this point I was suspicious, so I shut down our long-running 3000 300, and started trying the same tests on it.  Same result!  The T MEM LLSC test always fails!i   I conclude that the model 300 does not correctly implement whatever LLSC is trying to check.  I tried a 3000 600, and T MEM LLSC seems to work fine.   The rejuvenated 3000 300, with it's new memory added, booted VMS 6.2 just fine, as in the past. (I know, it's old.  The rejuvenation isn't complete yet.)o  My errors don't exactly match what you posted, but they're in the ballpark.  I think in your first post, you said your machine had tried to boot.  Maybe it really is an OS problem, not a memory/cache problem.  Did you ever hook up a CD and boot from a known good system disk?   = Does your machine give errors over the whole range of memory?rP Does it sometimes (always) pass the power-up tests, before failing the mem test?   I hope you didn't toss it yet.    d Once again, I wish someone could take a peek in the service manual and tell us what's going on here.   -- t Robert Deininger rdeininger@mindspring.com-   ------------------------------  % Date: Mon, 22 Jan 2001 21:16:32 -0600c7 From: "David J. Dachtera" <djesys.nospam@earthlink.net>8* Subject: Re: digital PrintServer 17 600 ps- Message-ID: <3A6CF790.AFFFE92B@earthlink.net>t   Eric Ebinger wrote:  >  > -----Original Message-----7 > From: David J. Dachtera <djesys.nospam@earthlink.net>0* > >Well, let's start with usual questions: > >  > >Make/model of printer?S > >.( > Uh, isn't the subject specific enough? > . > You might try the documentation available at@ > http://www5.compaq.com/products/printers/support/prt_docs.html  9 Sorry - my mistake. I was thinking LN17 instead of LPS17.   G I dunno if there was ever a version of the LPS17 software that providedtE TCP/IP support. I was kinda hoping Paul Anderson would chime in righte
 about here...m   -- h David J. Dachterab dba DJE Systems  http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/e  F This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings is to be expected.  @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.i   ------------------------------    Date: 22 Jan 2001 13:59:46 -0500* From: young_r@eisner.decus.org (Rob Young) Subject: Re: DS20 - Slow I/O+ Message-ID: <BJK1Rtzzk13w@eisner.decus.org>   N In article <94hui0$vsn$1@nnrp1.deja.com>, fooguy <jweisen@my-deja.com> writes:J > My only comment about that is: that doesn't explain the poor performance? > under NetBSD. As we all know, Unix does not care about recordnJ > attributes, a stream is a stream is a stream. Seeing similar performanceI > under different OSes tells me it's a hardware issue. Maybe it's time toA# > put NT on here and see how it is.a >   : 	Poor or expected?   Still copying same disk to same disk?- 	You have disk heads bouncing back and forth.r   			Rob   ------------------------------  # Date: Mon, 22 Jan 2001 18:35:44 GMTp" From: fooguy <jweisen@my-deja.com> Subject: Re: DS20 - Slow I/O) Message-ID: <94hui0$vsn$1@nnrp1.deja.com>-  H My only comment about that is: that doesn't explain the poor performance= under NetBSD. As we all know, Unix does not care about record-H attributes, a stream is a stream is a stream. Seeing similar performanceG under different OSes tells me it's a hardware issue. Maybe it's time too! put NT on here and see how it is.i    + In article <hlR2lSkNyZQf@eisner.decus.org>,b-   young_r@eisner.decus.org (Rob Young) wrote: F > In article <pXzL6RROqHCY@eisner.decus.org>, young_r@eisner.decus.org (Rob Young) writes:n >e7 > 	Received the following in email from someone that isl: > 	very familiar with internals and not affilited with VMS > 	engineering:  >e > --- Begin Quote ---i > F > Yes, fast_io is a major win. Internally, it avoids most of the setup and H > teardown time of buffer allocation, avoids an extra trip thru IPL 4 onF > every $qio, and avoids several buffer checks. John Hallyburton is to beH > commended for that effort. (Unfortunately he's no longer with VMS...heB > was another one who left due to the ambiance of the Palmer era.) > G > The fast_io will work with ANY disk...it's all generic stuff..and wasiH > discussed in outline (though without the label) as a sensible followonH > to $qio as far back as the non-disclosure talks about Alpha before itsH > introduction back in 1992. It was recognized even then that the bufferD > probing and other PALcode calls needed for $qio would produce much more5 > relative degradation on Alpha than they had on Vax.a > F > $copy generally copies by blocks, then resets file attributes at theD > end, but fastio is just, at low level, a helluva lot faster way to move > data.  >v > --- End Quote ---e >r	 > 				RobW >n > >  > > 	From your original post:O > > G > > "My test was copying several large files (Oracle dumps - each abouts6 > > 200MB) from one directory on the disk to another." > > = > > 	If you really have a need to speed this up , I will makea# > > 	two suggestions (if possible):a > >j2 > > 	1)  Copy from one physical volume to another. > >r > > 	2)  Use fast_io_copyd > >M@ > > 	Here is an example copy of a 781 MByte (1600000) block file! > > 	using COPY and FAST_IO_COPY:o > >t > >o > > 	Vanilla Copy results  > >q > > Accounting information: E > >  Buffered I/O count:        54  Peak working set size:       1504sE > >  Direct I/O count:       50123  Peak virtual size:         166768oE > >  Page faults:              393  Mounted volumes:                0n" > >  Images activated:           4- > >  Elapsed CPU time:          0 00:00:17.55 - > >  Connect time:              0 00:16:24.88p > >n: > >>>>>	16:24 , 50123 I/Os.  From one SCSI disk to another > > 	off a SWXCR.c > >e > > 	fast_io_copy results: > > A > > Using 16 buffers, with 127 blocks (65024 bytes) per transfer.o > >y( > > Reading DRIVE1:[DIRECTORY]FILE.DAT;1( > > Writing DRIVE2:[DIRECTORY]FILE.DAT;1 > >oG > >  ELAPSED:    0 00:08:55.70  CPU: 0:00:06.57  BUFIO: 1  DIRIO: 25200o	 FAULTS: 2=4 > >      Total Fast I/O Reads: 12599; Writes: 12599.. > > Maximum Outstanding Reads: 16; Writes: 16.5 > > Output file truncated at VBN 1600012 by 0 blocks.p; > > New end-of-file block is 1600007, first free byte is 0.j > >a> > >>>>>	8:56, 12599 I/Os.  Same source and destination drives. > >  > >lE > > 	Summary: FAST_IO_COPY is about twice as fast as regular COPY andi> > > 	you may be able to do better as you can supply BLOCKS and= > > 	BUFFERS as args to FAST_IO_COPY... from the source code:  > >b1 > >     if (!strncmp(argv[ibuf], "-nbuffers", 9))E	 > >     {t > >       ibuf++;t > >       if (ibuf < argc) > >       {r( > >         nbuffers = atoi(argv[ibuf]);< > >         if ((nbuffers == 0) || (nbuffers > MAX_BUFFERS))% > >           nbuffers = MAX_BUFFERS;h > >n > > 	same for -nblocks > >h> > > 	Also.. even at that 1.46 MByte/sec using FAST_IO_COPY but@ > > 	these are plain old SCSI drives and I don't know the ratingB > > 	of the SWXCR but it can't be that great.  You should do quite > > 	a bit better regardless.a > >lC > > 	Request:  Perhaps VMS engineering could add a COPY/FAST switchaC > > 	that hooks to this routine.  In future versions of VMS, MEMBOB=C > > 	is no longer an issue (butchering Fred K.s view/words but theye> > > 	will no longer be checking as they are to keep users fromD > > 	chewing up system memory via other mechanisms) so maybe this is% > > 	a supportable qualifier to COPY.= > >  > > 	Caveat: > >P4 > > 	You most likely will have to jack MAXBOBMEM ... > >o > >e% > > $ fic :== $sys$login:fast_io_copy ) > > $ fic DRIVE1:[DIRECTORY]SMALLFILE.DATt DRIVE2:[DIRECTORY]SMALLFILE.DATVA > > Using 16 buffers, with 127 blocks (65024 bytes) per transfer.e' > > Could not create buffer object #12.SF > > %SYSTEM-F-EXBUFOBJLM, exceeded systemwide buffer object page limit (MAXBOBMEM)e > > $ set proc/priv=all  > > $ mcr sysgen > > SYSGEN>  SHOW MAXBOBMEM B > > Parameter Name           Current    Default     Min.      Max. Unit > > DynamiceD > > --------------           -------    -------    -------   ------- ---- > > -------eE > > MAXBOBMEM                    1600       1600         0         -1  Pagelets   DE > >  internal value               100        100         0         -1a Pages      D > > SYSGEN>  SET MAXBOBMEM 2400h > > SYSGEN>  WRITE ACTIVE  > > SYSGEN>  WRITE CURRENT > > SYSGEN>  SHOW MAXBOBMEMxB > > Parameter Name           Current    Default     Min.      Max. Unit > > Dynamic D > > --------------           -------    -------    -------   ------- ---- > > -------oE > > MAXBOBMEM                    2400       1600         0         -1t Pagelets   DE > >  internal value               150        100         0         -1a Pages      D > > SYSGEN>  EXITt > >t% > > $ fic :== $sys$login:fast_io_copyi) > > $ fic DRIVE1:[DIRECTORY]SMALLFILE.DATd DRIVE2:[DIRECTORY]SMALLFILE.DATDA > > Using 16 buffers, with 127 blocks (65024 bytes) per transfer.f+ > > Search: DRIVE1:[DIRECTORY]SMALLFILE.DATi+ > > Reading DRIVE1:[DIRECTORY]SMALLFILE.DATh+ > > Writing DRIVE2:[DIRECTORY]SMALLFILE.DATu > >n > > @ > > 	Tests on small alpha running vms 7.1.  If you can't compile? > > 	fast_io_copy, let me know and I will drop it out to you asf > > 	a zip attachment. > >  > > 				Roba > >e >o   --- *********************************************n( "All I every wanted from life was to see, Larry Wall give Bill Gates a Perl Necklace."     Sent via Deja.coma http://www.deja.com/   ------------------------------  # Date: Mon, 22 Jan 2001 19:44:12 GMTs" From: fooguy <jweisen@my-deja.com> Subject: Re: DS20 - Slow I/O) Message-ID: <94i2i9$3td$1@nnrp1.deja.com>u  - In article <87puhiqj6q.fsf@prep.synonet.com>, /   Paul Repacholi <prep@prep.synonet.com> wrote:  <...SNIP...>  @ > What do you get from back/phys/group=0/nocrc <disk> nla0:./sav< > ( you must have the dot )                                ^- > or back/ima/group=0/nocrc <disk> nla0:./sav-   It takes 28 mins 50 seconds,   > --> > Paul Repacholi                               1 Crescent Rd.,9 > +61 (08) 9257-1001                           Kalamunda.sB >                                              West Australia 60760 > Raw, Cooked or Well-done, it's all half baked. >e   --- ********************************************* ( "All I every wanted from life was to see, Larry Wall give Bill Gates a Perl Necklace."     Sent via Deja.comm http://www.deja.com/   ------------------------------  # Date: Mon, 22 Jan 2001 21:30:59 GMT " From: fooguy <jweisen@my-deja.com> Subject: Re: DS20 - Slow I/O) Message-ID: <94i8qj$a6e$1@nnrp1.deja.com>c   Let's try that again:x  - From back/phys/group=0/nocrc dra0: nla0:./savo   I get 1.5MB/Sece   2600MB / 1730Sec.t  G Also, someone mentioned turning high water marking off. Surprise, thereaG was actually a performance drop to .93MB/Sec with the same sample data.e  ' This is starting to really piss me off.a  E At the request of our vendor, I'm going to try moving the hard drives,D (not the SSD) to the external cabinet and see what we get from that.G They suspect that the backplane in the DS20 may not be as unintelligente8 as first thought (Since our AS800 is kicking it's butt).  ) In article <94i2i9$3td$1@nnrp1.deja.com>,n%   fooguy <jweisen@my-deja.com> wrote:. >. > / > In article <87puhiqj6q.fsf@prep.synonet.com>,a1 >   Paul Repacholi <prep@prep.synonet.com> wrote:t > <...SNIP...> > B > > What do you get from back/phys/group=0/nocrc <disk> nla0:./sav> > > ( you must have the dot )                                ^/ > > or back/ima/group=0/nocrc <disk> nla0:./sava >g > It takes 28 mins 50 secondsg >a > > --@ > > Paul Repacholi                               1 Crescent Rd.,; > > +61 (08) 9257-1001                           Kalamunda.eD > >                                              West Australia 60762 > > Raw, Cooked or Well-done, it's all half baked. > >e >  > --/ > *********************************************b* > "All I every wanted from life was to see. > Larry Wall give Bill Gates a Perl Necklace." >s > Sent via Deja.coma > http://www.deja.com/ >    --- *********************************************w( "All I every wanted from life was to see, Larry Wall give Bill Gates a Perl Necklace."     Sent via Deja.com. http://www.deja.com/   ------------------------------  % Date: Mon, 22 Jan 2001 18:31:13 -0500 2 From: norm lastovica <norman.lastovica@oracle.com> Subject: Re: DS20 - Slow I/O* Message-ID: <3A6CC2C1.747BBD5D@oracle.com>  
 fooguy wrote:h >  > Let's try that again:- > / > From back/phys/group=0/nocrc dra0: nla0:./savt  , 	this test is not measuring any writes; only2 reads from the source disk.  make certain to avoid0 any operating system caches or controller caches2 (mount the disk /NOCACHE and set the disk no cache2 in the controller).  otherwise, run-to-run, you'll2 likely suffer variablility due to what is or ain't cached.b   >  > I get 1.5MB/Sece >  > 2600MB / 1730Sec.e > I > Also, someone mentioned turning high water marking off. Surprise, there|I > was actually a performance drop to .93MB/Sec with the same sample data.:  / 	HWM ought to have no effect since you are only-5 reading *unless* you have quite a lot of unallocated   space. r   > ) > This is starting to really piss me off.0 > G > At the request of our vendor, I'm going to try moving the hard driveseF > (not the SSD) to the external cabinet and see what we get from that.I > They suspect that the backplane in the DS20 may not be as unintelligenti: > as first thought (Since our AS800 is kicking it's butt). > + > In article <94i2i9$3td$1@nnrp1.deja.com>,i' >   fooguy <jweisen@my-deja.com> wrote:  > >~ > >t1 > > In article <87puhiqj6q.fsf@prep.synonet.com>,s3 > >   Paul Repacholi <prep@prep.synonet.com> wrote:u > > <...SNIP...> > >uD > > > What do you get from back/phys/group=0/nocrc <disk> nla0:./sav@ > > > ( you must have the dot )                                ^1 > > > or back/ima/group=0/nocrc <disk> nla0:./savg > >s > > It takes 28 mins 50 seconds0 > >e > > > --B > > > Paul Repacholi                               1 Crescent Rd.,= > > > +61 (08) 9257-1001                           Kalamunda.tF > > >                                              West Australia 60764 > > > Raw, Cooked or Well-done, it's all half baked. > > >e > >o > > --1 > > ********************************************* , > > "All I every wanted from life was to see0 > > Larry Wall give Bill Gates a Perl Necklace." > >b > > Sent via Deja.comB > > http://www.deja.com/ > >u >  > --/ > *********************************************0* > "All I every wanted from life was to see. > Larry Wall give Bill Gates a Perl Necklace." >  > Sent via Deja.comd > http://www.deja.com/   --  > norman lastovica / oracle rdb engineering / usa / 610.696.4685   ------------------------------  % Date: Mon, 22 Jan 2001 21:01:23 +0100o" From: "Hans Vlems" <hvlems@iae.nl>/ Subject: Re: Dual ethernet config for failover?e( Message-ID: <94i3bk$8b4$1@news.IAEhv.nl>  > Agreed, but the poster stated clearly he was running phase IV. Hence the advise.T   Hans  0 Robert Deininger heeft geschreven in bericht ...F >In article <94ei3d$53i$1@news.IAEhv.nl>, "Hans Vlems" <hvlems@iae.nl> wrote: >o >iL >> A machine that runs DECnet will replace the hardware mac address(es) withF >> the DECnet mac address (AA-00-04-00-xx-xx, where the last two bytes encodeJ >> the DECnet address). This happens on all LAN interfaces. So connect two of >> themsH >> to the same switch and the switch will shutdown one of the two ports.K >> Put them on different switches and even more dramatic things may happen,  >> since; >> ethernet does not allow multiple paths between segments.r > J >I think your warning does not apply with DECnet plus.  I haven't done it,K but I think you can have multiple LAN interfaces configured, and they DON'Tt! end up with the same MAC address.c >o >--  >Robert Deiningerj >rdeininger@mindspring.com   ------------------------------  % Date: Tue, 23 Jan 2001 04:54:00 +1000.? From: "Matt Muggeridge" <Matt.Muggeridge@compaq.com.spammffree> / Subject: Re: Dual ethernet config for failover? 5 Message-ID: <F61b6.27$cu.319@gazette.loc1.tandem.com>   3 > 1. How does DNS resolve a name to 2 IP addresses?e  L DNS uses round-robin when handing out addresses.  However, this doesn't giveE you failover.  In fact, if one of the interfaces fails, then DNS will G dutifully continue to handout both interfaces in a round-robin fashion, K which will mean some users will have intermittent trouble connecting.  That, would be a pain.  F To resolve this, you can add the load-broker and metric server to yourI network.  The load-broker will poll the metric server, running on each of G the participating nodes.  If the metric server fails to respond after 3mK attempts, the load-broker will remove that interface from the DNS database.s+ This requires Dynamic DNS to be configured.v  K This fixes the inbound traffic problem by preventing anyone from connectingsF to the bad interface, however, there is another problem.  The outboundL traffic will consult the local routing table to determine which interface toK use to send data.  So if the bad interface reamains configured, then it may L be selected for outbound traffic, which of course also causes the network toI fail.  To address this, you might develop a command procedure which pingslI the default router.  If it fails (after 3 attempts, say), then remove therD failed interface, (TCPIP SET NOINTERFACE).  This will remove the badG interface from the routing table, and so you should continue operating. G (There are other factors too - e.g. are interfaces in the same subnet.)c  H Then you can make use of the internet cluster alias, which requires yourJ nodes to be configured in a VMS cluster.  If one node in the cluster failsJ (rebooted etc) then one of the other cluster members will become the ownerH of the cluster IP address.  (The cluster alias doesn't give you failover7 between controllers, it is a failover between systems)..  I Finally, with multiple controllers, you could attach each controller to anF different subnet, then place a router between the subnets.  Then if anH interface fails, the router should be able to dynamically reroute to theI available interface.  (I haven't tried this configuration, but have heard C others speak of it, and in theory it should work - you will need an 2 excellent working knowledge of routing protocols).   Cheers,- Matt.-   ------------------------------  % Date: Tue, 23 Jan 2001 11:54:00 +1300n9 From: "Antony Wardle" <antony.wardle@nnnoospam.met.co.nz>s Subject: ftp1 Message-ID: <VL2b6.3857$cF2.79328@ozemail.com.au>n   Ok,1  , how do I change disk when I have ftp'ed into my axp 7.1 machine?b     Antony   ------------------------------   Date: 22 Jan 2001 15:39:00 PSTT From: Fairfield@SLAC.Stanford.EDU (Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515) Subject: Re: ftp3 Message-ID: <bv3R85MDVFyU@mccdev.slac.stanford.edu>h  2 In article <VL2b6.3857$cF2.79328@ozemail.com.au>, ?     "Antony Wardle" <antony.wardle@nnnoospam.met.co.nz> writes:   . > how do I change disk when I have ftp'ed into > my axp 7.1 machine?R  A         Use VMS syntax, eh?  Like: ftp> cd disk$user:[dir.subdir]4               -Ken -- iM  Kenneth H. Fairfield            |  Internet: Fairfield@SLC.Slac.Stanford.Edu.:  SLAC, 2575 Sand Hill Rd, MS 46  |  Voice:    650-926-2924:  Menlo Park, CA  94025           |  FAX:      650-926-3515N  -----------------------------------------------------------------------------B  These opinions are mine, not SLAC's, Stanford's, nor the DOE's...   ------------------------------  % Date: Mon, 22 Jan 2001 18:53:41 +0000a  From: steven.reece@quintiles.com$ Subject: Re: GS160 hardware questionH Message-ID: <OF9A7C4B46.30DF769F-ON802569DC.0067423B@qedi.quintiles.com>  J But then it depends upon what the customer's priorities are.  If they wantH availability and the ability for one instance of the operating system toI crash without affecting application availability then they would probably 6 go for OPS in a box or OPS across a number of systems.  F If, on the other hand, they want minimum footprint and have no worriesK about whether the application remains available then they would more likelye go with a single SMP system.   Horses for courses and YMMV.   Steve.   The SUNny Mr. Harrison wrote:e7 >>>Take for example a 20 CPU GS320, 5 QBBS, 5 instancesi6 of Oracle using OPS. Now what happens if your customer6 wants to increas the CPU count to say 24 to get better0 throughput. Well guess what with OPS you have to4 re-partition your data and your clients, this can be1 a costly process and is not necessary on a single 2 instance SMP system. This ongoing cost is just one& not measured or reported for TPC-C.