1 INFO-VAX	Wed, 29 Nov 2006	Volume 2006 : Issue 657       Contents:0 Re: Cluster connection lost when one link fails?0 Re: Cluster connection lost when one link fails?0 Re: Cluster connection lost when one link fails?0 Re: Cluster connection lost when one link fails?0 Re: Cluster connection lost when one link fails?0 Re: Cluster connection lost when one link fails?0 Re: Cluster connection lost when one link fails? Re: DECW$SERVER crashes (8.3)  Re: interested in wombats 4 Re: libXrender 0.9.0 for OpenVMS ALPHA 8.2 or later. Re: Looking for a bunch of URLs  Re: Looking for a bunch of URLs 1 Mac patches 31 os x critical flaws - more needed!  Re: OpenVMS Clustering Question  Re: OpenVMS Support Issues Re: OpenVMS Support Issues Re: OpenVMS Support IssuesC Re: OpenVMS/Alpha V8.3 + Motif V1.6 + Password locked console crash ? Re: Reloading Alpha drivers, was: Re: Mouse: thumbwheel support 4 Re: Thoughts on the book: DEC is dead, long live DEC4 Re: Thoughts on the book: DEC is dead, long live DEC4 Re: Thoughts on the book: DEC is dead, long live DEC! Re: VAX 11/730 SABACKUP questions ! Re: VAX 11/730 SABACKUP questions   F ----------------------------------------------------------------------    Date: 29 Nov 2006 01:36:26 -0800/ From: "Volker Halle" <volker_halle@hotmail.com> 9 Subject: Re: Cluster connection lost when one link fails? A Message-ID: <1164792986.529506.40000@16g2000cwy.googlegroups.com>    Malcolm,  , your errlog entry describes a LINK UP event:  >           15--<-12  11--<-08  07--<-04  03--<-00   :Byte OrderE  0000:    08020111  000000CA  00000001  0000000B   *................* E  0010:    00000008  00000003  00001103  00000001   *................* E  0020:    00000000  00000002  00000000  00000000   *................* E  0030:                                  00000000   *            ....*   C WEBES V4.4.4 decodes this (example from a similar errlog entry) as:   / LAN_Event_Packet_Fast_Eth_82559  Packet_Size 11  Packet_Revision 1 ! pkt_entry_type x0000 00CA Link Up % pkt_vms_Version_Generation x0802 0042 " pkt_vms_Version_Variant x0000 0004 pkt_LAN_LSB_Status x0000 0103 F pkt_EI_82559_Driver_State 1 Link is down, transmits are completed with error status pkt_LAN_Link_Transitions 3 pkt_LAN_Link_Soft_Errors 0 pkt_LAN_Link_Fatal_Errors 0  pkt_LAN_Link_Device_Resets 1  pkt_LAN_Link_Transmit_Timeouts 0  G Note that the pkt_EI_82559_Driver_State = 1 is decoded incorrectly. The  correct decode would be:  $ 1 = Link is up, transmits are issued; 3 = Link is down, transmits are completed with error status   / A Link Down pkt_entry_type would be: x0000 00CB    Volker.    ------------------------------  % Date: Wed, 29 Nov 2006 10:14:51 +0000 ! From: "R.A.Omond" <Roy@Omond.net> 9 Subject: Re: Cluster connection lost when one link fails? 4 Message-ID: <ekjmit$kbm$1$8300dec7@news.demon.co.uk>   Volker Halle wrote: 
 > Malcolm, > A > two new bits (12 and 13)  in LAN_FLAGS have been supported with C > VMS732_LAN-V0300 and higher (backport from V8.2). For a detailled B > description of those 2 bits, please see the VMS732_LAN-V0300 ECO > release notes: > : > http://eisner.decus.org/DECUServe/DECnotes/ECO/1473.HTML > G > The VC can only be closed, if there are no packets pending in the XMT F > cache of the VC(VC$A_BUF_FIRST must be 0). If those packets would beE > stuck in the LAN transmit pending queue, your system is in trouble.  > E > Increasing RECNXINTERVAL in the situation we've both seen would not H > have helped. We had RECNXINTERVAL=60 and crashed 72 seconds after linkG > down, in your case (RECNXINTERVAL=20), your system crashed 30 seconds  > after link down. > H > Setting LAN_FLAGS bit 12 should have reduced the transmit timeout fromB > about 3-4 to 1-1.25 seconds, but I can't see how this would have	 > helped.   