1 INFO-VAX	Sat, 17 Jul 2004	Volume 2004 : Issue 392       Contents:" Re: Audit Trail of Interactive DCL" Re: cluster_config extreme badness DECnet question @ Do customers need to support DECforms in DECwindows using Motif?D Re: Do customers need to support DECforms in DECwindows using Motif?D Re: Do customers need to support DECforms in DECwindows using Motif?D Re: Do customers need to support DECforms in DECwindows using Motif?D Re: Do customers need to support DECforms in DECwindows using Motif?D Re: Do customers need to support DECforms in DECwindows using Motif?D Re: Do customers need to support DECforms in DECwindows using Motif?1 Re: HSx ECB (cache battery) troubleshooting/specs  Re: Is a file open Re: Is a file open Re: Is a file open7 Re: Is it possible to route or tunnel the SCS protocol? 7 Re: Is it possible to route or tunnel the SCS protocol? 7 Re: Is it possible to route or tunnel the SCS protocol? 7 Re: Is it possible to route or tunnel the SCS protocol?   Re: LOGIN.COM comparison utility7 Re: Node A dismounting its system disk from other nodes 7 Re: Node A dismounting its system disk from other nodes 6 Re: OpenVMS Management Station -  cluster storage ????. Re: OpenVMS Marketing: Variations on "stealth". Re: OpenVMS Marketing: Variations on "stealth" Pathwork problems 1 Re: Turn-key OpenVMS E-mail, web server solution? 1 Re: Turn-key OpenVMS E-mail, web server solution? % Re: Useful 'phpbb' installation guide ) Re: VAX Users See the Writing on the Wall ) Re: VAX Users See the Writing on the Wall ) Re: VAX Users See the Writing on the Wall   F ----------------------------------------------------------------------  % Date: Fri, 16 Jul 2004 22:08:31 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>+ Subject: Re: Audit Trail of Interactive DCL + Message-ID: <40F8982F.70F6D852@comcast.net>   
 dooley wrote:  > f > lsk55@hotmail.com (Scott) wrote in message news:<926edf3b.0407141758.643c0457@posting.google.com>...J > > I have a command procedure that simulates the DCL command line prompt,J > > logging everything that the user types.  It works great, except that II > > lose up-arrow and RECALL.  I don't want to use ACCOUNTING because the K > > information I want to track is really pretty minor and I don't want the K > > the overhead.  I don't want to use a pseudo-terminal because it creates K > > another process (I've got over 2,000 users on a single Alpha.)  But I'd . > > sure like to get up-arrow and RECALL back! > > J > > So does anyone have any ideas about how I can track just what is typedI > > in at the DCL prompt in interactive mode?  (Is there a hook I can use  > > in the CLI or something?)  > > B > > Thank you for taking the time to consider my problem.  --ScottI > There was some freeware called PHOTO in the supervisor series software, H > however it was VAX only - I don't know if an alpha port was ever done.   Can the VAX version be VESTed?   D.J.D.   ------------------------------  % Date: Fri, 16 Jul 2004 21:39:47 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>+ Subject: Re: cluster_config extreme badness + Message-ID: <40F89172.1DBD7220@comcast.net>    Michael Moroney wrote: > 6 > David J Dachtera <djesys.nospam@comcast.net> writes: >  > >Something that bit me once: > I > >I had a system disk where [SYS0.SYSCOMMON] was the common tree's start I > >point, and [VMS$COMMON] was the alias. Even my MOVEROOT kit would have  > >clobbered that setup. > I > What does CLUSTER_CONFIG.COM do wrong if the backlink/alias is mis-set?   F Y'know, I didn't look to see if it even checks that. Based on Lurker'sE observation, it looked like it doesn't bother to check, just assumes.   H Actually, I looked at my MOVE_ROOT kit and REMV_ROOT.COM does check that= the DID of SYSCOMMON.DIR for that root matches the FID of the E [000000]000000.DIR for that volume. If not, it complains, then pukes.    D.J.D.   ------------------------------  # Date: Sat, 17 Jul 2004 05:00:37 GMT 7 From: Jay Olson <jay.olson@triton-software.com.no.spam>  Subject: DECnet question: Message-ID: <40F8B2A2.9050609@triton-software.com.no.spam>  I I have a question concerning DECnet task-to-task communications. We have  H an application which uses DECnet task-to-task communication to start up I and communicate with some other programs. These programs in turn talk to  E other systems with TCP/IP. On some days, these programs get busy for  G some reason (most likely TCP/IP network problems) and they may not get  6 around to reading from the DECnet channel for a while.  D It appears that the main program can do a $QIOW to send data to the E TCP/IP communication programs and the write will complete before the  G other end has read the data. In fact, the $QIOW will copmplete even if  H the other end doesn't have a read active. However, this can only happen I a limited number of times, after which the $QIOW does not complete until  " a read is issued on the other end.  B My question is what AUTHORIZE, SYSGEN, DECnet or other parameters I control how many DECnet writes can be made without the other end reading  H the data. The amount of data is definitely a factor. More writes can be I done if the amount written is smaller. BYTLM is definitely not a factor.  C   I adjusted some DECnet OSI TRANSPORT parameters (MAXIMUM RECEIVE  - BUFFERS and MAXIMUM WINDOW SIZE) to no avail.   L Any answers, hints, or pointers to appropriate manuals would be appreciated.   	- Jay Olson 	Triton Software group Ltd.    ------------------------------  % Date: Fri, 16 Jul 2004 15:39:28 -0400 7 From: "Rick McLaughlin" <rick.mclaughlin@nospam.hp.com> I Subject: Do customers need to support DECforms in DECwindows using Motif? * Message-ID: <40f83198@usenet01.boi.hp.com>  @ Do customers need to support DECforms in DECwindows using Motif?      J The DECwindows Motif support on the latest version of DECforms V3.3 is not working on Alpha.   L After doing some testing it seems that it has been broken since the DECforms V3.0. So this functionality   ' has not been working for about 4 years.       I There are two interfaces for DECforms: "Character Cell" forms and "Motif"  forms. "Character Cell" works   E and the "Motif" interface does not. Today users can have any terminal  emulator within the DECwindows  I environment (eg. DECterm) and get the "Character Cell" form to work. Just  like it works for VT terminals.   K A forms which is not defined as a "motif" (or "Pixel")  layout will work in  "character cell" mode under any   K terminal emulator under DECwindows or even any terminal emulator from a PC.       J The user defined DECforms-DECwindows logical which evokes a separate Motif window for a form itself   does not work.      L Your feedback will be helpful in determining whether the engineering work is done for DECforms I64.      