<<<   ------------------------------    Date: 22 Jan 2001 20:46:39 +0100) From: sfb@consultron.com (Stefan K. Berg)i1 Subject: How to delete directory with bad blocks?e2 Message-ID: <m3u26rz7xc.fsf@hermes.consultron.com>   Hi,n  N I just brought a MicroVAX 3100 running OpenVMS 6.0 back to life. The installedN system mostly works, but I have bad blocks in a directory - hopefully the disk6 isn't crashing on me, and this is an isolated failure.  B Question: How do I delete a directory which I can't access - I getM SYSTEM-F-PARITY, parity error when I enter it and tries to list the contents?t  + I've tried ANALYZE/DISK/REPAIR to no avail.   9 Is my only option to reformat the disk and reinstall VMS?e   Regards,   Stefan   ------------------------------  # Date: Mon, 22 Jan 2001 20:30:39 GMTi7 From: moroney@world.std.spaamtrap.com (Michael Moroney)k5 Subject: Re: How to delete directory with bad blocks? & Message-ID: <G7Kzn3.CwE@world.std.com>  + sfb@consultron.com (Stefan K. Berg) writes:n  C >Question: How do I delete a directory which I can't access - I get N >SYSTEM-F-PARITY, parity error when I enter it and tries to list the contents?  , >I've tried ANALYZE/DISK/REPAIR to no avail.  : >Is my only option to reformat the disk and reinstall VMS?    From a privileged account, do a @ $ SET FILE/NODIRECTORY disk:[whatever]bad_directory.DIR  then doM $ DELETE/ERASE bad_directory.DIR;   to get rid of it.  I suggest DELETE/ERASETG so the bad blocks are overwritten (and possibly revectored) now so thatOG they are gone.  If you get another parity error here, replace the disk..  G Then do ANALYZE/DISK/REPAIR, this will place any files that were in the  bad directory in [SYSLOST].    -Miken   ------------------------------  % Date: Mon, 22 Jan 2001 15:44:28 -05009  From: norm.raphael@jamesbury.com5 Subject: Re: How to delete directory with bad blocks?b4 Message-ID: <C22569DC.00712953.00@jklh21.valmet.com>  0 SET FILE/NODIRECTORY     ddcu:[dir]whatever.dir;/ Then DELETE/LOG/CONFIRM ddcu:[dir]whatever.dir;u9 Then ANAL/DISK should bring any files into ddcu:[SYSLOST]r        , sfb@consultron.com on 01/22/2001 02:46:39 PM   To:   Info-VAX@mvb.saic.com  cc:   Info-VAX@mvb.saic.comv2 Subject:  How to delete directory with bad blocks?           Hi,S  N I just brought a MicroVAX 3100 running OpenVMS 6.0 back to life. The installedN system mostly works, but I have bad blocks in a directory - hopefully the disk6 isn't crashing on me, and this is an isolated failure.  B Question: How do I delete a directory which I can't access - I getM SYSTEM-F-PARITY, parity error when I enter it and tries to list the contents?C  + I've tried ANALYZE/DISK/REPAIR to no avail.n  9 Is my only option to reformat the disk and reinstall VMS?i   Regards,   Stefan   ------------------------------  # Date: Mon, 22 Jan 2001 20:46:35 GMTk* From: Alan E. Feldman <alan48@my-deja.com>5 Subject: Re: How to delete directory with bad blocks?t) Message-ID: <94i677$7g4$1@nnrp1.deja.com>:  2 In article <m3u26rz7xc.fsf@hermes.consultron.com>,,   sfb@consultron.com (Stefan K. Berg) wrote: >_ > Hi,  > F > I just brought a MicroVAX 3100 running OpenVMS 6.0 back to life. The	 installednG > system mostly works, but I have bad blocks in a directory - hopefully  the disk8 > isn't crashing on me, and this is an isolated failure. >dD > Question: How do I delete a directory which I can't access - I getE > SYSTEM-F-PARITY, parity error when I enter it and tries to list the 	 contents?   % Could you be a little more vague? :-)y  A OK, did you set default to that directory, then ran the DIRECTORY-! command and got the parity error?B  F Did you do a DIRECTORY BADDIR.DIR where [BADDIR] is the directory-spec for the bad directory?  @ Can you produce the complete exact output of the *all* the error	 messages?S  ( Are you sure that this *is* a directory?  6 What do you get when you do DIRECTORY/FULL baddir.DIR?   > - > I've tried ANALYZE/DISK/REPAIR to no avail.   4 What, if any, messages were there from this command?   > ; > Is my only option to reformat the disk and reinstall VMS?- >   
 Probably not.   ( afeldman@gfigroup.ButItSaidItPrinted.com   --F NOTE: If you wish to e-mail me, please do NOT use the deja address. ItD is not a valid address. Instead, use the address below, removing the long wrong part first. Thanks.   Disclaimer: JMHO Alan E. Feldman  &-)( afeldman@gfigroup.ButItSaidItPrinted.com     Sent via Deja.como http://www.deja.com/   ------------------------------    Date: 22 Jan 2001 23:20:39 +0100) From: sfb@consultron.com (Stefan K. Berg)n5 Subject: Re: How to delete directory with bad blocks?s2 Message-ID: <m3puhfz0so.fsf@hermes.consultron.com>   Hi,   F Many thanks to Michael, Norm and everyone else who has kindly replied.  E Apparently most of my previous VMS knowledge has been overwritten due*D to recent Unix exposure - I've done "set file/nodir" before. In the * beginning of the previous decade or so. :)  , Alan E. Feldman <alan48@my-deja.com> writes:  ' > Could you be a little more vague? :-)2  E Well, if you think *that's* vague you should hear me on one of my bad: days. :)  C > OK, did you set default to that directory, then ran the DIRECTORYi# > command and got the parity error?R > H > Did you do a DIRECTORY BADDIR.DIR where [BADDIR] is the directory-spec > for the bad directory? > B > Can you produce the complete exact output of the *all* the error > messages?r > * > Are you sure that this *is* a directory? > 8 > What do you get when you do DIRECTORY/FULL baddir.DIR? >  > > / > > I've tried ANALYZE/DISK/REPAIR to no avail.  > 6 > What, if any, messages were there from this command? >  > > = > > Is my only option to reformat the disk and reinstall VMS?  > >a  J To be a bit more verbose: I followed Michael's procedure with a successful result as can be seen below.  J If bad blocks returns to bite me I suppose I will have to go hunting for a( replacement RZ24 or other suitable disk.   Again, thanks for all help.a   Regards,   Stefan  M -----------------------------------------------------------------------------i $ analyze/disk/repair dkb200:eM Analyze/Disk_Structure/Repair for _DKB200: started on 22-JAN-2001 22:55:53.68   . %ANALDISK-I-OPENQUOTA, error opening QUOTA.SYS" -SYSTEM-W-NOSUCHFILE, no such file; %ANALDISK-I-BBLHEADER, file (1859,1,1) [000000]SHAZAM.DIR;13 	contains suspected bad blocks5 %ANALDISK-W-READDIR, error reading directory [SHAZAM]e -SYSTEM-F-PARITY, parity error   $ dir/full [000000]shazam.dir    Directory DKB200:[000000]   ; SHAZAM.DIR;1                  File ID:  (1859,1,0)         a4 Size:            1/3          Owner:    [ADM,SYSTEM]" Created:   23-JUN-1993 13:54:14.56& Revised:   22-JAN-2001 22:54:56.15 (4) Expires:   <None specified>g Backup:    <No backup recorded>  Effective: <None specified>j Recording: <None specified>  File organization:  SequentialD File attributes:    Allocation: 3, Extend: 0, Global buffer count: 0K                     No default version limit, Contiguous, MoveFile disabled <                     Directory file, File contains bad blocks6 Record format:      Variable length, maximum 512 bytes4 Record attributes:  No carriage control, Non-spanned RMS attributes:     None Journaling enabled: None= File protection:    System:RWE, Owner:RWE, Group:RE, World:RE  Access Cntrl List:  None   Total of 1 file, 1/3 blocks. $ dir/full [shazam].= %DIRECT-E-OPENIN, error opening DKB200:[SHAZAM]*.*;* as input*/ -RMS-E-FND, ACP file or directory lookup failed  -SYSTEM-F-PARITY, parity error* $ set file/nodir dkb200:[000000]shazam.dir) $ delete/erase dkb200:[000000]shazam.dir;*   $ analyze/disk/repair dkb200: M Analyze/Disk_Structure/Repair for _DKB200: started on 22-JAN-2001 22:58:01.77k  . %ANALDISK-I-OPENQUOTA, error opening QUOTA.SYS" -SYSTEM-W-NOSUCHFILE, no such file $ sh dev/full dkb200:0  I Disk DKB200:, device type RZ24, is online, mounted, file-oriented device,c(     shareable, error logging is enabled.  O     Error count                    4    Operations completed               4534mO     Owner process                 ""    Owner UIC                  [ADM,SYSTEM]tO     Owner process ID        00000000    Dev Prot            S:RWPL,O:RWPL,G:R,WmO     Reference count               23    Default buffer size                 512 O     Total blocks              409792    Sectors per track                    38eO     Total cylinders             1348    Tracks per cylinder                   8e  O     Volume label        "OPENVMS060"    Relative volume number                0 O     Cluster size                   3    Transaction count                   179oO     Free blocks                23016    Maximum files allowed            102448 O     Extend quantity                5    Mount count                           1 O     Mount status              System    Cache name           "_DKB200:XQPCACHE"tO     Extent cache size             64    Maximum blocks in extent cache     2301aO     File ID cache size            64    Blocks currently in extent cache      0cO     Quota cache size               0    Maximum buffers in FCP cache        966>O     Volume owner UIC           [1,1]    Vol Prot    S:RWCD,O:RWCD,G:RWCD,W:RWCDj  N   Volume status:  subject to mount verification, protected subsystems enabled,=       file high-water marking, write-through caching enabled.o   $ dir dkb200:[syslost]! %DIRECT-W-NOFILES, no files found   M -----------------------------------------------------------------------------a   ------------------------------  % Date: Tue, 23 Jan 2001 11:50:03 +1300c9 From: "Antony Wardle" <antony.wardle@nnnoospam.met.co.nz>65 Subject: Re: How to delete directory with bad blocks? 1 Message-ID: <dI2b6.3855$cF2.79829@ozemail.com.au>W    well you might be able to do a :     set file/removea   on the directory.e  * don't know if that is the best way though.   antony    6 "Stefan K. Berg" <sfb@consultron.com> wrote in message, news:m3u26rz7xc.fsf@hermes.consultron.com... >  > Hi,  >tF > I just brought a MicroVAX 3100 running OpenVMS 6.0 back to life. The	 installed*K > system mostly works, but I have bad blocks in a directory - hopefully thea disk8 > isn't crashing on me, and this is an isolated failure. >tD > Question: How do I delete a directory which I can't access - I getE > SYSTEM-F-PARITY, parity error when I enter it and tries to list the 	 contents?: >2- > I've tried ANALYZE/DISK/REPAIR to no avail.u >c; > Is my only option to reformat the disk and reinstall VMS?b >b
 > Regards, >n > Stefan >    ------------------------------  # Date: Mon, 22 Jan 2001 23:14:09 GMTT* From: Alan E. Feldman <alan48@my-deja.com>5 Subject: Re: How to delete directory with bad blocks?n) Message-ID: <94iero$fus$1@nnrp1.deja.com>o  2 In article <m3puhfz0so.fsf@hermes.consultron.com>,,   sfb@consultron.com (Stefan K. Berg) wrote: >d > Hi,A >0H > Many thanks to Michael, Norm and everyone else who has kindly replied. >hG > Apparently most of my previous VMS knowledge has been overwritten dueeE > to recent Unix exposure - I've done "set file/nodir" before. In the , > beginning of the previous decade or so. :) >g. > Alan E. Feldman <alan48@my-deja.com> writes: >w) > > Could you be a little more vague? :-)t >kG > Well, if you think *that's* vague you should hear me on one of my bad 
 > days. :)  A I'm sorry I jumped the gun a bit. Maybe I'm just in a cruddy moodt@ lately. Sorry. But it would have been nice to have seen at least  ? > %DIRECT-E-OPENIN, error opening DKB200:[SHAZAM]*.*;* as input 1 > -RMS-E-FND, ACP file or directory lookup failedM  > -SYSTEM-F-PARITY, parity error  E You only put the last line in your original post and I don't remember6B that from bad directory format messages that I had many years ago.F That's why I wasn't sure what you were doing. The last line could show up in numerous scenarios.   F Actually, many years ago, mid 80's or so, I had a directory that afterG having set default to it, DIRECTORY would work fine on it from one nodenB but if I ran DIRECTORY on it from another node, it would give "badB directory file format" or similar message upon reaching files thatD began with S. These were two nodes in a cluster accessing the *same*B directory on the *same* disk! Same version of VMS, same everythingG else. (Well, obviously *something* was different, but I never found outeF what.) I had to copy the files out, nuke the directory, and rename theE temp dir back. And then I think it happened again! Never did find out G what really happened. So, you could have wiped out a directory that wasiE readable from another node. Not too likely, and it appears there wereg no files in it anyway.  B Hey, if Carl Lydick were still around, he'd flame you severely and mercilessly. (!)  E I would advise you to keep an eye on the error count for this disk bye< periodically, at least once per workday, running the command   $ SHOW ERROR   !.l  = If the error count starts increasing rapidly, BACKUP the diskrE immediately!!! I'd check all errors by running the ANAL/ERROR commande= (I think you could add /INCLUDE=disk-name to filter out otheroF messages). If the ANAL/ERROR output looks scary, contact your hardwareD maintenance company or just replace the disk. I think an occasional,& successfully revectored block is okay.  E > > OK, did you set default to that directory, then ran the DIRECTORY % > > command and got the parity error?e > >lF > > Did you do a DIRECTORY BADDIR.DIR where [BADDIR] is the directory- spec > > for the bad directory? > >rD > > Can you produce the complete exact output of the *all* the error
 > > messages?l > >a, > > Are you sure that this *is* a directory? > >a: > > What do you get when you do DIRECTORY/FULL baddir.DIR? > >e > > > 1 > > > I've tried ANALYZE/DISK/REPAIR to no avail.a > >n8 > > What, if any, messages were there from this command? > >  > > >s? > > > Is my only option to reformat the disk and reinstall VMS?e > > >o >nA > To be a bit more verbose: I followed Michael's procedure with a-
 successful > result as can be seen below. >1F > If bad blocks returns to bite me I suppose I will have to go hunting for aj* > replacement RZ24 or other suitable disk. >m > Again, thanks for all help.I >a
 > Regards, >  > Stefan >nH > ---------------------------------------------------------------------- -------J > $ analyze/disk/repair dkb200:iC > Analyze/Disk_Structure/Repair for _DKB200: started on 22-JAN-20011 22:55:53.68R >f0 > %ANALDISK-I-OPENQUOTA, error opening QUOTA.SYS$ > -SYSTEM-W-NOSUCHFILE, no such file= > %ANALDISK-I-BBLHEADER, file (1859,1,1) [000000]SHAZAM.DIR;1o  > 	contains suspected bad blocks7 > %ANALDISK-W-READDIR, error reading directory [SHAZAM]c  > -SYSTEM-F-PARITY, parity error >  > $ dir/full [000000]shazam.dird >4 > Directory DKB200:[000000]K >e4 > SHAZAM.DIR;1                  File ID:  (1859,1,0)6 > Size:            1/3          Owner:    [ADM,SYSTEM]$ > Created:   23-JUN-1993 13:54:14.56( > Revised:   22-JAN-2001 22:54:56.15 (4) > Expires:   <None specified>-! > Backup:    <No backup recorded>- > Effective: <None specified>  > Recording: <None specified>s  > File organization:  SequentialF > File attributes:    Allocation: 3, Extend: 0, Global buffer count: 0D >                     No default version limit, Contiguous, MoveFile disabled> >                     Directory file, File contains bad blocks8 > Record format:      Variable length, maximum 512 bytes6 > Record attributes:  No carriage control, Non-spanned > RMS attributes:     None > Journaling enabled: None? > File protection:    System:RWE, Owner:RWE, Group:RE, World:RE  > Access Cntrl List:  None >y > Total of 1 file, 1/3 blocks. > $ dir/full [shazam]p? > %DIRECT-E-OPENIN, error opening DKB200:[SHAZAM]*.*;* as inputv1 > -RMS-E-FND, ACP file or directory lookup failedg  > -SYSTEM-F-PARITY, parity error [...]Y --F NOTE: If you wish to e-mail me, please do NOT use the deja address. ItD is not a valid address. Instead, use the address below, removing the long wrong part first. Thanks.s   Disclaimer: JMHO Alan E. Feldman  &-)( afeldman@gfigroup.ButItSaidItPrinted.com     Sent via Deja.com  http://www.deja.com/   ------------------------------  % Date: Tue, 23 Jan 2001 00:34:02 +0000S4 From: John Laird <john@laird-towers.freeserve.co.uk>5 Subject: Re: How to delete directory with bad blocks?-8 Message-ID: <vujp6t0csnq3s303756vb70a2uhgc6jjih@4ax.com>  A On Mon, 22 Jan 2001 20:30:39 GMT, moroney@world.std.spaamtrap.com  (Michael Moroney) wrote:  , >sfb@consultron.com (Stefan K. Berg) writes: >tD >>Question: How do I delete a directory which I can't access - I getO >>SYSTEM-F-PARITY, parity error when I enter it and tries to list the contents?k >s- >>I've tried ANALYZE/DISK/REPAIR to no avail.l >.; >>Is my only option to reformat the disk and reinstall VMS?s >I! >From a privileged account, do a  A >$ SET FILE/NODIRECTORY disk:[whatever]bad_directory.DIR  then doYN >$ DELETE/ERASE bad_directory.DIR;   to get rid of it.  I suggest DELETE/ERASEH >so the bad blocks are overwritten (and possibly revectored) now so thatH >they are gone.  If you get another parity error here, replace the disk. >aH >Then do ANALYZE/DISK/REPAIR, this will place any files that were in the >bad directory in [SYSLOST].  @ This is all sound.  Additionally, if the blocks have been markedE "suspect", then the deletion of the directory file will trigger a badIC block scanning process (it has a very obvious name), which IIRC mayID attempt to read every block on the disk.  This may however be faultyG memory blocks on my part, as I have only encountered this once or twice G and it may simply have been that the suspect file was very large and itE5 merely looked as if the whole disk was being checked.D  - $ DUMP/HEAD/BLO=COU=0 disk:[000000]BADBLK.SYSa  E will show if any blocks have been sent there.  Ignore the very end ofpC the disk which is commonly marked "bad" to avoid a partial cluster.t     	Johna   -- f