H I think the 3-4 seconds is misleading.  I was at the particular customerH site that brought this long transmit timeout to attention (they sufferedG a 10 second "blip" which, for the customer, was like a gazillion years. F They had a lot of explaining to do to their customers; this was in the New York area).   I > Setting bit 13 (and restoring the pre-V8.2 behaviour) might have made a  > difference...   F I'm not on-site at the moment, but I'd hazard a guess that the problemC only occurs if the multiple paths for SCS traffic are all using the > same device (and hence device driver).  I proferred a possibleA workaround by asking them to test the scenario where they have an A additional cluster path over 100 Mbit Ethernet.  Alas, they never D got round to it.  They were satisfied by setting bit 12 of LAN_FLAGSB and RECNXINTERVAL down to 2 or 3 seconds - the goal was to achieveB failover in under 7 seconds.  The simulation of one of the GigabitD paths going down was simply to pull one of the cables.  I must admitB I was skeptical that this would cause the CLUEXIT, but it was 100%
 reproducable.   , This is a very interesting thread for me ;-)  	 Roy Omond  Blue Bubble Ltd.   ------------------------------    Date: 29 Nov 2006 04:23:49 -0800/ From: "Volker Halle" <volker_halle@hotmail.com> 9 Subject: Re: Cluster connection lost when one link fails? C Message-ID: <1164803029.414912.193030@h54g2000cwb.googlegroups.com>    My phone keeps ringing ;-)  G I have now heard about 2 other customer sites, which may have seen this  type of problem:  F Unexplained CLUEXIT crash with multiple LAN adapters running SCA after$ the failure of a single LAN adapter.  A If anyone else has seen this and still has access to such a dump, E please use the troubleshooting information provided in this thread to G collect evidence and post it here or mail it to me (volker dot halle at  invenate dot de).    --- , Volker Halle, Invenate GmbH, OpenVMS Support  # An OpenVMS crashdump analysis a day $ makes the Windows headaches go away.   ------------------------------  % Date: Wed, 29 Nov 2006 12:42:46 +0000 ! From: "R.A.Omond" <Roy@Omond.net> 9 Subject: Re: Cluster connection lost when one link fails? 4 Message-ID: <ekjv88$3es$1$8300dec7@news.demon.co.uk>   Volker Halle wrote:    > My phone keeps ringing ;-) > I > I have now heard about 2 other customer sites, which may have seen this  > type of problem: > H > Unexplained CLUEXIT crash with multiple LAN adapters running SCA after& > the failure of a single LAN adapter. > C > If anyone else has seen this and still has access to such a dump, G > please use the troubleshooting information provided in this thread to I > collect evidence and post it here or mail it to me (volker dot halle at  > invenate dot de).    Lieber Volker,  ? is there the common footprint in all cases that they're all the A same LAN adaptors (i.e. all DE500s, or all DEGXAs as is my case)?   > Beste Gruesse, und vielen Dank fuer die Arbeit in diesem Fall.  < (und ja, es ist wirklich muehsam, wenn das Telefon permanent klingelt, gell ? ;-)  	 Roy Omond  Blue Bubble Ltd.   ------------------------------    Date: 29 Nov 2006 05:00:26 -0800/ From: "Volker Halle" <volker_halle@hotmail.com> 9 Subject: Re: Cluster connection lost when one link fails? B Message-ID: <1164805226.147599.158620@80g2000cwy.googlegroups.com>   Roy,  # in our case, the SCA LAN config is:   : EWA0: DE500 failed and causing the crash, only SCA enabled< EWB0: DE500 not connected to network cable, only SCA enabledD EWC0: DEGXA-TB working all the time, SCA enabled and other protocols   Volker.    ------------------------------    Date: 29 Nov 2006 07:02:57 -0800! From: "Ian Miller" <gxys@uk2.net> 9 Subject: Re: Cluster connection lost when one link fails? B Message-ID: <1164812577.644272.92760@j72g2000cwa.googlegroups.com>   Richard B. Gilbert wrote: C > If you have a software contract, have the "crash dump weenies" in H > Colorado analyze the dump. If you don't have a contract you might wantA > to pay them by the hour.  That sort of analysis requires a good H > knowledge of VMS internals, access to the source listings, and lots ofH > practice.  The typical system manager might see one crash dump a year,% > these guys do four or more per day.   F If you don't have a support contract then there are a couple of people5 who can do crash dump analysis if you give them money   & Volker Halle (http://www.invenate.com)' John Travell (http://www.jomatech.com/)    ------------------------------  % Date: Wed, 29 Nov 2006 08:07:23 -0800 , From: "Malcolm Dunnett" <dunnett@mala.bc.ca>9 Subject: Re: Cluster connection lost when one link fails?  Message-ID: <456dafad$1@flight>   ? "Richard B. Gilbert" <rgilbert88@comcast.net> wrote in message  4 news:XJedneu-5KgmC_DYnZ2dnUVZ_oydnZ2d@comcast.com...  K >Analysis of the crash dump revealed that it was a resource problem rather  I >than a network problem.  Alas, I no longer remember what resource it was   @   I'm pretty sure this is a network problem ( see other notes in? this thread ). I'm able to precisely match the crashes to times * when one of the network switches rebooted.   > M > If you have a software contract, have the "crash dump weenies" in Colorado   > analyze the dump.   D     I have a contract and I may log a call. At the moment it appearsD that Volker is seeing the identical problem and has done significant@ crash dump analysis already, so I'm hoping he'll log the call asC he could do a better job of explaining to HP what they need to look  for.  B     I have saved a crash dump showing the problem which I can send7 to HP if they want it (or Volker if it would help him).   M > If you don't have a contract you might want to pay them by the hour.  That  = > sort of analysis requires a good knowledge of VMS internals   A    used to know them but that was back in the V4 days, I'm really  rusty and lots has changed!     >, access to the source listings  E     no source listings here since they stopped including the listings H microfiche in the distros - reading the fiche was a pain but better than nothing.   ------------------------------  % Date: Wed, 29 Nov 2006 09:23:00 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> & Subject: Re: DECW$SERVER crashes (8.3)6 Message-ID: <606a$456d97af$cef8887a$5419@TEKSAVVY.COM>  1 It crashed again. (DECW$SERVER runs on node BIKE)   D This time: WHEEL has SET DISP/CREATE/EXEC/TRANSPORT=DECNET/NODE=BIKE    From BIKE, I did:	 MC SYSMAN  SYSMAN> SET ENV/NODE=WHEEL( SYSMAN> DO CREATE/TERM/DETACHED/NOLOGGED  J (The first time, it seemed to just take forever, so I <CTRL-C> out of it, H and re-issued the command. This time, I heard the disk on WHEEL do some < work (indicating it was processing he command, and then ....  G Kaboom ! DECW$SERVER crashed and all my windows gone. (I was trying to  I document a Web server problem, and thankfully ALLIN1 has an auto-recover  	 feature).   , This is what the new DECW$SERVER looks like:  ( $ show proc/id=28E01157/quota/accounting  G 29-NOV-2006 09:08:28.50   User: JFMEZEI          Process ID:   28E01157 O                            Node: BIKE             Process name: "DECW$SERVER_0"    Process Quotas:    Account name: VAXIN G   CPU limit:                      Infinite  Direct I/O limit:       100 G   Buffered I/O byte count quota:    334704  Buffered I/O limit:     256 G   Timer queue entry quota:              14  Open file quota:        280 G   Paging file quota:                254976  Subprocess quota:        32 G   Default page fault cluster:           64  AST quota:              249 G   Enqueue quota:                       497  Shared file limit:        0 G   Max detached processes:                0  Max active jobs:          0    Accounting information: B   Buffered I/O count:     11001  Peak working set size:      45520B   Direct I/O count:        2042  Peak virtual size:         398496B   Page faults:             6085  Mounted volumes:                0   Images activated:           0 *   Elapsed CPU time:          0 00:00:05.62*   Connect time:              0 00:17:21.36   Soft CPU Affinity: off  & IS THERE ANYTHING BAD IN THE ABOVE ??? ----------------------    / This is what the DECW$SERVER_0_ERROR.LOG shows: ? 29-NOV-2006 08:49:08.3 Connection 6092410 is accepted by Txport = 29-NOV-2006 08:49:08.3 Access granted to: DECNET WHEEL SYSTEM       matched entry: DECNET * *J 29-NOV-2006 08:49:14.7 %DECW-I-ACCEPT_FAILED, could not accept connection 9 request from WHEEL::"0=SYSTEM/4 ....................X$X0" @ 29-NOV-2006 08:49:14.8 -SYSTEM-F-EXQUOTA, process quota exceededM 29-NOV-2006 08:49:47.7 Connection 6092410 is closed by Txport (status = 20e4) ? 29-NOV-2006 08:50:15.6 Connection 6092410 is accepted by Txport = 29-NOV-2006 08:50:15.6 Access granted to: DECNET WHEEL SYSTEM       matched entry: DECNET * *    I So the 08:49:47 would be when I typed in the CTRL-C. And 08:50:15 when I  H issued the second CREATE/TERM command. And this time, no indication the  server crashed.   ' The last error message before that was:   K 26-NOV-2006 03:21:47.5 AllocatePixmap failed, called from fbCreatePixmapBpp     H (and on the 25th there was a series of similar messages probably when I L tried to load some web pages that were rich in images). But on the 26th, it 5 was just the one line. (probably when trying to view  7 http://www.srh.noaa.gov/ridge/Conus/RadarImg/latest.gif   J Interestingly, right now, Mozilla does display the image. But if I try to J display the same web page (just a large GIF image) later on, after having L tried to load large pages, even though Mozilla is reduced to just one empty 4 window, it won't be able to load that image anymore.  N So it does look like Mozilla forgets to destroy pixmaps in some circumstances.   ------------------------------  % Date: Wed, 29 Nov 2006 10:06:28 -0500 + From: ben yates <winston19842005@yahoo.com> " Subject: Re: interested in wombats5 Message-ID: <pan.2006.11.29.15.06.28.67807@yahoo.com>   ; On Mon, 27 Nov 2006 20:52:14 -0600, David J Dachtera wrote:    > wombat@fancier.net wrote:  >>  > >> Can someone tell me whether Wombats live only in Australia, >> or also on other continents?  >>   >> Apart from zoos, of course. > < > In older versions of Datatrieve, see HELP ADVANCED WOMBAT.  4 Did they remove that from newer (?) versions of DTR?  C I remember that vividly, and wonder what the story behind that was?    ------------------------------  % Date: Wed, 29 Nov 2006 09:23:01 +0100 ( From: JOUKJ <joukj@hrem.nano.tudelft.nl>= Subject: Re: libXrender 0.9.0 for OpenVMS ALPHA 8.2 or later. ; Message-ID: <58ea6$456d4366$82a13c9d$1701@news1.tudelft.nl>    John E. Malmberg wrote:  > JOUKJ wrote: >  >> John E. Malmberg wrote: >>I >>> libXrender is the client side library for the Xrender extension.  