 Thank you,       Rick McLaughlin   ! DECforms Product Business Manager    Rick.McLaughlin@hp.com   ------------------------------  % Date: Fri, 16 Jul 2004 13:11:13 -0700 + From: "Barry Treahy, Jr." <Treahy@MMaz.com> M Subject: Re: Do customers need to support DECforms in DECwindows using Motif? ' Message-ID: <40F83661.9000300@MMaz.com>    Rick McLaughlin wrote:  A >Do customers need to support DECforms in DECwindows using Motif?  >  >  > K >The DECwindows Motif support on the latest version of DECforms V3.3 is not  >working on Alpha. > M >After doing some testing it seems that it has been broken since the DECforms  >V3.0. So this functionality > ( >has not been working for about 4 years. >  >    > @ My 2 cents, if it has been busted for four years and no one has I complained, it is probably a strong sign that it is dead.  We use DF but  7 have no need, or ever plan on needing, Motif support...    Barry    --    > Barry Treahy, Jr                       E-mail: Treahy@MMaz.com> Midwest Microwave                          Phone: 480/314-1320> Vice President & CIO                         FAX: 480/661-7028                            ------------------------------    Date: 16 Jul 2004 17:07:45 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen) M Subject: Re: Do customers need to support DECforms in DECwindows using Motif? 3 Message-ID: <ZobRlPGbxI14@eisner.encompasserve.org>   d In article <40f83198@usenet01.boi.hp.com>, "Rick McLaughlin" <rick.mclaughlin@nospam.hp.com> writes:  L > The DECwindows Motif support on the latest version of DECforms V3.3 is not > working on Alpha.     N > Your feedback will be helpful in determining whether the engineering work is > done for DECforms I64.    E There are lots of VMS customers who are running on VAX and have never  converted to Alpha.   B HP has been trying to convince the world that Itanium will be more cost-effective than Alpha.  C Penalizing those potential Itanium customers who never ran on Alpha  does not seem a good move.   ------------------------------  % Date: Fri, 16 Jul 2004 16:13:18 -0700 * From: "Jack Peacock" <peacock@simconv.com>M Subject: Re: Do customers need to support DECforms in DECwindows using Motif? 2 Message-ID: <rvqdnXK36-2S_GXdRVn-jg@mpowercom.net>  : "Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message- news:ZobRlPGbxI14@eisner.encompasserve.org... E > Penalizing those potential Itanium customers who never ran on Alpha  > does not seem a good move. > I Not supporting features that no one uses is hardly a penalty.  Four years K and no complaints is a pretty good indicator that DecForms on DecWindows is H not a significant roadblock to an Itanium upgrade.  The last time I usedG DECforms was on an 11/750, VMS V3, as a front end to DBMS-32.  That's a  whole lot of obsolete.   Jack Peacock   ------------------------------    Date: 16 Jul 2004 19:39:55 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen) M Subject: Re: Do customers need to support DECforms in DECwindows using Motif? 3 Message-ID: <fynimODebyDw@eisner.encompasserve.org>   _ In article <rvqdnXK36-2S_GXdRVn-jg@mpowercom.net>, "Jack Peacock" <peacock@simconv.com> writes: < > "Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message/ > news:ZobRlPGbxI14@eisner.encompasserve.org... F >> Penalizing those potential Itanium customers who never ran on Alpha >> does not seem a good move.  >>K > Not supporting features that no one uses is hardly a penalty.  Four years M > and no complaints is a pretty good indicator that DecForms on DecWindows is 4 > not a significant roadblock to an Itanium upgrade.  8 No complaints from anyone on Alpha (where it is broken).  E I presumed there was a DECforms on VAX where the DECwindows interface 5 worked and might be in active use without complaints.    > The last time I used? > DECforms was on an 11/750, VMS V3, as a front end to DBMS-32.   G I was more of an FMS person, because I started before DECforms existed.    ------------------------------  % Date: Fri, 16 Jul 2004 21:48:22 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> M Subject: Re: Do customers need to support DECforms in DECwindows using Motif? , Message-ID: <40F88565.901906C3@teksavvy.com>   Jack Peacock wrote: K > Not supporting features that no one uses is hardly a penalty.  Four years M > and no complaints is a pretty good indicator that DecForms on DecWindows is 4 > not a significant roadblock to an Itanium upgrade.  J No, au contraire. It is a pretty good indication that people who are usingJ DECforms are still probably on older versions, stable environments and notL lokely to upgrade VMS to Alpha or IA64 because their plans are to migrate to another platform.   J Also, it is my impresison thatr FMS is still far more in use than DECformsI ever was. DECforms appeared more or less at a time when Palmer started to ; sabotage the company and never had much time to take roots.     G Now, if you could make FMS run native in MOTIF, that would be way cool.    ------------------------------  % Date: Fri, 16 Jul 2004 22:33:56 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>M Subject: Re: Do customers need to support DECforms in DECwindows using Motif? + Message-ID: <40F89E24.F2F2F8E5@comcast.net>    Rick McLaughlin wrote: > B > Do customers need to support DECforms in DECwindows using Motif? > L > The DECwindows Motif support on the latest version of DECforms V3.3 is not > working on Alpha.  > N > After doing some testing it seems that it has been broken since the DECforms > V3.0. So this functionality  > ) > has not been working for about 4 years.  > K > There are two interfaces for DECforms: "Character Cell" forms and "Motif"  > forms. "Character Cell" works  > G > and the "Motif" interface does not. Today users can have any terminal   > emulator within the DECwindows > K > environment (eg. DECterm) and get the "Character Cell" form to work. Just ! > like it works for VT terminals.  > M > A forms which is not defined as a "motif" (or "Pixel")  layout will work in ! > "character cell" mode under any  > M > terminal emulator under DECwindows or even any terminal emulator from a PC.  > L > The user defined DECforms-DECwindows logical which evokes a separate Motif > window for a form itself >  > does not work. > N > Your feedback will be helpful in determining whether the engineering work is > done for DECforms I64. >  > Thank you, >  > Rick McLaughlin  > # > DECforms Product Business Manager  >  > Rick.McLaughlin@hp.com   My $0.02...   E As noted in another thread, my take is if it's worth doing, its worth  doing right.  / A product that is broken may as well not exist.   E I don't see that, however, as an excuse for not fixing it. In fact, I E see it as business opportunity to capture business that OpenVMS would  other lose to other systems.  G Straight up: I'd even partner with WRQ to advance the degree of synergy ! between your products and theirs.   G I realize very few people share this viewpoint, but it works for me and , that's enough to motivate me to stand by it.   D.J.D.   ------------------------------  % Date: Fri, 16 Jul 2004 22:17:29 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>: Subject: Re: HSx ECB (cache battery) troubleshooting/specs+ Message-ID: <40F89A49.79DE2D76@comcast.net>    Duncan Brown wrote:  > G > (Self-reply...)  I dissected one of the battery packs more thoroughly G > and did some more googling based on what I found, and I now know that ; > indeed those batteries are below-spec and need replacing.  > = > Each "battery" is really two Hawker Cyclon 2V 5.0AH X Cells D > shrinkwrapped together and wired in series to produce a nominal 4VF > stick.  The part number is 0800-0004, and the cheapest place I could > find them online is  > A > http://www.gotbatteries.com/ProductPage.asp?ProductNum=37L142S1  > G > The manufacturer's website is hard to track down but here is the page  > for these batteries: > . > http://www.enersysreservepower.com/cyc_b.asp > E > Note the application manual .pdf link at the bottom, that gives the J > specs that show that 4.15V is simply not enough for a 2-cell stick to beJ > called well-charged.  It also mentions the life characteristics on theseH > things and even if DEC designed their charging circuit perfectly theseH > things would only be good for a few hundred discharge/recharge cycles.L > I guess it becomes clear why they don't last more than a few years in use. > E > I went ahead and ordered 6 batteries to get the price break - in an G > unused state they have a long shelf life, so I'll be all set the next I > time they die.  Now I just get to have the fun of trying to rebuild the J > stick of two batteries, keeping the shrink tubing thin and smooth enough? > to be able to fit the stick back in the frame.  Wish me luck!   H Good luck. Rumor has it that cache batteries turn to mush if they're not1 kept charged, ragardless of the voltage readings.    D.J.D.   ------------------------------    Date: 16 Jul 2004 14:01:29 -0700" From: cstranslations@msn.com (Joe) Subject: Re: Is a file open < Message-ID: <d56d1c2d.0407161301.4eceb60@posting.google.com>  _ Roy Omond <Roy.Omond@BlueBubble.UK.Com> wrote in message news:<2lpmd6FfahndU1@uni-berlin.de>...  > David Froble wrote: k > > cstranslations@msn.com (Joe) wrote in message news:<d56d1c2d.0407150630.24c5e883@posting.google.com>...  > > ( > > The post by Hein seems the simplest. > G > This one crops up every so often, and Hein gives the same answer time  > and time again.  > D > Not only is Hein's suggestion the simplest, but it's also the only< > (clean) solution which does not introduce race conditions.  E To answer the "what is this for" question - roughly files come into a F system that decrypts it, the dumps them on the alpha and then they areC to get picked up and processed. Of course it shouldn't be picked up 4 for processing if the file is still "coming across."  D From time to time I forget the (painfully) obvious solution; in this, case, just try and open it exclusive access.   Thanks to all.   ------------------------------  % Date: Fri, 16 Jul 2004 23:53:46 -0400   From: John Santos <JOHN@egh.com> Subject: Re: Is a file open 6 Message-ID: <1040716233523.16558B-100000@Ives.egh.com>   On 16 Jul 2004, Joe wrote:  a > Roy Omond <Roy.Omond@BlueBubble.UK.Com> wrote in message news:<2lpmd6FfahndU1@uni-berlin.de>...  > > David Froble wrote: m > > > cstranslations@msn.com (Joe) wrote in message news:<d56d1c2d.0407150630.24c5e883@posting.google.com>...  > > > * > > > The post by Hein seems the simplest. > > I > > This one crops up every so often, and Hein gives the same answer time  > > and time again.  > > F > > Not only is Hein's suggestion the simplest, but it's also the only> > > (clean) solution which does not introduce race conditions. > G > To answer the "what is this for" question - roughly files come into a H > system that decrypts it, the dumps them on the alpha and then they areE > to get picked up and processed. Of course it shouldn't be picked up 6 > for processing if the file is still "coming across." > F > From time to time I forget the (painfully) obvious solution; in this. > case, just try and open it exclusive access. >  > Thanks to all.  E Aha!  The old "someone's sending me a file, and I don't want to start 4 processing it until I have the whole thing" problem!  A Over the years, I've seen two approaches to this.  One is for the @ sender to send the file with a temporary name and then rename itA when finished.  