 John Laird   ------------------------------  % Date: Mon, 22 Jan 2001 20:59:14 -0600-7 From: "David J. Dachtera" <djesys.nospam@earthlink.net> 1 Subject: Imagine 30GB in your Laptop or Multia...e- Message-ID: <3A6CF382.71794711@earthlink.net>2   Sorry if this URL wraps...  L http://www.infoworld.com/articles/hn/xml/01/01/22/010122hndrive.xml?0122mnpm   -- u David J. Dachterau dba DJE Systemss http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/s  F This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings is to be expected.  @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.    ------------------------------  % Date: Mon, 22 Jan 2001 22:34:48 -0500n% From: "Bill Stouffer" <farm@erie.net>a; Subject: Is Compaq dropping support for vms on vax platformr/ Message-ID: <t6puuq8mnfrr3b@corp.supernews.com>d  J Has anyone been informed or heard rumors about Compaq dropping support for OpenVMS on the VAX platform?   bill_stouffer@lord.com   ------------------------------  % Date: Mon, 22 Jan 2001 23:14:24 -0500e' From: Randy Hawley <rhawley@iquest.net>H? Subject: Re: Is Compaq dropping support for vms on vax platform$* Message-ID: <3A6D0520.82812F70@iquest.net>   Bill Stouffer wrote:  L > Has anyone been informed or heard rumors about Compaq dropping support for > OpenVMS on the VAX platform? >S > bill_stouffer@lord.com  N Last I heard, the VAX chip is no longer being produced by Compaq.  I have alsoL heard Compaq say that they will support VMS/Vax as long as customers wanted,M or some such verbage.  (Of course, I think that means that as long as you areRL willing to pay for it.  How long will it be before they price the support soJ high as to make migrating to Alpha pay off in 6 months is anybody's guess)  M I take care of VAXen at work that we expect Compaq to accept money from us toRJ support for at least another 5 years.  Probably they won't last that long,< though, before they are replaced by other systems or Alphas.  = (anybody know if VAX chips are being fabbed by anybody else?)        Randy Hawley   ------------------------------  # Date: Mon, 22 Jan 2001 20:55:33 GMTA2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: KZQSA5 Message-ID: <971b6.28$cu.194@gazette.loc1.tandem.com>i  R In article <7uFWdEjEdfB@pdp.xnc.com>, t.haeger@pdp.xnc.com (Thomas Haeger) writes:2 :Hello, I'am serching for a discription/docs for a+ :KZQSA (M5976-SA) Q-BUS to SCSI controller.n& :- JUMPER discription (CSR-Adress ...)? :- with driver I'am have to LOAD/SYSGEN under OVMS (DKDRIVER ?)e  J   I don't have the CSR and vector settings handy for the KZQSA, but I willL   warn you that you may/will need to change CSR and vector settings on otherI   Q-bus modules.  You will NOT want to try to connect this widget to SCSItL   magnetic disks -- that configuration is not supported and not recommended.J   You will not have to manually connect any device driver for the widget, I   the appropriate drivers for the supported SCSI tapes and the supported vJ   CD-ROM disks will be automatically configured once the CSRs and vectors #   are correct on all Q-bus modules.e  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Mon, 22 Jan 2001 21:35:14 +0000r) From: Christof Brass <brass@infopuls.com>o- Subject: Re: Linux worm and RedHat 7.0 brokenO, Message-ID: <3A6CA792.94B14667@infopuls.com>   Arne Vajhj wrote: >  > Paul Sture wrote:rH > > Before SuSe, I was thinking of moving to NetBSD/FreeBSD, which again/ > > Netcraft report as being popular out there.  > G > It is my impression that *BSD are very good server operating systems.iB > They are more robust than Linux. For desktop usage I think Linux< > has more applications. I many ways it is unfair that Linux > got all the publicity. > L > > Solaris? Yes, attractive, but guess what? AH's rantings posted here have8 > > rather put me off spending a single dollar with Sun! > $ > I think Solaris for x86 are free ! > > > But I would not choose that either. I do not think there are& > much apps available for Solaris x86. >  > Arne  @ A friend of mine who is very familiar with several UNIX variants3 is fond of FreeBSD because of its stability, systemz: administration, development style and politics, quality of? service if one can say so. We will see if Solaris will survive.S; SUN made it open source and there is a chance that the open 5 source comunity will be interested in it and put somea development effort in it. > My experience is very little but compared to Linux on the same" machine it is a little bit slower.> FreeBSD can execute most of the other UNIX's apps - especially< the Linux executable formats like Minix, System V, a.out and= ELF. I don't like the Linux way very much because the creatore jumped 29 into this business without knowing much - Linus Thorvaldsa: forgive me if I'm wrong - and by that repeating Micro$ofts9 mistake: to give away a "unique" chance for designing andi< implementing a real new OS. Linux has the old, obsolete UNIX; architecture and is by that like all other UNIXes much lesst modern than VMS. s= Linus Thorvalds is great in organising work and people and hel: might be a good coder. He is not a good system analyst and= architect IMHO. I feel sorry for the FreeBSD people who did ar9 great job for many years without getting as close as much:3 attention as they deserve compared with e.g. Linux.s@ I'll switch over to FreeBSD from Linux rather sooner than later.? I've already have a FreeBSD installation on one partition whichn? can be booted alternatively to Linux. And FreeBSD supports the -@ Alpha processor since about 6 months (Linux offers Alpha support
 much longer).d> From my point of view the Alpha seems the only processor which= can keep up with the Intels or even be in front as it was for > several years. If we look at SPARC - ridiculous; PowerPC - not= worth to be mentioned - ask Steve Jobs; MIPS not that bad buti? out of WS business more or less instead doing well in Sony PSs;e= PA-RISC a competitor but no market share, no OSs, no apps, noo? future. From that point of view I think DEC did a very good job3? compared to SUN, IBM, HP and MIPS even though e.g. SUN, IBM and @ HP could have spent much more money and probably did. IBM cannot; be beaten because they built up an incredible huge research ; branch where sometimes even Nobel price awarded results areo= achieved. IBM has a worldwide system developed to brain drain 7 other countries like India or Europe (I know it's not as "country"). @ There is no other industrial player comparable to IBM. They even@ compete with public universities. But there is still one problem@ with IBM. They use too much unethical practices like corruption = and black mail and they use methods which prevent them to get1; good results in the large. E.g. the OS/400 command language ? design is crap. WebSphere is a sometimes running piece of shit. = Or think about the miserably failed Taligent project/company.z= If you have the wrong system architecture you can put as much 9 money and people in you won't get out and find a straightx? forward elegant solution; unless you throw away and start over.u? You cannot replace one very good person by hundreds of mediocre # people. Sorry for losing the topic.:< Of course I'm on my way back to VMS and I will do as much as= possible on VMS because the architecture is clean in every of,? the three important dimensions. And the UNIX architecture will C2 never be clean; if it will it is not UNIX anymore.   ------------------------------    Date: 22 Jan 2001 16:46:00 -0500* From: young_r@eisner.decus.org (Rob Young)- Subject: Re: Linux worm and RedHat 7.0 brokent+ Message-ID: <Fud93jHIgI9h@eisner.decus.org>:  X In article <3A6CA792.94B14667@infopuls.com>, Christof Brass <brass@infopuls.com> writes:  9 	Regarding Unix and system research.  Rob Pike has a very 3 	downbeat view of Linux.  It's just another Unix int5 	his view.  Summarizing his lament:  Truly innovativel> 	systems research isn't being down any more.  Short term focus 	has us in its grip.  ( 	If you can read postscript, click here:  5 http://www.cs.bell-labs.com/cm/cs/who/rob/utah2000.psl  3 	Or you can follow the discussion from June 2000 inl
 	slashdot:  3 http://slashdot.org/articles/00/06/06/1151209.shtmlr  ) 	I've checked, and both links are active.r   				Roba   	[snip]-  @ > FreeBSD can execute most of the other UNIX's apps - especially> > the Linux executable formats like Minix, System V, a.out and? > ELF. I don't like the Linux way very much because the creators	 > jumped  ; > into this business without knowing much - Linus Thorvalds-< > forgive me if I'm wrong - and by that repeating Micro$ofts; > mistake: to give away a "unique" chance for designing ando> > implementing a real new OS. Linux has the old, obsolete UNIX= > architecture and is by that like all other UNIXes much lesso > modern than VMS. K? > Linus Thorvalds is great in organising work and people and hed< > might be a good coder. He is not a good system analyst and? > architect IMHO. I feel sorry for the FreeBSD people who did an; > great job for many years without getting as close as much 5 > attention as they deserve compared with e.g. Linux.nB > I'll switch over to FreeBSD from Linux rather sooner than later.A > I've already have a FreeBSD installation on one partition whichhA > can be booted alternatively to Linux. And FreeBSD supports the eB > Alpha processor since about 6 months (Linux offers Alpha support > much longer).i   	[snip]s   ------------------------------  % Date: Mon, 22 Jan 2001 22:35:55 +0100n2 From: martin@radiogaga.harz.de (Martin Vorlaender)- Subject: Re: Linux worm and RedHat 7.0 brokenn; Message-ID: <3a6ca7bb.524144494f47414741@radiogaga.harz.de>e  * Wayne Holland (wholland@tscnet.com) wrote: : Martin Vorlaender wrote:. : > Wayne Holland (wholland@tscnet.com) wrote:H : > : I had Suse also. [...] I moved over to Calder Open Linux 2.4 [...]M : > : Caldera also sports Webmin, a GUI system admin tool thats well endowed.E : > 7 : > As an aside to this already off-topic thread: [...]O : >   : > To get on-topic again: [...] :b. : Oooopss!  Sorry to have broken the thread...  E No offense meant. The diversity of threads in this newsgroup does notcC cease to astonish me. This bunch of people has a horizon wide open.," I appreciate being a part of it...   cu,o   Martin --J One OS to rule them all       | Martin Vorlaender  |  VMS & WNT programmer7 One OS to find them           | work: mv@pdv-systeme.dehN One OS to bring them all      |       http://www.pdv-systeme.de/users/martinv/> And in the Darkness bind them.| home: martin@radiogaga.harz.de   ------------------------------  % Date: Tue, 23 Jan 2001 01:26:25 +0000,) From: Christof Brass <brass@infopuls.com>m- Subject: Re: Linux worm and RedHat 7.0 broken , Message-ID: <3A6CDDC1.45528AA7@infopuls.com>   Rob Young wrote: > Z > In article <3A6CA792.94B14667@infopuls.com>, Christof Brass <brass@infopuls.com> writes: > B >         Regarding Unix and system research.  Rob Pike has a very< >         downbeat view of Linux.  It's just another Unix in> >         his view.  Summarizing his lament:  Truly innovativeG >         systems research isn't being down any more.  Short term focus  >         has us in its grip.k > 1 >         If you can read postscript, click here:  > 7 > http://www.cs.bell-labs.com/cm/cs/who/rob/utah2000.pse  ? Read that, thanks. Honestly I didn't read it before but I agree ? and recommend reading to all. It is easy reading, 23 very shortt> pages, takes about 10 minutes of reading. Unfortunately one of6 most interesting figures is not well displayed with my? ghostscript or the figure is damaged. It is the graph on page 4 8 named "New OS at SOSP". Is correctly displayed with your> equipment? I only see the frame and the text around not single dot inside. ; What he states on page 22 is not true. There are successful = component models like EJB. But what he claims with respect tos@ systems research was recently stated by Prof. Niklaus Wirth on a* similar subject, computer language design.   > < >         Or you can follow the discussion from June 2000 in >         slashdot:  > 5 > http://slashdot.org/articles/00/06/06/1151209.shtmls > 2 >         I've checked, and both links are active. > % >                                 Rob-  @ Thanks for the second link. Reading that material takes a lot of> time though. I already read the about 5% and supports my point of view.   ------------------------------  % Date: Mon, 22 Jan 2001 17:29:39 -0800g) From: Wayne Holland <wholland@tscnet.com> - Subject: Re: Linux worm and RedHat 7.0 broken[O Message-ID: <495B30C6E0CF7893.358229298D034398.7519932B40CC159A@lp.airnews.net>r   Martin Vorlaender wrote: > , > Wayne Holland (wholland@tscnet.com) wrote: > : Martin Vorlaender wrote:0 > : > Wayne Holland (wholland@tscnet.com) wrote:J > : > : I had Suse also. [...] I moved over to Calder Open Linux 2.4 [...]O > : > : Caldera also sports Webmin, a GUI system admin tool thats well endowed.2 > : >.9 > : > As an aside to this already off-topic thread: [...]c > : >o" > : > To get on-topic again: [...] > : 0 > : Oooopss!  Sorry to have broken the thread... > G > No offense meant. The diversity of threads in this newsgroup does notoE > cease to astonish me. This bunch of people has a horizon wide open.e$ > I appreciate being a part of it... >  > cu,i
 >   Martin > --L > One OS to rule them all       | Martin Vorlaender  |  VMS & WNT programmer9 > One OS to find them           | work: mv@pdv-systeme.debP > One OS to bring them all      |       http://www.pdv-systeme.de/users/martinv/@ > And in the Darkness bind them.| home: martin@radiogaga.harz.deC Whew!  Thank you!  I was afraid that this ng was getting to be like  alt.solaris.x86.E There ya gotta be real careful how you phrase a question.  You almostf3 have to grovel and snivel to get an answer there.   H    Too bad about the Linux worm, but I figured it would happen sooner or later.G Back in the 70's I researched into many systems for purchase.  It was aa? hard job to wade through all of the sales hype to get to what alE companies system can do.  The O/S was the very first thing to look atwG and then the software for it.  I finally settled on a pdp-11/45 then...:G and in the 80's settled on VMS on a Vax 785. One has to be careful whenc= the boss gives you a couple of million $ to go shopping with!v   ------------------------------  % Date: Mon, 22 Jan 2001 15:13:35 -0600G1 From: "Dave Gudewicz" <david.gudewicz@abbott.com>n: Subject: Re: Looking for Backup Scheduler for OpenVMS 7.218 Message-ID: <94i7ia$6rm$1@fizban.fizban.pprd.abbott.com>   Thanks Curtis.  K We are (or better put, will be) using ABS for VMS backups and we've had the L same experience as you so far and look forward to future versions, features, etc.   Dave...k  ? "Curtis Rempel" <vmsguy.no.spam.here@home.com> wrote in messageT4 news:v56a6.10794$K8.476800@news1.rdc1.ab.home.com...( > OK, here's what Curtis has to say  ;-) >tD > (actually, see another thread if you want to read my soapbox rant) >t7 > ABS Backup In A Nutshell (with apologies to O'Reilly)  > 9 > If you're using it for VMS backups, great, no problems.T > 9 > If you're gonna try UNIX, watch out for the land mines.  >tH > If you're gonna try NT with it, better start polishing up your resume. >b >k> > "Dave Gudewicz" <david.gudewicz@abbott.com> wrote in message4 > news:947qdf$5un$1@fizban.fizban.pprd.abbott.com...H > > I donno about  "anti".  We've tested the latest version (3.1) and it workss> > > for us.  I'd be interested to hear what Curtis has to say. > >n) > > Is it "perfect".  Nope.  Is anything?m > >mF > > We're aware of the scheduling issue and v3 provided the ability to > integrateh? > > external scheduling software, which we're lookin' into now.0 > >LG > > v3.1b of ABS is coming RSN followed by v4 sometime later this year.e > >p > > Dave...i > >  > >nA > > "andrew harrison" <andrew.nospam@uk.sun.com> wrote in messageo( > > news:3A670D36.EAA0808A@uk.sun.com... > > > Dave Gudewicz wrote: > > > >0F > > > > Archive Backup System (ABS) from Compaq might be worth a look. > > > >3D > > > > Client licenses for workgroup class systems are ~$400 a pop. > > > >2H > > > > Check http://www.openvms.compaq.com/openvms/storage/abspage.html > > > > for more details.? > > > > K > > > > We're just starting to implement it here, so not alot of experiencei > with > > it > > > > yet. > > > >e > > > > Dave...p > > > >e > > >/; > > > I would suggest you contact Curtis Rempel he seems toe8 > > > be distinctly unimpressed with ABS. He has renamed  > > > it the Anti Backup System. > > >e > > >o; > > > > "Tom Crabtree" <tccrab@sunset.net> wrote in messageo, > > > > news:3A652FE6.B077C362@sunset.net... > > > > > Hi all!-	 > > > > >aG > > > > > I'm fairly new to VMS, I only know enough to be dangerous, sot please > > be7 > > > > > kind and and understanding with my ignorance.-G > > > > > We have a small network of 5 Dec PWS 500AU's running Open VMSe 7.21.o > > (notJ > > > > > clustered)  Once a month I go around to each computer and load a > tapeK > > > > > and archive a data directory.  My collection of tapes is growing,r > andd > > my# > > > > > patience is wearing thin.IF > > > > > Where can I find an easy to use (and reasonably inexpensive)
 > softwareG > > > > > package that will allow me to set up a network wide scheduled  backupJ > > > > > from one node?  I'm thinking of something similar to Backup Exec > that > > > > > will work on VMS.hL > > > > > I know that a reasonably smart person can set up a scheduled batchK > > > > > process that will do this kind of job, but I'm not that smart andv > not F > > > > > that interested in becoming more than casually familiar with batch. > > > > > processes and DCL.	 > > > > >h > > > > > Any ideas?	 > > > > >r
 > > > > > TIAp	 > > > > >  > > > > > Tom C. > > > > > tccrab@sunset.netX > > > > > --K > > > > > -----------------------------------------------------------------5$ > > > > > My father used to tell me,H > > > > > "Son, there is nothing wrong with being scared. . . as long as? > > > > > you don't let it affect you until the danger is over. H > > > > > Being hysterical is okay, too . . . afterwards and in private.K > > > > > Tears are not unmanly . . . in the bathroom with the door locked.RE > > > > > The difference between a coward and a brave man is mostly aa > > > > > matter of timing"o	 > > > > >r: > > > > >       --------- Robert A. Heinlein -------------K > > > > > ------------------------------------------------------------------	 > > > > >c	 > > > > >R > > >  > > > -- > > > Andrew Harrison  > > > Enterprise IT ArchitectR > >  > >  >' >a   ------------------------------   Date: 23 Jan 2001 02:30:13 GMT' From: "Jim Strehlow" <jims@data911.com>S. Subject: Re: Lowercase filenames on VMS V7.2-10 Message-ID: <94iqbl$609@dispatch.concentric.net>  H I happened to encounter an ugly situation on OpenVMS Alpha with Oracle a couple of months ago.nK Someone created a table using Microsoft Access and used a Microsoft tool toiA export it directly to a new table in Oracle on the OpenVMS Alpha.RH Unfortunately, one column name out of 40 columns contained one lowercase
 character.  I Oracle internally converts everything to UPPERCASE ... except for the one-7 Microsoft tool that directly built the table in Oracle.nI Thus, no one could subsequently do anything with the data in that column.-D We had to drop the entire table as there was no way to get Oracle toC recognize a lowercase column name ... since it should never happen.e  1 That was for Oracle v8.05 on Alpha OpenVMS v7.1-2j  7 (We are probably just touching the tip of the iceberg.)s   Jim Strehlow, jims@data911.com& Senior Applications Analyst, DBA, etc.  B Rob Reffner wrote in message <3a64d906.2896922@news.cstone.net>...C >I have encounter a situation where lowercase filenames are causing.A >problems. I was not aware that VMS allowed this but I have founda? >several. Is there a way to prevent this and what is the cause?a >? >Thnaks in advance,  >s >Rob   ------------------------------    Date: 22 Jan 2001 16:38:43 -0500/ From: brooks@cuebid.zko.dec.nospam (Rob Brooks)8% Subject: Re: Multiple path devices...o- Message-ID: <KcT5UuW4vnQa@cuebid.zko.dec.com>h   > This unique solution? 8 	No, that is pretty much what multipath is designed for.  ! > And how to change primary path?f= 	You can't; as someone else noted correctly, it is simply theeD first path that was found when this device was configured.  The termE "primary" may be a bit misleading; when path verification occurs, thesI primary path isn't given any more weight than other paths when attemptingsD to chose which path to take.  Under certain circumstances, there mayC be an attempt to stay on the current path, which may be the primarySC path at the time, but it is because it is the current path, not thex. primary, that we may try to stay on that path.   ----M Rob Brooks    VMS Engineering -- I/O Exec Group     brooks!cuebid.zko.dec.comu   ------------------------------    Date: 23 Jan 2001 02:24:40 +0800, From: Paul Repacholi <prep@prep.synonet.com>* Subject: Re: New OpenVMS Times now on line- Message-ID: <87vgr7cumv.fsf@prep.synonet.com>h  " Paul Sture <paul@sture.ch> writes:  D > What???? Download IE over over a modem??? They've gotta be joking! > 0 > Who the hell are the clowns running this site?  6 Possible the same people who run Compaqs CSA web site?   -- r< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.:@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------    Date: 23 Jan 2001 02:17:15 +0800, From: Paul Repacholi <prep@prep.synonet.com>* Subject: Re: New OpenVMS Times now on line- Message-ID: <878zo3e9jo.fsf@prep.synonet.com>u  / goathunter@goatley.com (Hunter Goatley) writes:   B > On Fri, 19 Jan 2001 01:27:34 GMT, system@SendSpamHere.ORG (Brian! > Schenkenberger, VAXman-) wrote:I > >lJ > >Why is it everybody prints their .PDF files?  Why not just ship them in! > >postscript in the first place!f > >uD > I've become a PDF convert in the past few days.  I discovered thatD > (in the PC world) there are versions of TeX that produce PDF filesD > (pdftex and dvipdfm).  I have files that I produced using TeX thatC > I've never been able to give to PC users because they didn't know # > what to do with Postscript files.e  C Better still, ship the dvi ;) Or even better, the TeX or Doc files.e  ? Seriously, I have not found anything that does as good a job as > TeX, TeX fonts, and dvips. Even using type 1 CM fonts degrades the quality.  F > In my experience, the only people who routinely know what to do with2 > Postscript files are VMS people, sad as that is.  A Ironic really. The PS evangalists where the Mac brigade. RemembermA the LW1? PS is alive and well, and being still tuned and improveds$ in the printing and pre-press areas.   ...s  > > (As for viewing my PDF files on VMS, Ghostscript does a much > better job of it than Xpdf.)  D I've just installed GS 6.50. I can now read a pile of Compaq manualsD that would not work on earlier versions. Now I just need a VMS GV to go with it.    -- p< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.n@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------    Date: 23 Jan 2001 02:31:44 +0800, From: Paul Repacholi <prep@prep.synonet.com>* Subject: Re: New OpenVMS Times now on line- Message-ID: <87r91vcub3.fsf@prep.synonet.com>n  , WILLIAM WEBB <WWEBB1@email.usps.gov> writes:  G > Murphy's Law strongly suggests that important documentation should be / > accessible when your system is Tango Uniform.o >   > .PDF or .HTML is fine with me. > B > WWWebb (who is still quite partial to paper, but that's just me)   Or before it is running.  > My son coughted up for a 56K modem last year. It did not work.= Wouldn't dial out. OK, it is not recognising the dial tone ate? least some of the time, set it to blinf dial. Insert CD. Manual 5 is the dog. No Acrobat on his billybox. ( Games, OK )n  , So ring the support number, explain problem.  9 "Oh, you can download it for free from Adobe's web site."p  8 I kid you not... And no, he did not know the AT magic to set it to blind dial.a   -- r< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.a@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------    Date: 23 Jan 2001 02:21:53 +0800, From: Paul Repacholi <prep@prep.synonet.com>* Subject: Re: New OpenVMS Times now on line- Message-ID: <87zogjcuri.fsf@prep.synonet.com>   3 bill@triangle.cs.uofs.edu (Bill Gunshannon) writes:n  H > I am still trying to figure out why we needed a proprietary "Portable"F > Document Format when Postscript was already about as portable as you > could get.  ? AGREE! Forced obselescene. Just what we don't need. PS has it'seB 'issues' with portability ( Framemaker and US paper come to mind )" but at least it's turing complete.  ? Proprietry Dog Fuck has to be one of the better con-jobs pulleds on the industry.   -- e< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.y@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------    Date: 23 Jan 2001 02:18:39 +0800, From: Paul Repacholi <prep@prep.synonet.com>* Subject: Re: New OpenVMS Times now on line- Message-ID: <874ryre9hc.fsf@prep.synonet.com>r  ) carl@gerg.tamu.edu (Carl Perkins) writes:w  # > system@SendSpamHere.ORG writes... J > }Why is it everybody prints their .PDF files?  Why not just ship them in! > }postscript in the first place!nR > }VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM > E > Not everybody has a postscript printer. You don't need a postscript 9 > printer to print a PFD file (well, not on a PC anyway).   " So how does one print a pdf file?    -- e< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.a@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------  % Date: Mon, 22 Jan 2001 19:23:56 +0100   From: Paul Sture <paul@sture.ch>* Subject: Re: New OpenVMS Times now on line+ Message-ID: <VA.00000251.040f0a46@sture.ch>e  8 In article <01012209442653@beast.dtsw.army.mil>,  wrote:" > From: jamese@beast.dtsw.army.mil > Newsgroups: comp.os.vms , > Subject: Re: New OpenVMS Times now on line' > Date: Mon, 22 Jan 2001 09:44:26 -0500: > H > Paul Sture <paul@sture.ch> wrote in <VA.0000024b.2579af46@sture.ch> on" > Sun, 21 Jan 2001 21:12:36 +0100: > O > > I've found no way to print a straight Postscript document from a PC. It's, 9 > > ahem, frustrating. >  > Amen to that.i > J > What I do on my NT box is to go to the command prompt window, and enter: > # >     D:\> type filepath.ps >>lpt1:L >         or9 >     D:\> type filepath.ps >>\\server\printer_queue_namem > C > I have these at the top level of the disk as .BAT files, such as:o >  >     D:\> type hp4si.bat , >     type %1% >>\\server\printer_queue_name >  > then I can say:c >  >     D:\> hp4si filepath.ps > E > I can get duplex on our HP4Si, and others, is by another .BAT file:t >   >     D:\> type hp4si_duplex.bat4 >     perl -ple "BEGIN { print q(%%!),qq(\n),q(%%!),6 >     qq(\nstatusdict begin true setduplexmode end\n);& >     }" %1 >>\\server\hp4si_queuename > D > The .BAT file is actually one long line, wrapped here for mailing. > ! That look's really useful thanks.S ___s
 Paul Sture Switzerland    ------------------------------  # Date: Mon, 22 Jan 2001 23:33:15 GMT - From: goathunter@goatley.com (Hunter Goatley).* Subject: Re: New OpenVMS Times now on line1 Message-ID: <3a6cc250.116397761@swen.process.com>   E On 23 Jan 2001 02:17:15 +0800, Paul Repacholi <prep@prep.synonet.com>, wrote:  0 >goathunter@goatley.com (Hunter Goatley) writes: >hF >>> I've become a PDF convert in the past few days.  I discovered thatE >> (in the PC world) there are versions of TeX that produce PDF filesnE >> (pdftex and dvipdfm).  I have files that I produced using TeX that D >> I've never been able to give to PC users because they didn't know$ >> what to do with Postscript files. >sD >Better still, ship the dvi ;) Or even better, the TeX or Doc files. >,C 8-)  In this case, the documents were issues of a newsletter I usedmA to publish, which includes both text and images.  The issues were5D originally offset printed, with the images halftoned by the printer.  C With dvipdfm, I'm now able to distribute a printable file, with all: the images in place.  > I *wish* the world knew about TeX, but, alas, more people seem2 to have forgotten about TeX than Postscript, even.  @ >Seriously, I have not found anything that does as good a job as? >TeX, TeX fonts, and dvips. Even using type 1 CM fonts degrades 
 >the quality.i >hA The few pages I've printed from the dvipdfm PDF files look great.   G >> In my experience, the only people who routinely know what to do withi3 >> Postscript files are VMS people, sad as that is.s >oB >Ironic really. The PS evangalists where the Mac brigade. RememberB >the LW1? PS is alive and well, and being still tuned and improved% >in the printing and pre-press areas.a >e True.M   Hunter ------9 Hunter Goatley, Process Software, http://www.process.com/b9 goathunter@goatley.com     http://www.goatley.com/hunter/s   ------------------------------  # Date: Mon, 22 Jan 2001 23:37:34 GMTd- From: goathunter@goatley.com (Hunter Goatley) * Subject: Re: New OpenVMS Times now on line1 Message-ID: <3a6cc39e.116731991@swen.process.com>p  E On 23 Jan 2001 02:18:39 +0800, Paul Repacholi <prep@prep.synonet.com>r wrote:  * >carl@gerg.tamu.edu (Carl Perkins) writes: >d$ >> system@SendSpamHere.ORG writes...K >> }Why is it everybody prints their .PDF files?  Why not just ship them inj" >> }postscript in the first place!S >> }VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM  >> aF >> Not everybody has a postscript printer. You don't need a postscript: >> printer to print a PFD file (well, not on a PC anyway). >e# >So how does one print a pdf file? c >lE From the PDF viewer, you can select whatever printer drivers you have % installed: Postscript, PCL, FAX, etc.G  D Yes, you can do the same thing with a .PS file and Ghostscript, but,? as I noted in an earlier post, more Windows users are likely to$8 already have the Adobe Acrobat viewer installed than GS.   Hunter ------9 Hunter Goatley, Process Software, http://www.process.com/e9 goathunter@goatley.com     http://www.goatley.com/hunter/n   ------------------------------   Date: 23 Jan 2001 01:47:32 GMT From: clark@sander.stsci.edu ()e* Subject: Re: New OpenVMS Times now on line) Message-ID: <94inrk$rsp$1@tomm.stsci.edu>   $ SubjecNew OpenVMS Times - Hubble Add  @ Just looked at the new issue.  The last page really burns me up.  D Did you know that Hubble Space Telescope's top project is to get the/ picture generation off of VMS and onto Solaris?r  + That's right VMS is being replaced by theirf4 competitor's product ASP. To make matters even worse5 the recommendation to drop VMS was made by one of thec5 Space Telescope Science Institute's managers who alsob5 happens to have been one of the top members of DECUS,.# and a past strong supporter of VMS.1  7 Talk about false advertising. The HST project views VMSa5 as a dead platform. We can't even get Sybase database $ support anymore. Solaris wins again.   Kerry Clarkn CSCu! Space Telescope Science Institute    ------------------------------    Date: 22 Jan 2001 13:59:56 -05002 From: malmberg@eisner.decus.org (John E. Malmberg)= Subject: Re: No technical computing, was: expanding the niche + Message-ID: <LJb9yfptEcik@eisner.decus.org>u  , In article <94hnon$5g1@gap.cco.caltech.edu>,5  mathog@seqaxp.bio.caltech.edu (David Mathog) writes:0 > 8 > Replying in one place to a bunch of different replies: >a >>Keith Brown said: I >>At my site we happen to throw lots of controllers and disk at our DUNIX J >>systems as well as our NT and OpenVMS systems.  Your accertion that thisC >>is only done on VMS systems to increase IO performance is untrue.p >eN > I did not mean to imply that it was done ONLY to improve IO performance, butN > rather that in these _already_ large systems the slow intrinsic IO was beingJ > compensated for by extra hardware that would not have been required wereJ > the base VMS IO faster.  That is, in order to achieve a certain level ofH > performance on VMS you will probably have to add more storage hardwareF > than you might on other OS's.  Specific application example, let "1"? > be the file server speed that can be achieved with Samba on a D > Linux DS10 with an U2W controller and a single 27 Gb disk under anG > average (nonsaturating) load.  If you loaded VMS onto that exact same H > machine, and Samba as well (or Pathworks) you will be fortunate indeedA > to achieve .1 in file server performance in mixed reads/writes.a <snip>  C I assume that you did not do your tests on SAMBA 2.0.6 for OpenVMS.   ? Partly as a result of your previous postings, and to solve someo? issues I ran into, I think that I made some improvements there.w  B Unfortunately I only had time to test to make sure that everything; was functional before I had temporarily pack my stuff away.V  $ I do not have any benchmarking data.  D If you do get a chance to benchmark SAMBA 2.0.6 for OpenVMS, I thinkG you will see the best performance when transfering a FIXed format file.i  E IIRC: I put a default template to treat .EXE files as FIXed 512 bytes , when copied to OpenVMS from a LANMAN client.  I The OPLOCKs on SAMBA 2.0.6 may slow down the start of the copy as part ofc2 them are still being done with UNIX locking files.   -John- wb8tyw@qsl.network Personal Opinion Only-   ------------------------------    Date: 23 Jan 2001 03:33:05 +0800, From: Paul Repacholi <prep@prep.synonet.com>= Subject: Re: No technical computing, was: expanding the niche - Message-ID: <87itn7crgu.fsf@prep.synonet.com>   4 rdeininger@mindspring.com (Robert Deininger) writes:  ] > In article <87puhfeqwk.fsf@prep.synonet.com>, Paul Repacholi <prep@prep.synonet.com> wrote:  > 0 > > This is not yet fit for exposure yet, but... ...    > What was the actual working set for the VMS process?  What was WSEXTENT?  Also there's a system parameter involved, whose name I can't remember off the top of my head.  The point is, VMS likely borrowed working set as needed above 32MB.E > Make sure you are comparing similar situations between the two OSs.0  C WSextent was 65000. WSMAX, on looking seems to be 13000. That seemsL odd.    l > Also, was VMS racking up a lot of "soft" page faults?  If so, it wasn't really paging at all in the usual meaning of the word.  There's opportunity for mischief here, since the VMS system has more total memory that the oonix one.r  D True. as far as I could see, there was about 32M used. Few processesC where swapped. The system was 'normal', clustered, X, AMDS, Decnet,dE TCPIP, LAT. Nothing else busy, but it was my normal day to day setup.U# It was not cleaned out or anything.r   > C > > The test code is FFTW-TEST, version 1 from MIT. The standard DUbD > > generation was used. For VMS it was just thrown at the compiler! > > @ > > The top line give OS and size of data ( complex - complex ).= > > The data lines give the speed in MFLOPS for each subtest.f/ > > Each line is a run, most where run 5 times.r > > E > > There is no 3600x3600 DU number as it was to large for the kernel9 > > config.r > > F > > Yes, I KNOW these numbers are not 'optimum', and that there shouldF > > be al sorts of tunning done. How ever, I suspect that it IS prettyC > > close to the average for most boxes on researchers desks unlessh0 > > they have specialist support from their uni. > > @ > > Can you say paging?? Do not use for real work. You have been > > warned.- > >  > > VMS 3600x3600 Pe > > + > > 14.683201 14.465498 15.242671 15.426464k+ > > 24.153116 24.607297 24.137885 24.229562i+ > > 24.806644 24.698597 23.967847 24.714545Q+ > > 22.892734 24.046899 24.229562 24.540248i+ > > 22.397104 23.587522 24.532383 24.859004  > ^ > What do the 4 numbers on a line mean?  Why is the first line so much slower than the others?  A The FFTW-TEST runs 4 variants of the FFT. Each number is the ratenF in MFLOPS from the run. Note, the 'rate' is computed from the efective7 number of FLOPS, not by count. So the compiler matters.g  C I can't remember any reason for the first run being out. I may haveeA been doing a lot more monitoring and pocking. That was one of ther very early runs.   > > DU 1600x1600 P > > + > >  1.624575  1.846240  2.016829  1.905070 + > >  2.069622  1.935263  1.812045  1.785000t+ > >  2.097139  1.956157  1.933399  1.920702n+ > >  2.118218  1.722296  1.824570  1.814402.+ > >  1.905643  1.823318  1.834744  1.843191-+ > >  1.995768  1.677230  1.781479  1.830002a > >  > > VMS 1600x1600y > > + > > 34.491483 35.947588 40.367810 41.920418i+ > > 34.404384 37.480429 40.608453 41.791828c+ > > 34.978526 37.173631 40.070988 41.347909 + > > 33.933091 37.532055 40.851982 42.642053h > A > Ok, I guess you are saying DUNIX doesn't page as nicely as VMS.s@ > (Though I don't follow the details of your tests very well...)  < You can say that in spades. The results rocked my socks off!  < I expected VMS to do well under heavy paging, By a factor of? ( guessing ) 2 to 3. If anyone gets unix paging right, it woulda< be the DU folks. Like they can 'be inspired' by what the VMS* people know... Well that was the thinking.  : Also, if you look at the very first DU number. Top left of8 each set. They are consistantly lower, due to paging out> the stuff in memory. If you wait between runs, you can sandbag the first number of the line.J  ? The VMS system has 64 M in it at the moment, so I'll re-run thegC 1600x1600 test. I need to check wsmax too, that seems way too smallfB for how I usually run. And without a couple of Netscape windows...   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.t@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------  % Date: Mon, 22 Jan 2001 20:57:36 -0500f' From: "Bill Todd" <billtodd@foo.mv.com>r= Subject: Re: No technical computing, was: expanding the nicheo( Message-ID: <94io73$jpd$1@pyrite.mv.net>  = David Mathog <mathog@seqaxp.bio.caltech.edu> wrote in messageh& news:94hnon$5g1@gap.cco.caltech.edu...   ...l  ) > >Bill Todd" <billtodd@foo.mv.com> said:c > >fL > >Perhaps it's just due to frustration (which is something I can sympathizeD > >with), but in this post the bullshit far outweighs the substance. >iA > Everything I said is based on my own benchmarks which are quites0 > representative of the sorts of programs I run.  A Unfortunately, you chose to extrapolate and make statements abouteI enterprise-level systems which bear almost no resemblance to the sorts ofe programs you run.w   ...   C > >The fact that VMS disk-access-related performance *using default 	 settings* F > >lags drastically behind, say, Linux's has been more than adequately covered K > >in this forum.  But the leap from that to the assertion that VMS *can't*lK > >perform well when well-tuned is completely unfounded - just as unfoundedn asJ > >for VMS bigots to assert that Linux is intrinsically unsafe (especially whenI > >compared with VMS) because, by default, data gets written back to disk  onlyJ > >when Linux happens to feel like it, hence arbitrary amounts may be lost if > >power fails.t >oI > I will assert emphatically that on a DS10, using the benchmark programsoD > indicated above, VMS *cannot* perform as well as Linux on the sameE > hardware, no matter how you tune it, at least if it's written in C.e  J To say that certain programs won't run as fast (in their existing form) onF VMS is a much different statement from your generalization that VMS isJ intrinsically 'hideously slow'.  I suspect that I can, for example, easilyG write a program that cannot run as fast on Linux as it does on VMS if IdH choose to force all writes to disk and use block-level (through RMS, letL alone QIO-level) access (at least IIRC that Linux doesn't offer a reasonable4 way to avoid copying data through its system cache).    ForF > programs that are completely IO bound, and record oriented, the 2-3XL > slowness of VMS/RMS/ Compaq C record IO compared to that of Linux/Compaq C  K > (same compiler, same language, same hardware, only the OS differs) cannotrI > be overcome by any amount of RMS tuning. Specifically, on an IO limitednI > program VMS running RAMdisk->RAMdisk will run no faster than 1/3 to 1/2. thefB > speed of the same code running on Linux file cache-> file cache.  H I can believe that:  RMS does add a layer that Linux integrates into itsK normal file system, likely with better CPU efficiency.  Of course, RMS alsonL adds a great deal of function that Linux does not have, whether or not thoseK features may be useful to you (if they're not, there's always QIO, which istI arguably more comparable to the level of features in Unix, but admittedly H far less easy to use:  the Unix interface level *is* a good choice for a great many applications).   : OTOH, the above, if I read it correctly, is somewhat of anI apples-and-oranges comparison, in that the VMS code path runs through the.G RAMDISK device level while the Linux code never delves beneath the filerG system level:  a more equal comparison would be with something like RMSsK global buffers (assuming that they can be re-read before being written backuH to disk, which I don't know), and I suspect would lessen the difference.  H Oh, yes:  a 2- or 3-to-1 difference (ignoring the apples/oranges issue),K while significant, is far from the 10-to-1 figure you were throwing around.A  D The bottom line is that you're still confusing default behavior withL attainable behavior.  I agree whole-heartedly that VMS's default behavior isH inappropriate for a many applications and drastically needs improvement.J But VMS is flexible and efficient enough to make it *possible* to get very" good performance when you need it.   - bill   ------------------------------  % Date: Mon, 22 Jan 2001 22:14:16 -0600W% From: Keith Brown <kbrown780@isd.net>-= Subject: Re: No technical computing, was: expanding the niche0' Message-ID: <3A6D0518.A850990A@isd.net>j   David Mathog wrote:m > c > In article <94aa7u$4v5@gap.cco.caltech.edu>, mathog@seqaxp.bio.caltech.edu (David Mathog) writes:y > 8 > Replying in one place to a bunch of different replies: >  > >Keith Brown said:J > >At my site we happen to throw lots of controllers and disk at our DUNIXK > >systems as well as our NT and OpenVMS systems.  Your accertion that this D > >is only done on VMS systems to increase IO performance is untrue. > N > I did not mean to imply that it was done ONLY to improve IO performance, butN > rather that in these _already_ large systems the slow intrinsic IO was beingJ > compensated for by extra hardware that would not have been required wereJ > the base VMS IO faster.  That is, in order to achieve a certain level ofH > performance on VMS you will probably have to add more storage hardwareF > than you might on other OS's.  Specific application example, let "1"? > be the file server speed that can be achieved with Samba on alD > Linux DS10 with an U2W controller and a single 27 Gb disk under anG > average (nonsaturating) load.  If you loaded VMS onto that exact same!H > machine, and Samba as well (or Pathworks) you will be fortunate indeedG > to achieve .1 in file server performance in mixed reads/writes.  ThatwD > is, the average elapsed time for each operation would be about 10XG > longer on VMS.  (For the moment assume that the UPS on your Linux box,H > constitutes sufficient data integrity insurance, and that you can SYNCG > all disks before its battery goes).  How much extra hardware must you7G > add to the identical DS10, once it's running VMS, to achieve the samepA > level of file server performance?   