The local process doesn't look for the temp name, F it looks for the final name.  This method assumes the rename operationA is atomic (i.e. you will either see the original temp name or the C final name, but not both), and when you see the final name, you can A open it and process it without restriction (i.e. that there isn't A a little window during the rename operation where you can see the @ file with its final name, but you can't access it yet.)  The wayA renames work on VMS, this seems to be true.  I think RENAME takes A out an exclusive lock on the directory while messing with it, and B releases the lock only when done, so if you try to access the fileE during the rename, you will wait until the directory lock is released E and everything will be cool.  However, I haven't seen this documented C anywhere.  I've never seen it fail in a way that indicates it might  be a problem, though.   B One feature of this method is that if either system crashes during? the transfer (or the network goes down, etc.), you won't see an D incomplete file in the destination directory, but you will see staleE partial temp files there, so you may need to have an automated method . to clean them up (if this happens frequently.)  ? The second method is to send a flag file after sending the data ? file.  Don't process the data file until you see the flag file. = When done, delete both.  The main problem with this method is @ people who just don't get the concept and send you the flag fileA before sending the data file!  Or they'll try to make the process > more efficient by sending both the data file and the flag file> in parallel!  If you don't control both ends, I guarantee this? will happen sooner or later, and it may be hard to diagnose :-(   A (These issues apply whether it is using DECnet/COPY, FTP, NFS, or   any other file transfer method.)   HTH    --   John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Sat, 17 Jul 2004 00:13:36 -0400 4 From: "Hein RMS van den Heuvel" <hein_at_hp_dot_com> Subject: Re: Is a file open , Message-ID: <40f8aa13$1@usenet01.boi.hp.com>  G > To answer the "what is this for" question - roughly files come into a H > system that decrypts it, the dumps them on the alpha and then they areE > to get picked up and processed. Of course it shouldn't be picked up 6 > for processing if the file is still "coming across." > F > From time to time I forget the (painfully) obvious solution; in this. > case, just try and open it exclusive access.  E What if the transfer fails mid-way. The file might no longer be open, C and you might erroneously conclude that it is ready for processing. F For that case I would prefer a rename solution. Call it xxx.tmp during trnsfer.6 When and if the transfer completes, rename to xxx.dat.   fwiw,  Hein.    ------------------------------  % Date: Fri, 16 Jul 2004 14:21:26 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> @ Subject: Re: Is it possible to route or tunnel the SCS protocol?, Message-ID: <40F81C9F.94CFDDE8@teksavvy.com>   "Stanley F. Quayle" wrote:G > The problem is that they haven't been able to make the speed of light G > any faster.  That was the source of the limit on clustering distance.   K Back in 1992, when I setup a fault tolerant site, it was just simple 10mbps L ethernet bridged over fibre with cheap bridges (not under my control) with aM couple KM distance between buildings. And this was fully supported (even with 9 the then brand spanking new host based volume shadowing).   B Now, once you're on the fibre, it doesn't really matter if you areM tunnelled/encapsulated or not, you travel at same speed. The difference is at K the "box" level  where the boxes receive an ethernet packet and decide what  needs to be done to it.   G Limiting discussion to the 2 routers next to the 2 VMS hosts, would the G encapsulation process inside those routers be done quickly enough using P today's hardware that it would be "transparent" to the SCS protocol robustness ?  N I realise that if you are tunneling over an IP network consisting of many hopsJ and agregated traffic from many locations as well as bursty trtaffic, then you're askin for trouble.   I But would the mere fact of encapsulating SCS packets in IP packets over a  reliable IP link make it fail ?    ------------------------------  % Date: Fri, 16 Jul 2004 20:17:45 -0400 " From: Glenn Everhart <gce@gce.com>@ Subject: Re: Is it possible to route or tunnel the SCS protocol?4 Message-ID: <40f87080$0$10382$61fed72c@news.rcn.com>   JF Mezei wrote:  > "Stanley F. Quayle" wrote: >  > C >>The big issue is performance -- if there's too much delay, you'll C >>have to adjust the clustering SYSGEN parameters.  And, of course, ; >>this is completely unsupported by HP OpenVMS Engineering.  >  > I > If an all mighty Microvax II is capable of participating in an ethernet I > cluster, wouldn't modern routers be able to get SCS packets to a remote C > destination faster than it takes the MVII to process the packet ? A There have been reports of clusters working over IP and even over K SCSI busses. However, cluster communications is designed so that everything J connects to everything else. Nothing in there knows about handling routingE failures, and connection failures better be reported at the same time  if something fails.   C There are not a lot of examples of drivers that connect SCS either, C but it has been done to run the protocol over others.  I don't know C what presumptions there are for bandwidth; certainly 10MB is plenty A but at some point the fact that the initial design center was for * pretty decent bandwidth may assert itself.  > I have no further references at this point though. Some of theB mentions at DECUS symposia didn't even make it into session notes.   Glenn Everhart   ------------------------------  % Date: Fri, 16 Jul 2004 22:25:43 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>@ Subject: Re: Is it possible to route or tunnel the SCS protocol?+ Message-ID: <40F89C36.FAE96119@comcast.net>    JF Mezei wrote:  >  > Colin Butcher wrote:O > > If you want a reliable cluster - then no, you can't. SCS over ethernet type N > > networks is a layer 2 protocol that requires a high bandwidth, low latencyL > > (that's really important) datalink. Tunnelling etc. simply introduce too% > > much latency, so it doesn't work.  > L > Is there much difference in latency between  early 1990s  ethernet bridgesJ > (with fibre between 2 bridges), and a 2004 era high performance router ?   My take:  5 Early 1990's processor speeds <100MHz, Ethernet = 3Mb   - 2004 processor speeds >500MHz, Ethernet = 1Gb   : Yes, I'd expect to see significant differences in latency.  L > If bridging SCS and LAT with older slower bridges was acceptable, wouldn'tJ > today's networking hardware be able to perform tunneling with similar or > better performance ?  C If tunnelling did not introduce additional overhead, I'd venture to F guess "perhaps". I'd want to see it running reliably under one helluvaH load for an extended period (some tens of months) before I'd dare put it into production, though.  E I'd worry more about packet loss than I would about latency, however.    My $0.02...    D.J.D.   ------------------------------  + Date: Sat, 17 Jul 2004 04:59:06 +0000 (UTC) 7 From: moroney@world.std.spaamtrap.com (Michael Moroney) @ Subject: Re: Is it possible to route or tunnel the SCS protocol?( Message-ID: <cdabmq$ee4$2@pcls4.std.com>  * "Ralf van Diesen" <Ralf@nospam.nl> writes:  M >I am trying to figure out if it's possible to tunnel the SCS protocol, since < >SCS can not be routed I must look for a different solution., >Is there anybody who can help me with this?  F It is certainly possible to create a box to encapsulate SCS over some G other protocol with low enough latency to work.  So much depends on the L latency/reliability of whatever network you route it over.  IP is often bad.  J I think a handful of network boxes know what to do with SCS.  I have heardF of people running clusters over T1 lines in the early 90s (unsupportedI but that's because the minimum supported bandwidth is 10Mb Ethernet)  DEC D sold products that ran ethernet over things like microwave links for' multisite clusters, among other things.    --   -Mike    ------------------------------  % Date: Fri, 16 Jul 2004 22:14:26 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>) Subject: Re: LOGIN.COM comparison utility + Message-ID: <40F89992.1495C524@comcast.net>    John Brandon wrote:  >  > Before I go write my code... > N > Does anyone have a utility that will scan all LOGIN.COM on the disks looking- > for common characteristics?  Symbols (etc).   A I did that in DCL about eight or nine years ago. I may be able to G resurrect the code from the many archives I've kept, if you REALLY want  it.   B It was, however, based on a great many extremely brash assumptionsB determined by the needs of the 3rd-party app. it was meant to help manage.   O > I have a legacy system that some dev/admin had the great idea to add the same ( > symbols (etc) to each users LOGIN.COM.  " Typical "cut and paste" mentality.  O > My mission is to go out and consolidate these symbols (etc) into the SYLOGIN.    PLEASE!!! NOT SYLOGIN!!!  G If you must consolidate, put them in a separate .COM file and invoke it F in LOGIN, perhaps conditionally based on whatever criteria make sense.  D Best not to clog up SYLOGIN with application specific stuff (not all users will need it).   D.J.D.   ------------------------------  % Date: Fri, 16 Jul 2004 10:54:25 -0700 , From: Ken Fairfield <My.Full.Name@intel.com>@ Subject: Re: Node A dismounting its system disk from other nodes( Message-ID: <40F81651.8070700@intel.com>   JF Mezei wrote:   I > At boot time, node A gets its system disk mounted automatically, but in N > systartup_VMS, it does a MOUNT/CLUSTER of this drive so that it gets mounted% > on all others nodes in the cluster.  > M > What I would like to do is to dismount that drive from all other nodes when P > node A performs an orderly shutdown. However, issuing a DISMOUNT/CLUSTER failsO > because the drive is still in use by NODE A even if other nodes would be able  > to dismount it.  > N > Is there a clean way to have NODE A issue a command which will result in all& > other nodes dismounting that drive ? >  > All I can think of is  > $! > $MC SYSMAN > SET ENV/CLUSTER  > DO DISMOUNT $2$DKA200  > EXIT > $! > M > Would this work properly ? Is there a cleaner way of doing this (some magic # > switch in the DISMOUNT command ?)   % How about this (wrapped for clarity):    $ MCR SYSMAN Set E/C 9 Do If F$TrnLnm("SYS$SYSDEVICE") .Nes. "$2$DKA200:" Then -  _SYSMAN> Dismount $2$DKA200: Exit $   = Verify the exact string returned by translating SYS$SYSDEVICE  on NODE A first...   	-Ken  --  6 I don't speak for Intel, Intel doesn't speak for me...  
 Ken Fairfield ! D1C Automation VMS System Support " who:   kenneth dot h dot fairfield where: intel dot com   ------------------------------  % Date: Fri, 16 Jul 2004 14:24:44 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> @ Subject: Re: Node A dismounting its system disk from other nodes, Message-ID: <40F81D64.25E6A4DB@teksavvy.com>   Ken Fairfield wrote: > $ MCR SYSMAN	 > Set E/C ; > Do If F$TrnLnm("SYS$SYSDEVICE") .Nes. "$2$DKA200:" Then -  > _SYSMAN> Dismount $2$DKA200: > Exit  M Thanks. Hadn't though of putting "DCL" in DO commands, but you are right. The 2 above is far superior to my initial "DO DISMOUNT".   ------------------------------  % Date: Fri, 16 Jul 2004 21:35:33 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>? Subject: Re: OpenVMS Management Station -  cluster storage ???? + Message-ID: <40F89075.47194A22@comcast.net>    Ken Fairfield wrote: > " > Peter 'EPLAN' LANGSTOEGER wrote: > V > > In article <40F53285.2040406@uiowa.edu>, Rick Dyson <rick-dyson@uiowa.edu> writes: > > M > >>I have the dual HSG80 setup too, and I have v8.7F-4 on my HSG80s.  I have 1 > >>not checked recently to see if -4 is current.  > >  > > Q > > According to http://www.compaq.com/products/storageworks/softwaredrivers/acs/ + > > the current version seems to be V8.7F-7  > >  > > Time for an update ;-) > @ >      Probably not...  