On a very large system thisoG > "extraness" will be less evident, because you're already starting outIG > with RAID arrays with built in (large) cache and battery backup.   ItdH > could be that if you choose the right storage solutions they'll behave> > more or less memory (ie, they will effectively implement theC > equivalent of file caching), and then you'll do no worse than the F > intrinsic 2-3X performance factor that RMS offers (versus Linux) for+ > file IO which is really memory to memory.e > I > >----------------------------------------------------------------------w) > >Bill Todd" <billtodd@foo.mv.com> said:m > >hL > >Perhaps it's just due to frustration (which is something I can sympathizeD > >with), but in this post the bullshit far outweighs the substance. > A > Everything I said is based on my own benchmarks which are quitePG > representative of the sorts of programs I run.  If you feel that theyoC > were deficient in some way feel free to point out the error in myg > programs, they are at: > > >   http://seqaxp.bio.caltech.edu/pub/SOFTWARE/MYBENCHMARK.ZIP > H > If some scaling factor takes over on larger systems that renders theseM > benchmarks invalid that would be interesting.  So I invite everybody with aSF > spare GS$$$ and a few spare moments to run the benchmarks and see ifE > you're big expensive machine can run them any faster than my small,s! > inexpensive (sort of) DS10 can.- >  > >rM > >The fact that VMS disk-access-related performance *using default settings* N > >lags drastically behind, say, Linux's has been more than adequately coveredK > >in this forum.  But the leap from that to the assertion that VMS *can't*uN > >perform well when well-tuned is completely unfounded - just as unfounded asO > >for VMS bigots to assert that Linux is intrinsically unsafe (especially whencN > >compared with VMS) because, by default, data gets written back to disk onlyM > >when Linux happens to feel like it, hence arbitrary amounts may be lost if  > >power fails.j > I > I will assert emphatically that on a DS10, using the benchmark programsbD > indicated above, VMS *cannot* perform as well as Linux on the sameI > hardware, no matter how you tune it, at least if it's written in C. ForFF > programs that are completely IO bound, and record oriented, the 2-3XL > slowness of VMS/RMS/ Compaq C record IO compared to that of Linux/Compaq CK > (same compiler, same language, same hardware, only the OS differs) cannotwI > be overcome by any amount of RMS tuning. Specifically, on an IO limitediM > program VMS running RAMdisk->RAMdisk will run no faster than 1/3 to 1/2 thenB > speed of the same code running on Linux file cache-> file cache. > F > If you can provide an IO bound program, written in ANSI C, that doesL > better than that, RAMdisk->RAMdisk vs. filecache -> filecache, I'd be more > than happy to see it.u >  > >:I > >---------------------------------------------------------------------- ? > >Paul Repacholi                               1 Crescent Rd.,i > > D > >Sorry David, but I feel you are at least 99% wrong. And the tests# > >I've done come out very pro-VMS.  > H > The test you posted in the followup note was very unclear to me.  WereB > you measuring the behavior of VMS *paging* versus Unix *paging*?C > That's not at all the problem I was addressing.  If you're seeingeB > paging on Unix (linux in particular) it means that the amount ofK > RAM available for file cache has been reduced to zero.  At that point VMSaF > may well do better if for no other reason than it has locked up it'sM > /buffer/block memory and that will not evaporate when the free memory pages D > go to zero.  The "solution" for Unix is, well, just don't do that. >  > >-H > >If you are running a big tech job and it is doing lots of IO, then it( > >is time to rethink. Think P2 in fact. > - > I don't understand what you are suggesting.g >  > >bD > >It is half solved. The galaxy work, and more importantly, the new, > >MM layout on Alpha is a huge step foward. > E > Galaxy isn't going to do anything for small systems.  My DS10 seemseG > not to be running any faster since the GS models came out.  Upgradingt< > from a $5k system to a $250k system is also not an option! >  > >The lagardly part isrE > >the software. The LACK of VMS push in unis and Sci computing meanseC > >that near NO app start on, or are written with any considerationiF > >of VMS. So many things that VMS will do for the app get re-inventedE > >many different times over. So you are faced with a major re-design ( > >or band-aiding over the brain-damage. > >aN > >(Benchmarks - don't know what the program is doing or what's rate limiting.0 > >My comments referred to IO only, not paging.) > G > I know that my benchmarks do nothing but IO.  What exactly were yourst > doing? > 
 > Regards, >  > David Mathog > mathog@seqaxp.bio.caltech.edu @ > Manager, sequence analysis facility, biology division, CaltechL > **************************************************************************L > *                                RIP VMS                                 *L > **************************************************************************  	 Hi David,   H You obviously have spent much more time bench marking these two machinesH than I have. I did want to note one thing however. We compared IO on VMS@ and DUNIX with the HSZxx writeback cache turned off (just out ofG curiosity) and we found that IO performance tanks with either OS. Maybea> I've misinterpreted what you were saying but I expected betterA performance on the DUNIX box based on your comments and benchmark  results.     -- s Keith Brown  kbrown780@isd.netF   ------------------------------   Date: 22 Jan 2001 19:50 CSTE' From: carl@gerg.tamu.edu (Carl Perkins)o Subject: Re: Oldtimer forgot! - Message-ID: <22JAN200119502578@gerg.tamu.edu>s  0 Howard S Shubs <hshubs@mindspring.com> writes...4 }rdeininger@mindspring.com (Robert Deininger) wrote:C }>Each of the file names can be a simple name, a logical name that eJ }>translates to a file, or a full name with device, directory, file name,  }>etc. } I }Another way, if he's going to do this often, is to use a linker OPTions .( }file.  See the Linker docs for details. }--   3 Another way that saves a lot of typing is to defineq7 LNK$LIBRARY logical names. The first one is, of course,e9 just LNK$LIBRARY but you can also have LNK$LIBRARY_1, _2,t: etc. I have no idea what the upper limit o n the number is< but I know it is at least 6 since I've got some defined that, high (I would assume that it is at least 9).  + These are automatically used by the linker.o   --- Carl   ------------------------------  % Date: Mon, 22 Jan 2001 21:46:34 -0200o1 From: "Valdemir J. Santos" <valdemir-@uol.com.br>A# Subject: Printing problems in TCPIP 3 Message-ID: <000a01c084cd$8d9e5560$d2010001@desenv>o  , This is a multi-part message in MIME format.  + ------=_NextPart_000_0007_01C084BC.C9C93A20  Content-Type: text/plain;c 	charset="iso-8859-1"S+ Content-Transfer-Encoding: quoted-printableo  3 In my job we use TCPIP under OpenVMS V7.2-1, but=20M, we are experience troubles printing jobs.=20, When I send a job to the queue, I have this:    G Server queue PR$CMP_UNP_SCR_3, busy, on ALPHA1::, mounted form CART_132s   <32.0.0.17 >  4   Entry  Jobname         Username     Blocks  StatusD   -----      -------               --------           ------       = ------8    2177  NOTAS_33        PR_UNPCMP_SC      3  Processing7       530  CAD4180X        PR_UNPCMP_SC      1  Pendingn5    2510  NOTAS_33        PR_UNPCMP_SC      3  Pendingh5    2562  NOTAS_33        PR_UNPCMP_SC      3  Pendingr5    2583  NOTAS_33        PR_UNPCMP_SC      3  Pendingk5    3359  NOTAS_50        PR_UNPCMP_SC      3  Pendingr   =20o  :   Jobs stay in Processing status, but nothing is printed !  =    The only way to release these jobs is reseting the printere  =20    (Lexmark Optra 1255).  @    What can I to do to solve this problem ? Is there any command  H    in TCPIP to analyse the network ? Is there any log file to analyse=20      the printer under TCPIP ?  @    I have 200 printer under TCPIP in my job, and we have much=20      troubles like this !2  %    I use this command to init queues:    $ INIT/QUEUE/START-3 $           /OWNER=3D[1,4]-4 $           /DEFAULT=3DFLAG- $           /RETAIN=3DERROR- $           /DEVICE=3DSERVER-n% $           /PROCESS=3DTCPIP$LPD_SMB-  $           /FORM=3DCART_132-S* $           /DESCRIP=3D""''IP_ADDRESS'""-"< $           /PROTECTION=3D(S:E,G:R,W:W,O:D) ''SERVER_QUEUE'"    It=B4s right ?e        Thank you in antecipate...           + ------=_NextPart_000_0007_01C084BC.C9C93A20o Content-Type: text/html; 	charset="iso-8859-1"g+ Content-Transfer-Encoding: quoted-printableo  > <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> <HTML><HEAD>3 <META content=3D"text/html; charset=3Diso-8859-1" =2 http-equiv=3DContent-Type>9 <META content=3D"MSHTML 5.00.2919.6307" name=3DGENERATOR>1 <STYLE></STYLE>. </HEAD>9 <BODY bgColor=3D#ffffff>G <DIV><FONT face=3DArial size=3D2>In my job we use TCPIP under OpenVMS =6 V7.2-1, but=20
 </FONT></DIV>5F <DIV><FONT face=3DArial size=3D2>we are experience troubles printing = jobs.=20
 </FONT></DIV>3J <DIV><FONT face=3DArial size=3D2>When I send a job to the queue, I have=20 this:</FONT></DIV> <DIV>&nbsp;</DIV>tE <DIV><FONT face=3DArial size=3D2><BR>Server queue PR$CMP_UNP_SCR_3, =t busy, on=20l9 ALPHA1::, mounted form CART_132<BR>&nbsp; &lt;32.0.0.17 =o &gt;</FONT></DIV>t <DIV>&nbsp;</DIV>n6 <DIV><FONT face=3DArial size=3D2>&nbsp; Entry&nbsp;=20: Jobname&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20A Username&nbsp;&nbsp;&nbsp;&nbsp; Blocks&nbsp; Status<BR>&nbsp;=20h! -----&nbsp;&nbsp;&nbsp;&nbsp; =20u: -------&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20! &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20aB --------&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; &nbsp;&nbsp;=20J ------&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;  ------<BR>&nbsp;&nbsp; 2177&nbsp;=205 NOTAS_33&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20 J PR_UNPCMP_SC&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 3&nbsp; Processing<BR>&nbsp;=20 &nbsp;&nbsp;&nbsp; 530&nbsp; =5 CAD4180X&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20 4 PR_UNPCMP_SC&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1&nbsp; = Pending<BR>&nbsp;&nbsp;=20@ 2510&nbsp; NOTAS_33&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=204 PR_UNPCMP_SC&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 3&nbsp; = Pending<BR>&nbsp;&nbsp;=20@ 2562&nbsp; NOTAS_33&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=204 PR_UNPCMP_SC&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 3&nbsp; = Pending<BR>&nbsp;&nbsp;=20@ 2583&nbsp; NOTAS_33&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=204 PR_UNPCMP_SC&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 3&nbsp; = Pending<BR>&nbsp;&nbsp;=20@ 3359&nbsp; NOTAS_50&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=204 PR_UNPCMP_SC&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 3&nbsp; = Pending<BR>&nbsp;&nbsp;=20
 </FONT></DIV>b <DIV>&nbsp;</DIV>hI <DIV><FONT face=3DArial size=3D2>&nbsp; Jobs stay in Processing status, =c but nothing=20 is printed !</FONT></DIV>  <DIV>&nbsp;</DIV> G <DIV><FONT face=3DArial size=3D2>&nbsp;&nbsp; The only way to release =e these jobs is=20! reseting the printer</FONT></DIV> 5 <DIV><FONT face=3DArial size=3D2>&nbsp; </FONT></DIV>f> <DIV><FONT face=3DArial size=3D2>&nbsp;&nbsp; (Lexmark Optra = 1255).</FONT></DIV>s <DIV>&nbsp;</DIV>,I <DIV><FONT face=3DArial size=3D2>&nbsp;&nbsp; What can I to do to solve =e this problem=20o# ? Is there any command</FONT></DIV>p <DIV>&nbsp;</DIV>lG <DIV><FONT face=3DArial size=3D2>&nbsp;&nbsp; in TCPIP to analyse the =t network ? Is=20L+ there any log file to analyse </FONT></DIV>n <DIV>&nbsp;</DIV>VG <DIV><FONT face=3DArial size=3D2>&nbsp;&nbsp; the printer under TCPIP =i ?</FONT></DIV> <DIV>&nbsp;</DIV>aH <DIV><FONT face=3DArial size=3D2>&nbsp;&nbsp; I have 200 printer under = TCPIP in my=20# job, and we have much </FONT></DIV>m <DIV>&nbsp;</DIV>vB <DIV><FONT face=3DArial size=3D2>&nbsp;&nbsp; troubles like this = !</FONT></DIV> <DIV>&nbsp;</DIV>fJ <DIV><FONT face=3DArial size=3D2>&nbsp;&nbsp; I use this command to init =   queues:</FONT></DIV> <DIV>&nbsp;</DIV>r< <DIV><FONT face=3DArial size=3D2>$ INIT/QUEUE/START-<BR>$=208 &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; = /OWNER=3D[1,4]-<BR>$&nbsp;=20n. &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;=20J /DEFAULT=3DFLAG-<BR>$&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs= p;&nbsp;=20eJ /RETAIN=3DERROR-<BR>$&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs= p;&nbsp;=20 B /DEVICE=3DSERVER-<BR>$&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; = &nbsp;&nbsp;=20rJ /PROCESS=3DTCPIP$LPD_SMB-<BR>$&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&= nbsp;&nbsp;&nbsp;=20J /FORM=3DCART_132-<BR>$&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=J sp;&nbsp;&nbsp;/DESCRIP=3D""''IP_ADDRESS'""-"<BR>$&nbsp;&nbsp;&nbsp;&nbsp=( ;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20A /PROTECTION=3D(S:E,G:R,W:W,O:D) ''SERVER_QUEUE'"<BR></FONT></DIV>oB <DIV><FONT face=3DArial size=3D2>&nbsp;It=B4s right ?</FONT></DIV> <DIV>&nbsp;</DIV>e <DIV>&nbsp;</DIV>a= <DIV><FONT face=3DArial size=3D2>&nbsp;&nbsp; Thank you in=20e antecipate...</FONT></DIV> <DIV>&nbsp;</DIV>f <DIV>&nbsp;</DIV>x <DIV>&nbsp;</DIV>t <DIV>&nbsp;</DIV></BODY></HTML>   - ------=_NextPart_000_0007_01C084BC.C9C93A20--,   ------------------------------  % Date: Tue, 23 Jan 2001 13:13:47 +1100a/ From: "Phil Howell" <phowell@snowyhydro.com.au>f' Subject: Re: Printing problems in TCPIP 1 Message-ID: <LK5b6.4020$cF2.82513@ozemail.com.au>i  0 In my job we use TCPIP under OpenVMS V7.2-1, but) we are experience troubles printing jobs.h, When I send a job to the queue, I have this:    G Server queue PR$CMP_UNP_SCR_3, busy, on ALPHA1::, mounted form CART_132n   <32.0.0.17 >  4   Entry  Jobname         Username     Blocks  StatusG   -----     -------               --------           ------      ------n8    2177  NOTAS_33        PR_UNPCMP_SC      3  Processing7       530  CAD4180X        PR_UNPCMP_SC      1  Pending,5    2510  NOTAS_33        PR_UNPCMP_SC      3  Pending 5    2562  NOTAS_33        PR_UNPCMP_SC      3  Pendinge5    2583  NOTAS_33        PR_UNPCMP_SC      3  Pendingv5    3359  NOTAS_50        PR_UNPCMP_SC      3  Pendingo    :   Jobs stay in Processing status, but nothing is printed !  =    The only way to release these jobs is reseting the printeri      (Lexmark Optra 1255).  ' I've never used these printers but ....   @    What can I to do to solve this problem ? Is there any command  $    in TCPIP to analyse the network ?  = Use tcpiptrace to identify traffic between the 2 ip addressesi ($ help tcpiptrace)6!  Is there any log file to analysea      the printer under TCPIP ?  F The location of logfiles for printers is defined in your printcap file ($ show log tcpip$printcap)i  =    I have 200 printer under TCPIP in my job, and we have muchs      troubles like this !e  %    I use this command to init queues:t   $ INIT/QUEUE/START-a $           /OWNER=[1,4]-S $           /DEFAULT=FLAG- $           /RETAIN=ERROR- $           /DEVICE=SERVER-e# $           /PROCESS=TCPIP$LPD_SMB-  $           /FORM=CART_132-e( $           /DESCRIP=""''IP_ADDRESS'""-": $           /PROTECTION=(S:E,G:R,W:W,O:D) ''SERVER_QUEUE'"  
  Its right ?   $ You "should" use  tcpip$lprsetup.exe  > Try to ping the printer's network card by address $ tcpip ping <nnn.nnn.nnn.nnn>i then by name $tcpip ping <name>aK Stop/reset the queue and try to telnet to the printer and enter "/" to showg
 configurations?  if you get connection refused then something else is using it.e; Reset the printer and try increasing the timeout parameter?n9 Are any non-vms jobs going to this printer by any chance?  Phil   ------------------------------  # Date: Tue, 23 Jan 2001 02:56:29 GMTS* From: martin.hunt@nbnz.co.nz (Martin Hunt)* Subject: Re: Queue has lowercase attribute0 Message-ID: <3a6cf20d.722940533@news.xtra.co.nz>  1 On Fri, 19 Jan 2001 16:42:06 +1100, "Phil Howell"a" <phowell@snowyhydro.com.au> wrote:  E >If you stop the queue you can do a $set queue /nolowercase <quename>d  F I can't see anything about that command in help, but it used to exist.  4 >(but you might not be able to change it back again)5 >I always thought that "lowercaseness" was associatedt) > with the terminal device not the queue.e  B I would've thought so, too. But this shows as a queue attribute. IC wouldn't have thought that there would be much used for the commandn7 now, with any uppercase-only printers being in museums.   9 >What type of queues are thay (printer/server/terminal) ?o   Terminal queues, using lat.l  % >are they using dcps/ucx/telnet/lat ?oB >If  dcps then have a look at sys$startup:dcps$execution_queue.com >Philr6 >Martin Hunt <martin.hunt@nbnz.co.nz> wrote in message+ >news:3a675991.356224263@news.xtra.co.nz... I >> I have a print queue which shows the "Lowercase" attribute with a SHOW I >> QUEUE/FULL command. I can't find the qualifier to remove (or add) thist4 >> attribute. I think it used to be INIT/QUEUE/LOWERH >> I need to either remove this attribute or add it to another queue, soG >> that ASSIGN/QUEUE will work between these two queues. I am currentlysF >> getting the message "inconsistent queue type" when I try using thisD >> command (need to use ASSIGN/QUEUE while one printer has a fault). >> --h >> Martin Hunt >> Technical Specialistt >> National Bank of New ZealandZ >  >    -- Martin Hunta Technical Specialist National Bank of New Zealand   ------------------------------  # Date: Tue, 23 Jan 2001 03:03:07 GMTw* From: martin.hunt@nbnz.co.nz (Martin Hunt)* Subject: Re: Queue has lowercase attribute0 Message-ID: <3a6cf3a4.723346677@news.xtra.co.nz>  F On Tue, 23 Jan 2001 02:56:29 GMT, martin.hunt@nbnz.co.nz (Martin Hunt) wrote:  2 >On Fri, 19 Jan 2001 16:42:06 +1100, "Phil Howell"# ><phowell@snowyhydro.com.au> wrote:L >xF >>If you stop the queue you can do a $set queue /nolowercase <quename> >rG >I can't see anything about that command in help, but it used to exist.e >s5 >>(but you might not be able to change it back again)S6 >>I always thought that "lowercaseness" was associated* >> with the terminal device not the queue. >bC >I would've thought so, too. But this shows as a queue attribute. I D >wouldn't have thought that there would be much used for the command8 >now, with any uppercase-only printers being in museums.  E I tell lies, sorry. I have just checked again, and found that all ourl@ print queues show as having the lowercase attribute. I must haveB missed it before, possibly because it was appearing in a different position on the line.t  B So, what would prevent ASSIGN/QUEUE being accepted? Generic queuesD being directed to the source queue can cause problems, but I tracked all those done.r   >n: >>What type of queues are thay (printer/server/terminal) ? >o >Terminal queues, using lat. >u   -- Martin Hunti Technical Specialist National Bank of New Zealand   ------------------------------  % Date: Mon, 22 Jan 2001 21:40:08 -0600n7 From: "David J. Dachtera" <djesys.nospam@earthlink.net>s* Subject: Re: Queue has lowercase attribute- Message-ID: <3A6CFD18.9390451F@earthlink.net>y   Martin Hunt wrote: > [snip]D > So, what would prevent ASSIGN/QUEUE being accepted? Generic queuesF > being directed to the source queue can cause problems, but I tracked > all those done.   D Can you post the output of SHOW QUEUE/FULL for both the queue you'reA trying to redirect and the queue you're trying to redirect it to?r   -- a David J. Dachterao dba DJE Systemse http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/a  F This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings is to be expected.  @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.o   ------------------------------  % Date: Tue, 23 Jan 2001 15:41:44 +1100,/ From: "Phil Howell" <phowell@snowyhydro.com.au>r* Subject: Re: Queue has lowercase attribute1 Message-ID: <rV7b6.4109$cF2.78709@ozemail.com.au>n  5 Martin Hunt <martin.hunt@nbnz.co.nz> wrote in messageg* news:3a6cf20d.722940533@news.xtra.co.nz...3 > On Fri, 19 Jan 2001 16:42:06 +1100, "Phil Howell" $ > <phowell@snowyhydro.com.au> wrote: >yG > >If you stop the queue you can do a $set queue /nolowercase <quename>b >uH > I can't see anything about that command in help, but it used to exist. >e6 > >(but you might not be able to change it back again)7 > >I always thought that "lowercaseness" was associatedn+ > > with the terminal device not the queue.N >pD > I would've thought so, too. But this shows as a queue attribute. IE > wouldn't have thought that there would be much used for the command 9 > now, with any uppercase-only printers being in museums.  >d; > >What type of queues are thay (printer/server/terminal) ?k >  > Terminal queues, using lat.a >  With lat I think you have to :-f stop the queue set the device as /nospooled- set the device to /lowercase or  /nolowercaset set it back to /spooled  start the queue againa# ps. I have never used assign /queuec$ as a work-around for broken printersH I define a logical name for the broken queue to point to a working queue# $ define/sys/exec printera printerb*& It should work (application dependent) Phil   ------------------------------    Date: 22 Jan 2001 18:59:29 -07001 From: nothome@spammers.are.scum (Malcolm Dunnett)d1 Subject: Re: RA230+ (KZPAC-CA) and 18/36 GB Disks , Message-ID: <3ialjoyyb7iH@malvm1.mala.bc.ca>  + In article <3A6C2EDB.C984A3E7@compaq.com>,  1    Hans Bachner <Hans.Bachner@compaq.com> writes:r >> fK >> I have a few questions about the older Raid Array 230+ that Digital sold J >> a few years ago.  I believe the 3-channel one was KZPAC-CB, it is a PCI4 >> device and is supported in at least OpenVMS v7.1. > V6.2-1H3 and V7.1 (and later)  > P > The 32 GB limit is a limit for the *logical* volumes. You can always put sevenN > 9 GB disks into a single BA356 and create a RAID 5 array with a net capacityM > of ca. 54 GB. You then need to create two logical partitions of, say, 27 GBsO > each (though the partitions may be different in size - the only point to take K > care of is that a single logical partition can't exceed the 32 GB limit).l > D     Clever workaround. Any idea why Compaq/Digital never bothered toG release V3.x firmware for the controller though? I've got it on genericTC DAC 960s ( on Prioris servers ) and it's quite happy to create 90GB 1 logical volumes ( 3 45GB disks in a Raid 5 set ).R  E     You'd think they'd want their "enterprise" operating system to bec@ able to support storage volumes at least as well as a PC can :-)   ------------------------------  % Date: Mon, 22 Jan 2001 22:17:16 +0000 ) From: Christof Brass <brass@infopuls.com>iW Subject: Re: Rehash of Andrew's Java on OpenVMS FUD.  (was: Re: GS160 hardwarequestion)-, Message-ID: <3A6CB16C.1358CF9F@infopuls.com>   Jordan Henderson wrote:e< > I wouldn't have said it was a little out of date had I not= > known about the existence of 1.3, would I have?  Of course,e: > I still contend that 1.2 is just a "little out of date".9 > The gross majority of existing Java commercial softwaret > still supports it. > 8 > If we're talking about inaccuracies, you just recently: > accused me of saying that 1.2 was the most recent, which > you've been unable back up.s >   = I'm familiar with the Java language and standards developmentP since 1995._@ It is true that 1.2 is the most important version. VisualAge for> Java has only recently been upgraded to 1.2. A lot of software? is happy with even earlier versions. 1.3 is a no-issue compared = with the lot of vapor "standards" issued by SUN like J2EE. If = you read that and look around who is supporting that you willT; recognise that there isn't anything around which comes even > close to that. SUN learned a lot from Micro$oft in introducing= the future by paperwork. As the only good thing in Windows isP9 that it lowered and continues to lower the prices of UNIXt> systems and SW, the major good thing in Java and the JVM is to? have something to fight Micro$oft. Technically speaking the JVM ; approach is very outdated, it has been done by the UCSD andm@ adopted by Apple in the late seventies. The most painful fact is= that since at least 1990 better techniques have been publicly = around. It is suprising how far SUN came with this oldy. Javae= itself is an oldy also. It is not innovative as it introducesN@ not a single new idea and the way the different things have been9 put together is far from beeing elegant. Anyway the majorC@ backlash is already going on. The Java hype is coming to the end= despite the fact that IBM puts a lot of money into Java which > will ensure that Java will stay and play an important role but+ it will soon loose its pacemaking position.a   ------------------------------    Date: 22 Jan 2001 17:24:24 -0500/ From: jordan@lisa.gemair.com (Jordan Henderson)nW Subject: Re: Rehash of Andrew's Java on OpenVMS FUD.  (was: Re: GS160 hardwarequestion)n* Message-ID: <94ibuo$mgb$1@lisa.gemair.com>  * In article <3A6C7B09.73FED248@uk.sun.com>,2 andrew harrison  <andrew.nospam@uk.sun.com> wrote: >Jordan Henderson wrote: >> S: >> >Either you were b******TING when you said that OpenVMS7 >> >was a "little late" or you didn't know about 1.3 inn7 >> >which case you were just stating an opinion withoutD >> >being equipped to make one.t >> > >> n5 >> I claim I did know about 1.3.  You claim I didn't.a6 >> Seeing as other vendors hadn't come out with 1.3 at4 >> the time, I felt justified in calling it a little >> late. >> 2 >R0 >I know what you claim, the problem is your past- >postings do not support this claim and your 21 >allegation that OpenVMS's JVM support is only a  
 >little late.= >r  ) This is bizaare.  If you review what was b0 ACTUALLY SAID at the time, you see quite clearly" that I was very well aware of 1.3.  4 In the article <8sfeif$l3$1@lisa.gemair.com>, we see, (AH> Andrew Harrison, JH> Jordan Henderson):  ; AH>>Umm well yes and no. The devil is in the detail and it -7 AH>>all depends on the Java implimentation on OpenVMS. b AH>0 JH> > JH> This is an example of how Andrew likes to spread FUD here. JH> K JH> In no way can you say that the 1.2 Java that's available for OpenVMS isnI JH> NOT Java, but that's exactly what he's saying above.  In response to 5K JH> the statement "It's Java" he says "yes and no", which implies that "no"_ JH> is in some way true. JH> J JH> So, Andrew has succeeded in spreading Fear Uncertainty and Doubt about JH> using Java on OpenVMS. JH> K JH> Of course, Compaq has stated their commitment to Java, cross platforms, I JH> but is a little behind Sun (surprise surprise) and IBM.  I wonder how_L JH> many _deployed_ applications require Java 1.3 today?  Probably not many.L JH> In the meantime, you can certainly develop on the Compaq platforms while JH> they Compaq plays catch up.=  @ This is where this statement that you insist is so inaccurate isA introduced.  Here, I say that Compaq is a "little behind" (not a s$ "little late" as we've been saying).  @ Note the reference to Java 1.3?  Note that I didn't say anything= about 1.2 being the most up-to-date as you earlier contended?     1 >You are trying to have both when for you its an O >either or question. >&2 >Either you knew that 1.3 was the current release 0 >in which case you also new that the little late2 >claim was untrue or you did not in which case the2 >little late claim is just a new definition on the1 >word little. Over 12 months after 1.2.X had beenD0 >made available on other plaforms is not little. >I  5 No Andrew, I both knew that 1.3 was the newest, and Ia6 contended at the time (see above) that it was not that far behind.2  : In computer programming languages, compliance to standards9 it's just not that typical for everyone to be at the sameV9 standard instantly.  Only someone trying to make a brighte8 line distinction for marketing purposes would think that7 a year behind the most current standards is a big deal.t  8 The C9X language standard was adopted almost 2 years ago5 and I don't believe (I could be wrong), that there is3; _any_ fully compliant compiler available today.  Of course,b; there's not a huge marketing machine behind C9X blaring outs9 recriminations at those who fail to live up to the latestp" standard, like there is with Java.  6 I still challenge you to show where commercial off the% shelf software widely requires 1.3.  N  5 >> You however keep sidestepping the point here.  Youp5 >> did say that I claimed that 1.2 was the latest.  Ap >> clear innacuracy. >> b >&3 >You don't get it do you, you didn't claim directly&8 >that 1.2.X was the latest release but you demonstrated 6 >that you were unaware it was by saying fact that the : >1.2.x had just made it onto OpenVMS ment that JVM support# >on OpenVMS was only a little late.s >=0 >Now you contend that you did know that 1.3 was - >the latest release, which if true ment that =0 >you knew that your "little" statment was untrue, >when you made it. You cannot use little to - >describe the availability of a product when W+ >you have no release date for that product.D ><  >Which do you want it to be ???? >4  3 Neither of the above.  Your either/or leaves out an<5 important possibility.  I knew 1.3 was the latest andh4 I didn't think that 6 months or a year was that much
 of a deficit.&  2 >> This debate really isn't about Java, it's about3 >> accuracy, mine and yours.  You keep spouting offC0 >> huge inaccuracies and picking at nits from my >> postings. >> 1 >30 >This is not picking nits, you either knew that 0 >1.3 was the current release and hence knew that2 >your little statement wasn't true or you did not. > = >> >SPEC JBB2000 the server side Java benchmark would be muchr9 >> >more applicable to the market the Alpha boxes runninge? >> >any OS play in and no Compaq does not have industry leadingH! >> >performance for SPEC JBB2000.. >> > >>  > >> I'm not sure why Compaq hasn't done JBB2000 on their larger; >> systems, but the price performance looks excellent, withn7 >> similarly sized Compaq systems beating the similarlyg6 >> sized competition.  Compaq has the best 2 processor5 >> result, the best 4 processor result and the best 8n >> processor result. >>   >_5 >Firstly SPEC does not measure price performance, and_4 >secondly Compaq have not published an 8 CPU JBB20007 >result. They do not have the best 8 CPU result becausen >they have not published one.d >   < You are right.  I misread the table.  Compaq doesn't have an 8 CPU result published.5  ? Now, are you going to answer the question as to why Sun hasn't d published any JBB2000 numbers?  5 >> > Oh?  Just about branding?  Let's take a look at:  >> >E >> > http://www.zdnet.com/intweek/stories/news/0,4164,2404696,00.htmlt >> > >I/ >Wonderfull the URL you posted is exactly what r/ >I said, a reference to a branding dispute longi$ >since resolved between Sun and IBM. >$  1 Funny that you would snip (without any editing ton. show that you have snipped) my quote from that/ URL that demonstrate very clearly that IBM was,i. in fact, concerned about standardization.  If + anyone is interested, just access the above  URL and find it for yourself.     - I'd quote it again, but you'd just bluster onL, and say that I've proven your point and that1 I clearly don't understand the difference betweenO/ branding and standardization, blah, blah, blah.  - >The branding dispute was not about what say e+ >should be included in for example JMS etc  * >but about what components JMS etc had to / >be included in a platform for it to be brandede >as being J2EE compliant.  >n2 >IBM wanted a reduced set of components Sun wanted* >all the origionally specified components. >s3 >Nor would the dispute have stopped IBM rolling outm5 >their own Java platform, if it hadn't been resolved i6 >which it was before your posting IBM simply would not4 >have been able to brand say WebSphere as being J2EE >compliant.z >.1 >IBM have subsequently signed up to the branding a2 >scheme as have the other major middleware vendors >such as BEA etc.9 >n2 >It is ironic that in producing this post you have1 >demonstrated that you don't grasp the differencec1 >between branding (what Sun and IBM were arguing ' >about) and standardisation. > / >And I did respond to it then pointing out your ( >comprehension gap (branding/standards). >y  2 No, you did not respond to the point I made above 2 from article <8u9hsv$rfu$1@nnrp1.deja.com>.  I can1 find no article that references this post nor can.. I find an article where I made the claim that 1 IBM was concerned about Java Standardization, youn+ said "no, just branding, you clearly don't o0 understand the difference between stadardisation- and branding" and I proved it with the above  0 elided quote (for those interested go back to my2 last post, as usual there are a lot of interesting0 questions that Andrew is dodging there) and you  went silent.  0 I don't know why I bother.  It's really pathetic/ how you spin and lie.  It's absolutely as clearh0 as it can be that you are trying to bluster your/ way through being caught out without an answer,  again.   >Regards >Andrew Harrison >Enterprise IT Architect   -Jordan Hendersona jordan@greenapple.com    ------------------------------  % Date: Mon, 22 Jan 2001 19:23:58 +0100s  From: Paul Sture <paul@sture.ch>$ Subject: Re: Remote boot via DECnet?+ Message-ID: <VA.00000252.040f1350@sture.ch>o  J In article <OFB900B0EA.55756277-ON802569DC.004A4655@qedi.quintiles.com>,   wrote:" > From: steven.reece@quintiles.com > Newsgroups: comp.os.vmsp& > Subject: Re: Remote boot via DECnet?' > Date: Mon, 22 Jan 2001 13:36:29 +0000  >  > G > It would be interesting to see what kind of configuration the variouscF > systems have.  I'm using DECnet-Plus and always seem to get the sameJ > problems as Bart - DECnet shuts down too early and kills off the sessionM > that's shutting the system down.  This is whether I'm running with a DECneth- > over IPP connection or just on DECnet-Plus.s > K > Two ways to get around it.  The first has already been suggested which isy% > to use SYSMAN to reboot the system.tG > Another way is to connect a DECserver up to the serial console on thenH > system and connect to the appropriate port on the DECserver, login andK > reboot the system.  This I can recommend if you have to do remote support K > of a system since you can watch the system as it reboots as well as crashd< > it or bring it down to console level if it goes diodes up. >  >eC I usually use rlogin for a shutdown. But is that from habit or not?   M Let's clear up version numbers first. Alpha V7.2-1, DECnet/Plus V5.0D.01 (as d< per startup message). The VAX has V7.2 with DECnet Phase IV.   Alpha first:  8 A quick poke around in shutdown.com reveals the logical H SHUTDOWN$DECNET_MINUTES. Hmm, it looks as though setting it to -1 would M disable DECnet shutdown. It has a default of 6 minutes on both Alpha and VAX.e  K Probing further, sys$system:sys$net_services_shutdown.com checks to see if hN sys$manager:net$shutdown.com exists, and executes it if it does. That in turn M checks to see if the terminal name begins with "RT", and outputs an error if oI it does. From a brief glance I don't think that code gets executed on an D immediate shutdown.e  
 Here goes:  E SHUTDOWN message on SPEEDY from user PAUL at _ALPHA1$RTA1:   18:43:27e7                                                      ^^r  K ALPHA1 will shut down in 5 minutes; back up shortly via automatic reboot.  t Please log off node ALPHA1.f
 Standalone    ; ***********************************************************e; Cannot shut down DECnet/OSI for OpenVMS from this terminal.    -- Exiting NET$SHUTDOWN.COM.; ***********************************************************w    ..wait for Alpha to come back...  N Try a shutdown of the VAX, with a 6 minute delay. It says the network will be M shut down, and sure enough I can't do a SET HOST to it anymore. The terminal ,F session is still there though, and shutdown is continuing OK. As I've ( highlighted above, it _is_ an RT device.  : Again back to the shutdown files to see what's happening. 7 sys$manager:net$shutdown doesn't exist. For DECnet V4, hL sys$system:sys$net_services_shutdown.com _does_ exist. In turn, that does a    $ stop/network DECnetn  N My conclusion is that it got fixed somewhere in V7.2, and my continued use of K rlogin is down to force of habit. I'm pretty sure that the problem existed rM with V7.1, and vaguely recall getting bitten by this recently on a system we e3 deliberately kept back at 7.1 for support purposes.h >  > John Santos wrote/quoted: K > >>>I've done this lots of times, it hasn't ever failed that I can recall.tE > However it is possible that all the DECnet-plus hosts I've rebootedqF > remotely, I was connected on a DECnet-over-IP connection rather thanE > on a raw DECnet connection.  I think that would require both DECnet A > and TCP/IP to still be up when SHUTDOWN.COM actually shuts down 0 > the system (runs OPCCRASH.EXE?), but it works. >  > ' > On Sat, 20 Jan 2001, Bart Zorn wrote:t > = > > It depends on the version of DECnet that you are running.e > >aK > > DECnet IV: it will probably work, as the RTA session stays up until the/L > > AS2100 actually goes down to reboot. DECnet-plus, OTOH, stops at an veryD > > early stage during shutdown (Way too early, IMHO). You lose your > connection, > > and the shutdown procedure gets aborted. > >iG > > With both versions of DECnet you could use the SYSMAN SHUTDOWN NODE-
 > command,F > > with /AUTO. In that way the shutdown procedure does not run in the	 > contexteA > > of your RTA session and you are not susceptible for prematureD > disconnects.<<<v >m   ___c
 Paul Sture SwitzerlandB   ------------------------------    Date: 22 Jan 2001 13:51:18 -05002 From: malmberg@eisner.decus.org (John E. Malmberg)$ Subject: Re: samba 2.0.3 : NT client+ Message-ID: <YBW$a249sNKz@eisner.decus.org>l  ( In article <3a6c4da1@news.kapsch.co.at>,0 eplan@kapsch.network (Peter LANGSTOEGER) writes:- > In article <8yqn4FCMnjgK@eisner.decus.org>, / > wb8tyw@qsl.network (John E. Malmberg) writes:aA >>One thing to do to improve directorly listing time on SAMBA fora= >>OpenVMS, is to modify the parameter "case sensitive = YES".t >>A >>(If you have the option set to use case sensitive names that is : >> only available on SAMBA 2.0.3, this may not work right) >>< >>On OpenVMS, the file lookup is always case insensitive, so9 >>changing this parameter from the default will result inn? >>significantly less code being executed, thus an optimization.? >v > Are you sure about that ?s >o@ > I changed this parameter with V1.9 and then I had to enter the9 > filenames as lowercases on the client side or else theyn< > wouldn't be findable.  I switched then back to original...  5 I have only tested this with SAMBA 2.0.6 for OpenVMS.:    = The released SAMBA V1.9.x can not deal with certain legal VMS7< file specifications, and has other interesting problems that7 caused me to not use it's file name handling functions.   ? As such when I was doing SAMBA V1.9.x stuff I did not use them.o  = Because I had the routines done for port I was doing of SAMBAi8 for OpenVMS when 2.0.3 was released, I did not use those routines either.  > In the few tests that I ran on SAMBA 2.0.6, I saw a modest but noticable improvements.o  ; If I get a chance to produce an update to the SAMBA_VMS FAQ.: I will make sure to mention the behavior you saw on 1.19.?   -Johnu wb8tyw@qsl.network   ------------------------------  % Date: Mon, 22 Jan 2001 19:59:41 +0100   From: Paul Sture <paul@sture.ch>U Subject: Re: Script to read directory of all drives, clustered, shared, etc. at once?o+ Message-ID: <VA.00000253.042fcb93@sture.ch>   1 In article <94hocm$ptl$1@nnrp1.deja.com>,  wrote:n! > From: john_20_28_2000@yahoo.come > Newsgroups: comp.os.vms S > Subject: Script to read directory of all drives, clustered, shared, etc. at once?s% > Date: Mon, 22 Jan 2001 16:50:34 GMTt >  > Is there a DCL script to:h > F > Read the directory of all drives, clustered, shared, etc. at once byI > just running a command file and maybe saving it to an output?  I wantedo? > to be able to use wildcards in the command file with the DIR.  > E > I guess you could create a DCL command file that has all your driverF > names and accepts a parameter for DIR from the command line.  But, I? > did not know if this was the correct way to proceed.  Thanks.s >  > R Here's a snippet of code which processes each disk it finds. The device names are M returned in a random order, so you may wish to sort the output. Note that it dP excludes NFS mounted disks. This doesn't check whether a disk is write protected $  say := write sys$output $  node = f$getsyi("NODENAME") $!8 $  say " %SYS-I-RUNNING  ''F$ENVIRONMENT("PROCEDURE")'"  $!