While it's a bit hard to get the releaseA > notes for any of the patch levels to ACS, we stopped at V8.7F-3 ? > which has a fix for a potential controller crash when running 5 > FRUTIL (I think) to replace a redundant controller.  > ? >      OTOH, V8.7-7 BREAKS HSZTERM!  I.e., you lose the ability < > to $Set Host/Scsi $1$DGAxxx.  Also, near as we could tell,= > none of the fixes in -4 through -7 corrected anything for a ; > VMS environment with a dedicated SAN.  A lot of the fixes < > apply to Windows environments with zoning and such like... > E >      We've asked that the next point release (V8.8F or ???) restore A > the ability to use HSZTERM, but we don't have a commitment yet.   B I wrote to Ken privately, but I'll post here to the group as well:  H You really shouldn't SET HOST/SCSI to a DGA device - use the GGA device,H instead. For HSZs, use the virtual console device, not any of the active disk or tape devices.   G Here's a bit of DCL to help you find which GGA is associated with which  DGA:   $ @GGFIND 1    ...or...   $ @GGFIND $1$DGA1    $ ty [exe]ggfind.com   $ on warning then goto err_exit  $ say := write sys$output $ $ if    f$type( p1 ) .eqs. "INTEGER" $ then( $       p1 = f$FAO( "$1$DGA!UL:", 'p1' )  $       say f$fao( "(!AS)", p1 ) $ endif & $ if    .not. f$getdvi( p1, "exists" ) $ then $       say f$mess( %x90a ) ! $       say f$fao( " \!AS\", p1 )  $       exit %x1000090a  $ endif 	 $gg_loop:   $ devc = f$device( "_$1$GGA*:" )) $ if devc .eqs. "" then goto exit_gg_loop ! $ wwid = f$getdvi( devc, "wwid" )  $ gosub rd_wwid  $ &ggseg = devc  $! sh sym &ggseg $ goto gg_loop $exit_gg_loop: $ wwid = f$getdvi( p1, "wwid" )  $ gosub rd_wwid  $ say &ggseg $ exit $!
 $err_exit: $ sh sym/all $ exit $!	 $rd_wwid: ' $ elem_0 = f$elem( 0, "-", wwid ) + "-"  $ rdstr = wwid - elem_0  $ ggseg = "GG_" - (         + F$ELEM( 0, "-", rdstr) + "_" -(         + F$ELEM( 1, "-", rdstr) + "_" -          + F$ELEM( 2, "-", rdstr) $ return   D.J.D.   ------------------------------  % Date: Fri, 16 Jul 2004 18:45:34 -0400g# From: "John Smith" <a@nonymous.com>o7 Subject: Re: OpenVMS Marketing: Variations on "stealth"o, Message-ID: <FNydnaVCP64Hx2Xd4p2dnA@igs.net>   JF Mezei wrote:n > Michael Austin wrote:sE >> who cares -- as long as they are doing it, it can't be bad news...n > > > I have a friend in paliative care. Whenever her pain "breaksD > through", the nurse gives her some more pain killer. This way, herE > last weeks have been fairly "good" in terms of quality of life, butn7 > the underlying cancer has continued to eat her bones.S >CE > VMS is in a very similar situation, except the people here stronglyfF > believe that its disease can be cured. However, current and previousE > owners of VMS have only given VMS token amounts of drugs to keep it F > alive and shutting out the pain (the complainers), instead of fixing+ > the cancer that Palmer injected into VMS.  > F > So far, under HP, I have seen nothing but paliative care of VMS withA > token mentions whenever the "pain" was too loud (pain being theoC > complain.os.vms voices, and I believe it was Marcello that coinedn > that expression).i > B > And just as my friend exceeded all expectations and has survived@ > months instead of days since they stopped attempts to cure herF > cancer, VMS has exceeded all expectations and survived over a decadeE > despite all attempts at euthanasia (Palmer, Capellas and Stallard).  >r >eD > Yes, on the surface, those mentions of VMS are good and one cannotF > fault them. However, the question remains if whether VMS is still inD > paliative care or whether HP is on a solid plan to stop the cancerD > and rebuild VMS into a strong and healthy product that is not onlyE > profitable for HP, but also allowed to compete in the "IT olympics" ) > against all brands of Unix and Windows.e     Good analogy JF.  ! I hope your friend pulls through.    ------------------------------  % Date: Fri, 16 Jul 2004 21:46:57 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>7 Subject: Re: OpenVMS Marketing: Variations on "stealth"d+ Message-ID: <40F89321.F5303394@comcast.net>    JF Mezei wrote:t >  > David J Dachtera wrote:a > >dL > > With OpenVMS getting exposure in the trade rags (or at least the on-lineI > > versions there of), perhaps the marketing folks at hp know more aboutAH > > stretching their advertising dollars than we give them credit for... > P > Nop. It is probably Sue who found a loophole against the prohibition to marketO > VMS and is able to get VMS mentioned without anything tracing it back to her.   G O.k. I guess I should explain. Perhaps the subtle humor was misapplied.m  H Once upon many moons ago, around about the wonderous days of yore, thereG was a Bugs Bunny cartoon wherein the Easter Bunny - rather a "sad sack" H character - dupes Bugs into delivering Easter Eggs. Then, in an aside toG the "audience", he explains, "Every year I get some dumb bunny to do mya
 work for me."n  G So, I was casting "us" (those who try to contribute to OpenVMS's marketDH exposure, though I myself have yet to make a documentable impact) as theH "Bugs Bunny" types, and hp's marketeers as the ones using other people's1 resources (which is a common method of leverage).S  @ To carry the "joke" a bit further, later on in the cartoon, BugsD confronts him again when he tells Bugs, "You've got to deliver a lotE more eggs! You'll give the Easter Rabbit a bad name", where upon Bugsm< replies, "I already have a bad name for the Easter Rabbit!".  A (Insert an appropriate, if scathing comment here, if you like...)o   D.J.D.   ------------------------------   Date: 17 Jul 2004 03:56:45 GMT) From: tomarsin2015@aol.com (Tomarsin2015)D Subject: Pathwork problems: Message-ID: <20040716235645.11545.00001461@mb-m11.aol.com>   HellosJ I am using a 3100-80 running 7.3-1 and Pathworks V6.0D as a primary domainO controller for a couple of Win2000 pcs. The problem is I have one user and only K one user that cannot be validated by the controller. Every other user worksoJ okay. I have deleted the user and then added the user back with no luck. IK tried having the user type his password and username in upper/low case, trytJ loggin on from a different pc, but no luck. Any thoughts would be helpful. tks    ------------------------------  # Date: Fri, 16 Jul 2004 20:08:35 GMTo1 From: Michael Austin <maustin@firstdbasource.com> : Subject: Re: Turn-key OpenVMS E-mail, web server solution?< Message-ID: <7BWJc.2810$tn5.2161@newssvr22.news.prodigy.com>   jf.pieronne@laposte.net wrote: <snip>  ( > Found webcalendar, from the home page:I > WebCalendar is a PHP-based calendar application that can be configured eK > as a single-user calendar, a multi-user calendar for groups of users, or p, > as an event calendar viewable by visitors. > % > http://webcalendar.sourceforge.net/i > ' > Install and run out of the box on VMSw >  > I have build a demo system:h1 > http://www.pi-net.dyndns.org/calendar/index.phpb > @ > use demo/demo as username/password (may be remove at any time) >  <snip>  Q a bit plain but that can be easily fixed...  I would prefer that it did not rely iP on cookies, bu that too can be changed using sessions().  Yes, a bit more work, P but not a bad start.  And I like the one thing that even exchange has a problem & with -- search, export -- pretty cool.  P Again, Hey Bob, let me see you download, install and make this one work on your L beloved Purveyor - with no changes.  Don't get me [too] wrong, Purveyor was H great - when all you needed was a simple server.  Things are a bit more  complicated now...   -- M Michael Austin.D Consultant - Available.h@ Donations welcomed. Http://www.firstdbasource.com/donations.html :)   ------------------------------  % Date: Fri, 16 Jul 2004 21:51:17 -0500V2 From: David J Dachtera <djesys.nospam@comcast.net>: Subject: Re: Turn-key OpenVMS E-mail, web server solution?+ Message-ID: <40F89425.C99097C5@comcast.net>w  
 dooley wrote:4 > g > David J Dachtera <djesys.nospam@comcast.net> wrote in message news:<40F49CCD.5CA7E6AA@comcast.net>... J > > Is anyone aware of a company prodiving a turn-key OpenVMS solution forK > > e-mail, web services and/or both? I thought I followed a link on GoogleeJ > > Groups to one once, but I didn't print anything out from that site and > > don't a URL to go back to. > Check out this one > http://www.versamail.net.au/ > Phil  ' I'm looking for software, not services.n   D.J.D.   ------------------------------   Date: 16 Jul 2004 18:03:46 GMT From: healyzh@aracnet.comc. Subject: Re: Useful 'phpbb' installation guide, Message-ID: <cd95a2023f2@enews1.newsguy.com>  / dieter rossbach <dieter.rossbach@gmx.de> wrote:nH > Porting phpBB to VMS was mit really a problem, I did that a while ago, > but C > I still have problems with MySQL currupting database tables. This-H > happens as soon as I add a new forum in phpBB. I have this problem forH > month now, with MySQL 4.0.16 ,17, 18 and even with the new 4.1.3-beta.  7 I've got MySQL 4.0.20 and PHPBB 2.0.6 (I got PHPBB from H http://www.4ovms.dyndns.org/phpbb/) running on OpenVMS 7.3-2.  How can IH tell if I'm getting corruption or not?  I've setup several forums, but I  don't really have any posts yet.   		Zane   ------------------------------  % Date: Fri, 16 Jul 2004 15:29:54 -0500t/ From: Chris Scheers <chris@applied-synergy.com> 2 Subject: Re: VAX Users See the Writing on the Wall3 Message-ID: <40F83AC2.6C6159E5@applied-synergy.com>f   "Barry Treahy, Jr." wrote: >  > jlsue wrote: > J > >>But still, even in this scenario, you are still running four instancesB > >>of C-V, which would still require that your apps can easily beL > >>distributed.  If the app is a multi-threaded app that requires localizedJ > >>multi-processor CPU horsepower, I don't think that either the Alpha or) > >>Windows C-V can presently solve that.8 > >> > >> > > M > >True, you'll still have the potential single-CPU-per-VAX-instance to worry H > >about.  My point is that you'll need fewer actual boxes to manage it.D > >This can often be less expensive than lots of Wintel/amd servers. > >x > >iO > >>>While "c." above is still somewhat true, you can probably get sufficientlyuP > >>>fast Alpha server such that a single CPU virtual VAX may be fast enough forM > >>>most VAX 78xx workloads.  And certainly a 32-way Marvel would be able tofF > >>>run enough images to reduce the total number of boxes you'd need. > >>>f > >>>o > >>>.G > >>Yes, but at what cost?  How many folks are running VAX because theyeI > >>could not afford the migration then, and/or now, or how many couldn'ttH > >>migrate because of their VAX apps or environment?  VUP per CPU, a PCL > >>running C-V will produce faster emulated results than Alpha and throwingK > >>a 32-way Marvel isn't a realistic migration solution for 99% of the VAX ? > >>shops where as $8k in PC's, well that is a different story.e > >> > >> > >fN > >I don't think that most of the servers I've seen implemented for productionJ > >use, including those under consideration for C-V, are in the $8K range.N > >And if it takes 24 of those $8K boxes to handle your workload (a real-worldF > >example I've seen), then 2 16-CPU alphas might actually be cheaper. > >i > >aB > Auh, perhaps you skimmed the early part of my post, but that $8kJ > represents four $2k boxes and those four boxes running C-V could produceF > the equivalent total VUPS of a 7840 where as two 16-way Alphas wouldH > still have to be running multiple instances of C-V to produce the sameJ > potential horse power and I would have to believe that Alpha iron is far4 > more expensive than a total of $8k in PC hardware. > ? > Just an side, I also mentioned the $2k price as one of my C-VrE > systems...  This is just an example of an entry-level hosting priceaH > that, if compounded by four systems, can reach the realm of processingE > you suggest for a 7840 but if you are actually running that type ofSH > load, most likely you would spend more on systems with better hardwareJ > redundancy and even if you reached $5k per box, four $5k boxes are still) > cheaper than the dual Alpha approach...     . Don't forget to add in the cost of C-V itself.  G And, if you require any special hardware, add in the cost of that.  (IfF it is even available.)  G -----------------------------------------------------------------------a$ Chris Scheers, Applied Synergy, Inc.  C Voice: 817-237-3360            Internet: chris@applied-synergy.com     Fax: 817-237-3074    ------------------------------  # Date: Fri, 16 Jul 2004 20:31:50 GMTw From: dittman@dittman.net 2 Subject: Re: VAX Users See the Writing on the Wall4 Message-ID: <WWWJc.508$lz2.145@nwrddc03.gnilink.net>  1 Stanley F. Quayle <squayle@insight.rr.com> wrote:44 > On 16 Jul 2004 at 3:28, dittman@dittman.net wrote:C >> Couldn't you use VMware ESX to run multiple instances of Windowsw, >> running multiple instances of CHARON-VAX? > F > You could, but that wouldn't produce something that would replace a & > single high-end multi-processor VAX.  / That was in answer to the limitation presented:t  D >>>Other things to consider for CHARON-VAX Windows vs OpenVMS Alpha: >>>u= >>>1.  On Windows, only one instance of C-V can run at a timet  D Using ESX would allow more than one instance of CHARON-VAX to run at a time.n  I >> The only reason I could think of what that wouldn't work is due to thesH >> USB dongle, but that would be a limitation that SRI could work around >> if necessary. > E > It's not a limitation.  I could see someone paying for one license e; > and then running a bunch of instances on a single system.m  H I believe with ESX only one of the virtual machines would have access toG the USB dongle, so the other virtual machines couldn't read the dongle.,  K Not that I'm a fan of dongles anyway, as they tend to get lost when repairs> or upgrades are done.m -- e Eric Dittman dittman@dittman.nets   ------------------------------  % Date: Fri, 16 Jul 2004 13:58:24 -0700h+ From: "Barry Treahy, Jr." <Treahy@MMaz.com>t2 Subject: Re: VAX Users See the Writing on the Wall' Message-ID: <40F84170.4070809@MMaz.com>n   Chris Scheers wrote:   >"Barry Treahy, Jr." wrote:f >  o >  >>jlsue wrote: >> >>     >>J >>>>But still, even in this scenario, you are still running four instancesB >>>>of C-V, which would still require that your apps can easily beL >>>>distributed.  If the app is a multi-threaded app that requires localizedJ >>>>multi-processor CPU horsepower, I don't think that either the Alpha or) >>>>Windows C-V can presently solve that.  >>>> >>>> >>>>         >>>>M >>>True, you'll still have the potential single-CPU-per-VAX-instance to worryaH >>>about.  My point is that you'll need fewer actual boxes to manage it.D >>>This can often be less expensive than lots of Wintel/amd servers. >>>  >>> 	 >>>      o >>>nO >>>>>While "c." above is still somewhat true, you can probably get sufficientlygP >>>>>fast Alpha server such that a single CPU virtual VAX may be fast enough forM >>>>>most VAX 78xx workloads.  And certainly a 32-way Marvel would be able towF >>>>>run enough images to reduce the total number of boxes you'd need. >>>>>  >>>>>r >>>>>u >>>>>          e >>>>> G >>>>Yes, but at what cost?  How many folks are running VAX because theyyI >>>>could not afford the migration then, and/or now, or how many couldn't H >>>>migrate because of their VAX apps or environment?  VUP per CPU, a PCL >>>>running C-V will produce faster emulated results than Alpha and throwingK >>>>a 32-way Marvel isn't a realistic migration solution for 99% of the VAXe? >>>>shops where as $8k in PC's, well that is a different story.r >>>> >>>> >>>>         >>>>N >>>I don't think that most of the servers I've seen implemented for productionJ >>>use, including those under consideration for C-V, are in the $8K range.N >>>And if it takes 24 of those $8K boxes to handle your workload (a real-worldF >>>example I've seen), then 2 16-CPU alphas might actually be cheaper. >>>e >>>-	 >>>      4 >>>EB >>Auh, perhaps you skimmed the early part of my post, but that $8kJ >>represents four $2k boxes and those four boxes running C-V could produceF >>the equivalent total VUPS of a 7840 where as two 16-way Alphas wouldH >>still have to be running multiple instances of C-V to produce the sameJ >>potential horse power and I would have to believe that Alpha iron is far4 >>more expensive than a total of $8k in PC hardware. >>? >>Just an side, I also mentioned the $2k price as one of my C-V E >>systems...  This is just an example of an entry-level hosting price H >>that, if compounded by four systems, can reach the realm of processingE >>you suggest for a 7840 but if you are actually running that type ofcH >>load, most likely you would spend more on systems with better hardwareJ >>redundancy and even if you reached $5k per box, four $5k boxes are still) >>cheaper than the dual Alpha approach...  >>     >> >f >h/ >Don't forget to add in the cost of C-V itself.d >  e >hI Yes, but that is a given regardless of the target system.  Additionally,  I the discussion was running multiple instances of C-V and I know that the 'I Windows version has been specifically set to check and prevent that from e= happening, so I have to believe the Alpha version does too...a  H Additionally, there is all this talk of vitualization software, and you E folks complain about just running Charon on a striped down Windows?  sI Never the less, we have a lot of software, besides C-V, that use dongles  G and most of them use built in counters and date/time stamps to prevent cH concurrent copy abuse...  Would you really want to bet you company on a 5 copy of software that you are using illegally anyway?-  H >And, if you require any special hardware, add in the cost of that.  (If >it is even available.)i >  e >   F Special hardware?  What hardware is available on a Alpha box that you I cannot get on a X86 platform?  If you referring to VAX hardware, that is lA an entirely different issue of which even Alpha cannot resolve...)     Barrye   -- e  > Barry Treahy, Jr                       E-mail: Treahy@MMaz.com> Midwest Microwave                          Phone: 480/314-1320> Vice President & CIO                         FAX: 480/661-7028                        i   ------------------------------   End of INFO-VAX 2004.392 ************************