 $ set noon $!-- $  set process /priv=all! $  on control_y then goto abortedn $!% $! Other initialization stuff in heres $!( $  date = f$cvtime(,"COMPARISON","DATE")D $  devclass = 1    ! search only for disks (see f$device for details& $!           of other devclass values)/ $  device_context = 0  ! Initialize the contextj $!H $  device_context = device_context + 1 ! for multiple passes of the code $get_disk_name:D $!2 $  disk_name = f$device(,devclass,,device_context)+ $  if disk_name .eqs. "" then goto all_donerA $  if .not. f$getdvi(disk_name, "EXISTS") then goto get_disk_namei> $  if .not. f$getdvi(disk_name, "MNT") then goto get_disk_name@ $  if f$getdvi(disk_name, "SHDW_MEMBER") then goto get_disk_nameJ $  if f$getdvi(disk_name, "FOR") then goto get_disk_name ! mounted foreign $!) $! Omit the "foreign" disks e.g. $1$DNFS1  $!6 $  disk_type = F$GETDVI(DISK_NAME, "DEVICE_TYPE_NAME")B $  if f$locate("Foreign disk", disk_type) .ne. f$length(disk_type) $  then 
 $      say ""oG $      say "%SYS-I-FOREIGNDISK, ", disk_name, ", type = ", disk_type, -         " - Skipped"l $      goto get_disk_names $  endif $! $! Now look at the disk  $!	 $  say ""r( $  volume = f$getdvi(disk_name,"VOLNAM")4 $  say "Processing disk ", disk_name, ", volume ", -
        volume.+ $  call get_disk_usage ! Report on the disk.' $  if $status .eq. 40 then goto abortedo $! $  goto get_disk_name 	 $aborted:d $  say "Aborted by CTRL/Y" $ !n
 $all_done: $  exitd $! $get_disk_usage: subroutinep $! $  on control_y then exit 40 $!  $! Do the work on each disk here $!	 $  returnH# $  endsubroutine   ! get_disk_usagee     ___ 
 Paul Sture SwitzerlandV   ------------------------------    Date: 22 Jan 2001 20:41:29 -0500/ From: jordan@lisa.gemair.com (Jordan Henderson)yU Subject: Re: Script to read directory of all drives, clustered, shared, etc. at once?w* Message-ID: <94ing9$5rh$1@lisa.gemair.com>  ! > From: john_20_28_2000@yahoo.comp > Newsgroups: comp.os.vms,S > Subject: Script to read directory of all drives, clustered, shared, etc. at once?e% > Date: Mon, 22 Jan 2001 16:50:34 GMTI >  > Is there a DCL script to:  > F > Read the directory of all drives, clustered, shared, etc. at once byI > just running a command file and maybe saving it to an output?  I wanted ? > to be able to use wildcards in the command file with the DIR.i > E > I guess you could create a DCL command file that has all your drivedF > names and accepts a parameter for DIR from the command line.  But, I? > did not know if this was the correct way to proceed.  Thanks.  >  >   E Here's a snippet of Perl that will execute a command for each locally ; mounted disk.  The command line to get a directory list fort, all locally mounted disks is something like:  , $ perl foreachdisk.pl "DIR DISK:[000000...]"  E Where DISK will be replaced with each disk device in the system whichwC is locally mounted on your machine and the string will be executed DD for each one.  You can change the replacement string, DISK being the default, with the -d parameter.t  1 No warrantees, implied or otherwise, etc.  Enjoy.a   ---cut here--- #   foreachdisku # H #       foreachdisk <-d REPLACE_NAME> <command string with REPLACE_NAME> #r( #       The default REPLACE_NAME is DISKK #       foreachdisk will execute the command on each mounted disk replacingg) #       REPLACE_NAME with each disk foundh   my $replace_name;r
 my $template;l  & open( DEVICES, "SHO DEV D/MOUNTED|" );  / $_ = <DEVICES>; $_ = <DEVICES>; $_ = <DEVICES>;b  E if ( eof( DEVICES ) ) { die "Internal problem getting device names" }    if ( $ARGV[ 0 ] =~ /-d/ ) {e     $replace_name = $ARGV[ 1 ];r     $template = $ARGV[ 2 ];h }  else {t     $replace_name = "DISK";n     $template = $ARGV[ 0 ] }    while (<DEVICES>) { !     if ( not /(remote mount)/ ) {h          my @line    = split /:/;         my $DISK    = $line[0];p          my $command = $template;  ,         $command =~ s/$replace_name/$DISK/g;           system $command;     }Y }d --- cut here ---   -Jordan Hendersonr jordan@greenapple.com    ------------------------------  % Date: Mon, 22 Jan 2001 19:24:51 +0000s  From: steven.reece@quintiles.comT Subject: Re: Script to read directory of all drives, clustered, shared,etc. at once?H Message-ID: <OFA7FE5899.70C5453A-ON802569DC.006A60FA@qedi.quintiles.com>  , You could use something like the following :   $! $! $! SRR - 22-JAN-2001 $! $ IF "''P1'".EQS.""o $ THENH $       WRITE SYS$OUTPUT "%CHK-E-NOPARAM, no parameter supplied to find" $       EXIT $ ENDIFu $!	 $GETDISK:h $ DISK=F$DEVICE("*","DISK")4# $ IF DISK.EQS."" THEN GOTO FINISHEDr $!x $! (here you need to check whether the disk is a shadowset member or not and also whether the disk is accessible to you) $! $ FILESPEC="''DISK'"+"''P1'" $ DIRECTORY 'FILESPECa $ GOTO GETDISK $! $!
 $FINISHED: $ WRITE SYS$OUTPUT "Finished." $ EXIT  H john underscore 20 underscore 28 underscore 2000 at yahoo dot com asked: >>>Is there a DCL script to:  D Read the directory of all drives, clustered, shared, etc. at once byG just running a command file and maybe saving it to an output?  I wanted = to be able to use wildcards in the command file with the DIR.m  C I guess you could create a DCL command file that has all your driveoD names and accepts a parameter for DIR from the command line.  But, I@ did not know if this was the correct way to proceed.  Thanks.<<<   ------------------------------  # Date: Mon, 22 Jan 2001 22:40:51 GMTr2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)< Subject: Re: Seeking OpenVMS docs (was Re: Oldtimer forgot!)5 Message-ID: <TF2b6.33$cu.225@gazette.loc1.tandem.com>s  N In article <VA.00000243.22c4151e@sture.ch>, Paul Sture <paul@sture.ch> writes:G :In article <3A6A6024.4684C644@earthlink.net>, David J. Dachtera wrote:v :h :[snip]CL :> > Hi, getting to the linker docs is what I don't have.  I'm logging in to :> > a remote system.qC :> > If I'd only asked two months earlier I could have had the fullnF :> > documentation set.  Wouldn't know where I'd park all of them tho.H :> > I checked Compaq's site and couldn't find any detailed linker docs.  F   Please check the OpenVMS FAQ.  You will find pointers to the OpenVMSF   and compiler documentation, to the shareable image cookbook, and you;   will find details of how to run programs with parameters.B  E   Please also remember to pick a title that makes sense and that willoC   attract the attention of folks that might have the information or &   the answer to the question you seek.  H :> Well, stuff *IS* hard to find over there. We've been hounding the "Q"G :> about it for more than two years now - no improvement, in fact hingso :> continue to deteriorate.y  M   If you select "products & services / servers", you get a clickable picture aI   of the AlphaServer systems and a reference to Tru64 UNIX, OpenVMS, and uF   Linux.  If you pick "products & services / workstations", you get a F   clickable  picture of the AlphaStation series.  From AlphaServer or I   AlphaStation you can get directly to OpenVMS.  If you pick "Enterprise oH   solutions and Services" and then "software", you can get to OpenVMS.  I   If you click on "Products & Services" software, you can get to OpenVMS.iI   If you find the FAQ (comp.answers or MIT RTFM), you can get to OpenVMS.d  I   If you get to the OpenVMS main page, the documentation link on the main G   page contains the available OpenVMS documentation, various integratedtE   products, and a link to the layered product documentation download.tB   Getting to the Commercial product documentation is tougher, thatD   requires some navigation or a look at the FAQ -- I've asked that aF   link from the OpenVMS documentation area over to that area be added,F   and was chatting with the webmaster about consolidating all of this (   documentation onto the same website...  K   Please describe to me how I could have the main page navigation improved,eK   or navigation on the OpenVMS website.  (Well, short of replacing the main @   www.compaq.com contents with the OpenVMS website, that is. :-)  M :Not how I see it. For example, a colleague tells me that the Ada User Guide nL :has just appeared in pdf format on the Compaq website, meaning that he can  :print it out.  E   The OpenVMS commercial product website has a variety of informationdH   and documentation available.   (There's a browser-(in)tolerance bug atH   the C documentation website due to be fixed tonight, resulting from a    latent error in the HTML...)  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Mon, 22 Jan 2001 17:15:56 -08005) From: Wayne Holland <wholland@tscnet.com> < Subject: Re: Seeking OpenVMS docs (was Re: Oldtimer forgot!)O Message-ID: <A45D990AEC947E79.20DF80A89791D6C7.66807AF07BA350C7@lp.airnews.net>q   Hoff Hoffman wrote:: > P > In article <VA.00000243.22c4151e@sture.ch>, Paul Sture <paul@sture.ch> writes:I > :In article <3A6A6024.4684C644@earthlink.net>, David J. Dachtera wrote:w > :b	 > :[snip]nN > :> > Hi, getting to the linker docs is what I don't have.  I'm logging in to > :> > a remote system.tE > :> > If I'd only asked two months earlier I could have had the fulllH > :> > documentation set.  Wouldn't know where I'd park all of them tho.J > :> > I checked Compaq's site and couldn't find any detailed linker docs. > H >   Please check the OpenVMS FAQ.  You will find pointers to the OpenVMSH >   and compiler documentation, to the shareable image cookbook, and you= >   will find details of how to run programs with parameters.t > G >   Please also remember to pick a title that makes sense and that willsE >   attract the attention of folks that might have the information or ( >   the answer to the question you seek. > J > :> Well, stuff *IS* hard to find over there. We've been hounding the "Q"I > :> about it for more than two years now - no improvement, in fact hings  > :> continue to deteriorate.s > N >   If you select "products & services / servers", you get a clickable pictureJ >   of the AlphaServer systems and a reference to Tru64 UNIX, OpenVMS, andG >   Linux.  If you pick "products & services / workstations", you get auG >   clickable  picture of the AlphaStation series.  From AlphaServer or6J >   AlphaStation you can get directly to OpenVMS.  If you pick "EnterpriseH >   solutions and Services" and then "software", you can get to OpenVMS.K >   If you click on "Products & Services" software, you can get to OpenVMS.eK >   If you find the FAQ (comp.answers or MIT RTFM), you can get to OpenVMS.d > K >   If you get to the OpenVMS main page, the documentation link on the main I >   page contains the available OpenVMS documentation, various integrated G >   products, and a link to the layered product documentation download.dD >   Getting to the Commercial product documentation is tougher, thatF >   requires some navigation or a look at the FAQ -- I've asked that aH >   link from the OpenVMS documentation area over to that area be added,G >   and was chatting with the webmaster about consolidating all of this** >   documentation onto the same website... > M >   Please describe to me how I could have the main page navigation improved,yM >   or navigation on the OpenVMS website.  (Well, short of replacing the mainaB >   www.compaq.com contents with the OpenVMS website, that is. :-) > N > :Not how I see it. For example, a colleague tells me that the Ada User GuideM > :has just appeared in pdf format on the Compaq website, meaning that he cant > :print it out. > G >   The OpenVMS commercial product website has a variety of informationdJ >   and documentation available.   (There's a browser-(in)tolerance bug atI >   the C documentation website due to be fixed tonight, resulting from a'  >   latent error in the HTML...) > P >  --------------------------- pure personal opinion ---------------------------N >    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com" Hello Hoff, nice to hear from you.E When I was managing a system I had a 4 hour response time contract onsF software and hardware back in the 80's.  Sometimes one gets in a hurryH and just don't have the time to get the answers quickly out of the booksC we had then, so I just called DEC on line for the sometimes obviouslB answers.  I had to do this then when my boss was breathing down my@ back.  I realize web pages are hard to organize for such a largeE organization, but I feel a more organized thread is needed to the webr pages.... like OpenVMS docs ->F Overview of Docs -> then a branching tree to the area of interest.  OrA something like that.  HP's site got worse.  I was looking for the 6 technical specs on one of their printers and got lost.-      I really like VMS tho.  Sure do miss it.n   ------------------------------  # Date: Mon, 22 Jan 2001 19:37:40 GMTd" From: falk@arc.ab.ca (Alfred Falk)0 Subject: Re: Source for used StorageWorks disks?9 Message-ID: <Xns903180735BDCFfalkarcabca@205.233.108.180>e  / Paul Repacholi <prep@prep.synonet.com> wrote in # <87bst3gtuc.fsf@prep.synonet.com>: u  5 >rdeininger@mindspring.com (Robert Deininger) writes:e >2E >> The containers don't go all that cheap.  I think quite a few folksb& >> have learned how to open them. :-)  > 1 >Yes, I've collected a stack of 'dead' bricks. ;)a >cA >Question, I have a RZ-28 I think, FR-PCWVR-AY, black 'cuda. It'sT< >labeled a wide, but is all grey, not green. Is this normal?1 >I'd understood that grey = narrow, green = wide.   " It's more like grey=old, blue=new., Best to look at the label on front of brick.- Model numbers ending in -VA=narrow, -VW=wide. 5 Inside the SCSI diamond you should see something like      	10t     	--o     	F10' Top number: Transfer rate in Mbytes/sechH Bottom: Bus speed S=5/sec (slow), F10=10 MB/sec (fast), F20 (ultra), ...: to Lower Left of diamond: SCSI bus width: N=narrow, W=WideJ To lower right of diamond: shelf compatability: N=narrow, W=Wide, N/W=both  J If it's grey and doesn't have a label on the front, chances are it's slow ( narrow.  Look at model number on bottom.  @ ----------------------------------------------------------------A   A L B E R T A         Alfred Falk               falk@arc.ab.ca f@ R E S E A R C H         Information Systems Dept   (780)450-5185+   C O U N C I L         250 Karl Clark Roado1                         Edmonton, Alberta, Canadae http://www.arc.ab.ca/   T6N 1E4s  http://www.arc.ab.ca/staff/falk/   ------------------------------  # Date: Mon, 22 Jan 2001 23:51:28 GMTo2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)B Subject: Re: To Hoff Hoffman: Any news about the VMS DHCP client ?5 Message-ID: <4I3b6.36$cu.444@gazette.loc1.tandem.com>i  _ In article <3A6C215F.2E62320E@oracle.com>, norm lastovica <norman.lastovica@oracle.com> writes:u :t :Peter LANGSTOEGER wrote:s :>  a :> In article <3a6b5ede@newsfeed.vitts.com>, "John Gemignani, Jr." <john@ossc.DELETE.net> writes::> :> >TCP/IP Services for OpenVMS V5.1 submitted this past week.( :> >It includes support for DHCP client. :>  , :> May I ask someone for a copy of the kit ? :01 :	I believe that when you sign up for the vms 7.316 :field test, you get a copy of the tcp/ip services 5.15 :field test software as well.  otherwise, you'll need  :to contact compaq directly.  H   QA-MT3AD-H8.  US$40.  The second version of the OpenVMS V7.3 Software I   Developer's Kit (SDK2).  Includes OpenVMS VAX or Alpha V7.3 EFT2/SDK2, e=   TCP/IP Services V5.1, and a variety of other bits and kits.a  G   Formal OpenVMS External Field Test (EFT) sites have rather more of aneH   investment (at both EFT customer and OpenVMS Engineering) than do SDK G   sites, in terms of the increased levels of engineering interactions, rE   frequency of engineering contacts and feedback, and formal problem y'   reporting tools and formal responses.s  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Tue, 23 Jan 2001 16:44:57 +1200 . From: Nivlesh Chandra <NChandra001@itc.gov.fj>2 Subject: Urgent:Error recovering files from backupO Message-ID: <791C2856E8FDD211BAFB0008C759919501A62B3A@exchange01.govnet.gov.fj>o  L I have a tape that contains a  backup that was done using the backup commandE in VMS. Now when I try and recover the file I get the following errord  4 excessive error rate reading $2$MKA300:[000000]*.*;* software header crc erroro  * I used the following to restore the files   ! $ mount $2$mka300 /media=compact fL $ backup $2$mka300:[000000]bak-13-12-99.bck;1 /save disk44:[recover...]*.*;*. /log /list=disk$scratch:bak.lis /nocrc /rewind  H Then since I was getting so much error.. I thought I would just copy theF saveset onto the disk and then recover .. I used the following command  ? $ backup $2$mka300:[000000]*.*;* disk44:[test]*.*;* /log /nocrcs  # but received the same error message)  J Please I really need to get the files in this saveset and if someone could( help me I would really appreciate it ...   Please respond...P   Niv)   ------------------------------  # Date: Tue, 23 Jan 2001 05:46:17 GMTt) From: "Rob Brown" <robbrown@shaw.wave.ca>R6 Subject: Re: Urgent:Error recovering files from backupJ Message-ID: <01c08500$4a096dc0$0f754718@cs918188-a.edmw1.ab.wave.home.com>  9 Nivlesh Chandra <NChandra001@itc.gov.fj> wrote in articleNF <791C2856E8FDD211BAFB0008C759919501A62B3A@exchange01.govnet.gov.fj>...F > I have a tape that contains a  backup that was done using the backup command G > in VMS. Now when I try and recover the file I get the following error/ > 6 > excessive error rate reading $2$MKA300:[000000]*.*;* > software header crc errorn  D Others will tell you that to specify VMS version number and also theJ complete text of the error messages, including the facility, severity, and ident.  5 > $ backup $2$mka300:[000000]bak-13-12-99.bck;1 /save  disk44:[recover...]*.*;*0 > /log /list=disk$scratch:bak.lis /nocrc /rewind(                                   ^^^^^^@ /nocrc is *always* a bad idea isn't it?  I quote from the BACKUP documentation:  4 >>> note that use of /NOCRC reduces processing time ( >>> but increases the risk of data loss.  A I admit that I am unclear about the relationship between /CRC and  /GROUP_SIZE.  H In this case I am not sure what it does.  It can be either an input saveH set qualifier or an output save set qualifier, but you have specified itJ for an output file.  Did you create the saveset with /NOCRC?  My immediate: inclination would be to leave it off and see how it works.  J > Then since I was getting so much error.. I thought I would just copy theH > saveset onto the disk and then recover .. I used the following command > A > $ backup $2$mka300:[000000]*.*;* disk44:[test]*.*;* /log /nocrcm  < That is not the command to copy the saveset onto disk.  Use:      $ mount/override=id $2$mka300:2   $ copy/log $2$mka300:[000000]*.*;* disk44:[test]  
 Good luck.   - Rob.   ------------------------------  # Date: Mon, 22 Jan 2001 22:46:09 GMTc2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: Vax 7830 or 7730o5 Message-ID: <RK2b6.34$cu.225@gazette.loc1.tandem.com>t  c In article <3A6B4675.F2F1ACD9@decus.fi>, Veli =?iso-8859-1?Q?K=F6rkk=F6?= <korkko@decus.fi> writes:r7 :...must be correctly set. Otherwise even VMS V7.2 willC9 :recognize a VAX 7000-8x0 as VAX 7000-7x0. I think it was ; :BOOTVAX_COMPAT and the trick was to allow V5.5-2H4 run on e9 :-8x0. You may have this setting wrong and therefore VMS C; :recognizes system as a -7x0. It still runs just like real a :-8x0 if it really is a -8x0.   H   It definitely wasn't the VAX 7000 model 700 that was emulated for use K   on earlier OpenVMS VAX releases, as that platform is not supported prior VM   to OpenVMS V6.1 -- only the model 600 and platforms that can be configured ]K   to emulate the 600 will bootstrap prior to V6.0.  That the 700 would not  L   work on V5.5-2 and V5.5-2H4 was found to be an impediment to sales, hence 4   the provision for emulation of the 600 on the 800.  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------   Date: 23 Jan 2001 02:02:19 GMT' From: "Jim Strehlow" <jims@data911.com>v Subject: Re: WRQ drops LAT ???0 Message-ID: <94ionb$60o@dispatch.concentric.net>  L About nine months ago I phoned WRQ tech support with a LAT-related question.: Someone at WRQ told me they were dropping support for LAT.  I About three months ago we bought a couple of extra versions of Reflection F for new desktops, received version 8.00.076 of Reflection for UNIX andI Digital, expected not to find LAT support, and still found LAT was there.D  J Our company supports many customers who have several hundred LAT connectedI devices. The customers have been satisfied for years; but, as support foro? the old stable technology goes away, we will help them migrate.O   Jim Strehlow, jims@data911.com Senior Applications Analyst    ------------------------------  % Date: Mon, 22 Jan 2001 18:04:33 -0500 - From: JF Mezei <jfmezei.spamnot@videotron.ca>e$ Subject: X-windows paradigm question, Message-ID: <3A6CBC77.16066267@videotron.ca>  M In an X-windows, exactly what processing is done by the server (the X-windowsnB terminal) and the client (the host with the application running) ?  L For instance, if one has a data entry window, does every keystroke result inP the client code running, or does the x-terminal do a lot of the work by itself ?  L I am trying to compare the x-terminal paradigm with that of HTTP. HTTP seemsN very very similar to the IBM CICS paradigm where once the screen is displayed,L the mainframe is detached from the terminal and the terminal "reattaches" toG its application (establishing context etc) once the user presses ENTER.e  N Are there applications where HTTP can yield response times that are comparableM to that of 3270 terminals in the past ? Seems that HTTP has a lot of overheado to process transactions.  I If the new industry paradigm is to go away from client-server and instead L return to the "dumb terminal" approach because of simplified support, lesserM complexity, easier management etc, I am just wondering which of the X-windowsr' or HTTP/HTML paradigms would be better.    ------------------------------   Date: 22 Jan 2001 15:52:35 PSTT From: Fairfield@SLAC.Stanford.EDU (Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515)( Subject: Re: X-windows paradigm question3 Message-ID: <bEXu4vN+yiCG@mccdev.slac.stanford.edu>   , In article <3A6CBC77.16066267@videotron.ca>,4     	JF Mezei <jfmezei.spamnot@videotron.ca> writes:O > In an X-windows, exactly what processing is done by the server (the X-windows D > terminal) and the client (the host with the application running) ?  H         The server provides the graphics display services to the client.H     Well, and keyboard  and  mouse  services  (key  &  button  presses &     releases).  N > For instance, if one has a data entry window, does every keystroke result inI > the client code running, or does the x-terminal do a lot of the work bya
 > itself ?  H         I'd say both.  Every key  press  generates an "X event" which isH     transmitted from the server to the client.  The client may choose toH     process,  or  act  upon,  that or not.  Try running XEV  for  a  fewF     minutes.  It may open your eyes to what's going on in X-windows...  O > I am trying to compare the x-terminal paradigm with that of HTTP.  HTTP seems E > very very similar to the IBM CICS paradigm where once the screen istI > displayed, the mainframe is detached from the terminal and the terminal J > "reattaches" to its application (establishing context etc) once the user > presses ENTER.  H         Agreed.  HTTP sends and  receives  "batches"  of data, otherwiseB     known as "pages".  Very similar to handling of 3270 terminals.  E > Are there applications where HTTP can yield response times that areeO > comparable to that of 3270 terminals in the past ?  Seems that HTTP has a lotr& > of overhead to process transactions.  H         Doesn't this depend on  the  context?   If you're sending singleH     key-strokes,   or  prompting  for  user  input,  etc.,  HTTP   seemsH     inappropriate.  If you're asking the user to fill  out  forms,  thenH     all  the  "editing" of filling in fields, etc., can done locally (inH     the browser for HTTP, in the 3270 for the IBM paradigm) and only theH     composed  results  are  sent,  possibly   in  a  very  efficient  or     compressed form.  K > If the new industry paradigm is to go away from client-server and instead-N > return to the "dumb terminal" approach because of simplified support, lesserO > complexity, easier management etc, I am just wondering which of the X-windows-) > or HTTP/HTML paradigms would be better.a  H         I'd venture that it's application  dependent.  Filling out formsH     in  a browser/3270 terminal is fine.  Doing single-key responses  toH     prompts is better done in "dumb  terminal"  or  XTERM/DECterm  mode.H     BTW, I really hate editing in a form displayed by Netscape.  Give me%     a terminal and an editor anytime!3           -Ken --M  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:      650-926-3515N  -----------------------------------------------------------------------------B  These opinions are mine, not SLAC's, Stanford's, nor the DOE's...   ------------------------------  + Date: Tue, 23 Jan 2001 00:30:23 +0000 (   )w3 From: Christopher Smith <chriss@Mufasa.pubserv.com>c( Subject: Re: X-windows paradigm questionJ Message-ID: <Pine.LNX.4.05.10101230016070.16967-100000@Mufasa.pubserv.com>  $ On Mon, 22 Jan 2001, JF Mezei wrote:  O > In an X-windows, exactly what processing is done by the server (the X-windows D > terminal) and the client (the host with the application running) ?  N > For instance, if one has a data entry window, does every keystroke result inR > the client code running, or does the x-terminal do a lot of the work by itself ?  I Well, I can tell you this, just from experience in running X apps on slowuG networks.  The only thing the X terminal seems to do is the graphics --tI even scaling of a viewport style window (something with scrollbars, wherewG you could just clip a region and re-draw the decorations...) is done bye the client application.     N > I am trying to compare the x-terminal paradigm with that of HTTP. HTTP seemsP > very very similar to the IBM CICS paradigm where once the screen is displayed,N > the mainframe is detached from the terminal and the terminal "reattaches" toI > its application (establishing context etc) once the user presses ENTER.g  @ I'm not sure there's a lot to compare, here, but if you think ofI html/cgi/whatever as alike to CICS, then I imagine it's correct to assume ( that X11 is simmilar to a dumb terminal.  H The reason for this, of course, is that the terminal really does quite aD bit of work when you're using html, etc...  There have recently beenH hybrids which might offer you the best of both worlds, for instance, theI relatively "smart" xterms called "network computers," which have built-insJ java VM, or something like that so that more of the work can be transferedI to the terminal.  Of course, your application would need to be written to  take advantage of this.O  P > Are there applications where HTTP can yield response times that are comparableO > to that of 3270 terminals in the past ? Seems that HTTP has a lot of overheadl > to process transactions.  G Well, HTTP is a protocol based on TCP, based on UDP, so there is enoughnH overhead for (probably slightly less) three protocols. :)  There's a lotE of overhead, but response time can still be good, especially with the H respectable network pipes available today, and depending on application.I I've seen so-called "intranet" applications function perfectly well using) HTTP.   G I can't answer your question directly, though, since I have no idea how- quickly a 3270 will respond.  K > If the new industry paradigm is to go away from client-server and insteadfN > return to the "dumb terminal" approach because of simplified support, lesserO > complexity, easier management etc, I am just wondering which of the X-windowst) > or HTTP/HTML paradigms would be better.c  G I would say that it's _heavilly_ application dependant, and that if youeI require any flexability or "real time" response, X11 is the clear choice.dJ On the other hand, if your application lends itself to a sort of "fill outE the form and we'll do this in a batch" system, HTML may be the betteraF answer.  So, for some kind of simulation, for instance, or the like, XH would clearly be the choice, but for data entry, depending on the amountD of data that will be entered at any given time, HTTP and HTML may be better.   F Honestly, my position is that there should be a device available whichH does a combination of the two (or uses two simmilar methods).  Note thatJ this hybrid beast wouldn't have to be a "network computer," but could evenC be a low-powered desktop system with enough of the operating systemeJ installed to net-boot from a server, and find application programs in someG directory which could be on a remote disk.  You could then either use a G web-type interface, or something else to run applications on the users'eB behalf.  If the machine was small enough you could design it to be- fan-free, and look pretty much like an xterm.V   Regards,   Chris'  O ===============================================================================h@ "My two cents"			(http://rootworks.com/twocentsworth.cgi?128562)= Christopher Smith(chriss@pubserv.com)			Prgramer^W Programmerr Prime Synergy of Champaign, IL. % -------------------------------------eI "Where a calculator on the ENIAC is equipped with 18,000 vacuum tubes andnH weighs 30 tons, computers in the future may have only 1,000 vacuum tubes; and weigh only 1.5 tons." -- Popular Mechanics, March 1949 rO -------------------------------------------------------------------------------=   ------------------------------  % Date: Tue, 23 Jan 2001 02:11:52 +0000s) From: Christof Brass <brass@infopuls.com>t( Subject: Re: X-windows paradigm question, Message-ID: <3A6CE868.874F552F@infopuls.com>   Christopher Smith wrote: > & > On Mon, 22 Jan 2001, JF Mezei wrote: > Q > > In an X-windows, exactly what processing is done by the server (the X-windowsRF > > terminal) and the client (the host with the application running) ? > P > > For instance, if one has a data entry window, does every keystroke result inT > > the client code running, or does the x-terminal do a lot of the work by itself ? > K > Well, I can tell you this, just from experience in running X apps on slowgI > networks.  The only thing the X terminal seems to do is the graphics --6K > even scaling of a viewport style window (something with scrollbars, where I > you could just clip a region and re-draw the decorations...) is done bym > the client application.o  ; This is basically true although the X11 definition allows X5> display servers to do more, i.e. to offer more services to the= client like restoring the contents of an uncovered window and < the like. But it is stated in the X11 manual that the client> (the app) has to be programmed in a way that it can do the job@ itself without relying on these optional features of a certain X= display server. But all of these optional features are simple7? raster operations so the fact remains that all substantial workA> is done by the client (the app) and network traffic for events> may not be neglected although in most cases there no problems.@ There are a lot of events each mouse move generates one which is@ transfered over the network but the X11 commands from the client; to the X display server are intelligent. To copy bitmaps orc  pixmaps is reduced to a minimum.   > P > > I am trying to compare the x-terminal paradigm with that of HTTP. HTTP seemsR > > very very similar to the IBM CICS paradigm where once the screen is displayed,P > > the mainframe is detached from the terminal and the terminal "reattaches" toK > > its application (establishing context etc) once the user presses ENTER.h > B > I'm not sure there's a lot to compare, here, but if you think ofK > html/cgi/whatever as alike to CICS, then I imagine it's correct to assume.* > that X11 is simmilar to a dumb terminal. > J > The reason for this, of course, is that the terminal really does quite aF > bit of work when you're using html, etc...  There have recently beenJ > hybrids which might offer you the best of both worlds, for instance, theK > relatively "smart" xterms called "network computers," which have built-in0L > java VM, or something like that so that more of the work can be transferedK > to the terminal.  Of course, your application would need to be written toI > take advantage of this.e  < This is the way to go. HTML can be extended by JavaScript to= have local code executed as much as needed. But I prefer realo9 client/server apps replacing the good old DEC VTs by an Xa@ display. The IBM style has major disadvantages and unfortunately; all the applications I have used had incredible ugly screen > designs as compared to VT200 based applications. And the UI of= these 3270 apps was very strange in that you could completelyu< destroy the contents of the screen by moving around with the< text cursor and erasing text and putting in new at arbitrary@ positions. It was a real mess. Fortunately this application in a? public library has recently replaced by an HTLM interface which.& can also be used through the Internet.   > R > > Are there applications where HTTP can yield response times that are comparableQ > > to that of 3270 terminals in the past ? Seems that HTTP has a lot of overhead0 > > to process transactions. > I > Well, HTTP is a protocol based on TCP, based on UDP, so there is enougheJ > overhead for (probably slightly less) three protocols. :)  There's a lotG > of overhead, but response time can still be good, especially with the4J > respectable network pipes available today, and depending on application.K > I've seen so-called "intranet" applications function perfectly well usingI > HTTP.o  > As far as I know HTTP is completely independent and the others> are also. Normally HTTP is used on top of TCP/IP which in turn8 is independent of UDP. In fact TCP is based on IP and is= responsible for establishing connection (counting packets ande< putting it into the correct sequence). IP is connectionless,( whereas TCP is connection oriented IIRC.   > I > I can't answer your question directly, though, since I have no idea how  > quickly a 3270 will respond. > M > > If the new industry paradigm is to go away from client-server and insteadIP > > return to the "dumb terminal" approach because of simplified support, lesserQ > > complexity, easier management etc, I am just wondering which of the X-windowsm+ > > or HTTP/HTML paradigms would be better.e > I > I would say that it's _heavilly_ application dependant, and that if you[K > require any flexability or "real time" response, X11 is the clear choice.-L > On the other hand, if your application lends itself to a sort of "fill outG > the form and we'll do this in a batch" system, HTML may be the better3H > answer.  So, for some kind of simulation, for instance, or the like, XJ > would clearly be the choice, but for data entry, depending on the amountF > of data that will be entered at any given time, HTTP and HTML may be	 > better.0 > H > Honestly, my position is that there should be a device available whichJ > does a combination of the two (or uses two simmilar methods).  Note thatL > this hybrid beast wouldn't have to be a "network computer," but could evenE > be a low-powered desktop system with enough of the operating systemeL > installed to net-boot from a server, and find application programs in someI > directory which could be on a remote disk.  You could then either use abI > web-type interface, or something else to run applications on the users'eD > behalf.  If the machine was small enough you could design it to be/ > fan-free, and look pretty much like an xterm.l > 
 > Regards, >  > Chris  > Q > ===============================================================================fQ > "My two cents"                  (http://rootworks.com/twocentsworth.cgi?128562)-O > Christopher Smith(chriss@pubserv.com)                   Prgramer^W Programmers! > Prime Synergy of Champaign, IL.o' > -------------------------------------tK > "Where a calculator on the ENIAC is equipped with 18,000 vacuum tubes andLJ > weighs 30 tons, computers in the future may have only 1,000 vacuum tubes< > and weigh only 1.5 tons." -- Popular Mechanics, March 1949Q > -------------------------------------------------------------------------------   ? I basically agree but I don't like the HTTP approach because itd? is a major step backwards in UI design. If you had a little bits> more high level features in the X display server this would be> perfect to me. Of course we need the window manager to be able? to intercept locally to avoid traveling all the mouse movementsd= over the network. If we could split up the window manager and-> the other X clients in a pure screen and generic data oriented? part on one side which is run locally and the application logic 8 which is run remotely we can do all what we want without9 bringing down the networks. And we could do that over thev@ Internet although there is LBX low bandwidth X for such purposes! already which could be used also.e   ------------------------------   Date: 22 Jan 2001 17:34:39 PSTT From: Fairfield@SLAC.Stanford.EDU (Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515)? Subject: [Q]: Resetting privileged password on a DECserver 200?P3 Message-ID: <0$CQJFxSP7Oe@mccdev.slac.stanford.edu>e  H         We have a pair of  old  DECserver  200's that are clunking alongH     just fine.  Unfortunately, it's been so long since we've messed withH     them, no one remembers the privileged password.  Not the maintenanceB     password, mind you, that one is stored in the DECNET database.  H         What is the procedure to set  a new password?  It's very unclearH     to  me,  reading  the doc set that came with TSM 2.1 (the  docs  areH     V1.6, BTW), whether when you do a USE SERVER  server-name,  followedH     by  SET PRIVILEGED, just _which_ password is required.  Indeed, whenH     I do a SHOW SERVER ALL CHARACTERISTICS,  both a login password and aH     maintenance  password are listed, but neither one is accepted by SET@     PRIVILEGED (nor are the default strings for those accepted).  /         Any hints would be greatly appreciated!                Thanks, Keny -- eM  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:      650-926-3515N  -----------------------------------------------------------------------------B  These opinions are mine, not SLAC's, Stanford's, nor the DOE's...   ------------------------------  # Date: Tue, 23 Jan 2001 04:02:17 GMTt+ From: rjordan@mars.mcs.net (Richard Jordan)oC Subject: Re: [Q]: Resetting privileged password on a DECserver 200? 0 Message-ID: <dn7b6.161$9K.3928@news.goodnet.com>   Ken,H      if I recall correctly, you need to do a hard reset on the DECserverF 200 to get back to the factory passwords; this will also reset all theF server and port settings to default.  Press and hold the red button onD the back of the server in while powering the server on.  I think yoyE need to hold it in until the POST completes, but I'm not certain.  IfdB you have a terminal on port 1, 9600,N,8,1, it may tell you what it	 is doing.d  * I assume you know the default passwords...   Rich Jordanf rjordan@mcs.nete   ------------------------------  # Date: Tue, 23 Jan 2001 04:52:57 GMTm From: LBohan@dbc.spam_less..comsC Subject: Re: [Q]: Resetting privileged password on a DECserver 200?t8 Message-ID: <d83q6toqdlfdvsblb9psl3vpsl885j6r71@4ax.com>  = On 22 Jan 2001 17:34:39 PST, Fairfield@SLAC.Stanford.EDU (Ken 4 Fairfield; SLAC: 650-926-2924; FAX: 926-3515) wrote:  I >        We have a pair of  old  DECserver  200's that are clunking alongtI >    just fine.  Unfortunately, it's been so long since we've messed witheI >    them, no one remembers the privileged password.  Not the maintenance C >    password, mind you, that one is stored in the DECNET database.o >cI >        What is the procedure to set  a new password?  It's very unclearfI >    to  me,  reading  the doc set that came with TSM 2.1 (the  docs  are I >    V1.6, BTW), whether when you do a USE SERVER  server-name,  followediI >    by  SET PRIVILEGED, just _which_ password is required.  Indeed, whenrI >    I do a SHOW SERVER ALL CHARACTERISTICS,  both a login password and aoI >    maintenance  password are listed, but neither one is accepted by SETiA >    PRIVILEGED (nor are the default strings for those accepted).a >a0 >        Any hints would be greatly appreciated! >  >            Thanks, Ken  $ Well..., there's always brute force.   Yank the power cord. o  , Hold down the square red (S1) button, while  plugging back in.  .  $ This will reset to factory defaults.   ------------------------------   End of INFO-VAX 2001.045 ************************