It  J >>> is required for building the current Cairo graphics package, which is 5 >>> used by GTK+ 2.x for displaying on X-11 displays.  >>> G >> Do you have a working version of cairo and gtk2? I'm a little stuck  C >> with cairo giving me wrong results except for the ones that use  : >> Xrender (but I cannot do that on a "native" VMS-screen. >  > K > I have cairo working what appears to be good enough for my purposes.  It  H > will just take me some time to get all the components packaged up for  > distribution.  > I > I am seeing some tests fail because a few pixels are out of place.  It  J > looks like rounding errors.  A search on google shows that LINUX on x86 I > runs IEEE floating point at a different precision then the rest of the  3 > world.  I do not know if that is the real reason. H I had some discussion with one of the authors of cairo and he said that I "rounding errors" were likely to occur (due to different implementations  F of IEEE precission and different algorithms for "standard" functions)  but were harmless. > F > And at least one test does not agree with the library source and is  > failing because of that.G I normally test the fill-rule test and do not see the "stars" but some  ( triangles. And that is completely wrong. > F > GTK+ 2.x requires true-color X11 displays on any platform.  I could 9 > display them fine on a DS10 and a Radeon graphics card. G I think it needs a 24-bit display. But that is a limitation of gtk2 on  E all platforms. I tried to do it on a XP1000 with a Elsa-gloria card.  A (but it is already some time ago and in the mean time I upgraded  I DecWindows to verson 1.6 and installed some "experimental" drivers which  E I got from engeneering to make the DirectColor visual work properly.)  So I have to test agian. > 1 > My last tests with cairo were with Cairo 1.0.2. I I tested 1.0.6 and many versions I extracted from CVS over the last year.                         Jouk    ------------------------------    Date: 29 Nov 2006 01:45:18 -0800! From: "Ian Miller" <gxys@uk2.net> ( Subject: Re: Looking for a bunch of URLsC Message-ID: <1164793517.891739.220640@j72g2000cwa.googlegroups.com>    other collections of links  5 http://eisner.encompasserve.org/~miller/hplinks.htmlx 6 http://eisner.encompasserve.org/~miller/vmslinks.htmlx0 http://www.openvms.org/pages.php?page=OpenSource. http://www.openvms.org/pages.php?page=Software0 http://www.openvms.org/pages.php?page=FAQs_Howto/ http://www.openvms.org/pages.php?page=Developer    ------------------------------  % Date: Wed, 29 Nov 2006 15:28:37 +0200 4 From: Mike Rechtman <michael.rechtman.nospam@hp.com>( Subject: Re: Looking for a bunch of URLs& Message-ID: <456DA725.46BECBCF@hp.com>   Beach Runner wrote:  >  > Moved and lost bookmarks.  >  > Tools, > The site with t4,  > availability manager
 > freewareF > cross reference on various hardware platforms and supported firmware >  > ThanksA Got a number of OpenVMS-related links on http://alpha.mike-r.com/    Mike --  E --------------------------------------------------------------------- E Usual disclaimer: All opinions are mine alone, perhaps not even that. ? Mike Rechtman                            *rechtman@tzora.co.il* F Kibbutz Tzor'a.                          Voice (home): 972-2-9908337  B   "20% of a job takes 80% of the time, the rest takes another 80%"E ---------------------------------------------------------------------  -----BEGIN GEEK CODE BLOCK-----  Version: 3.1; GCM/CS d(-)pu s:+>:- a+++ C++ U-- L-- W++ N++ K? w--- V+++$ 6 PS+ PE-- t 5? X- tv-- b+ DI+ D-- G e++ h--- r+++ y+++@ ------END GEEK CODE BLOCK------    ------------------------------    Date: 29 Nov 2006 05:47:37 -0800 From: bob@instantwhip.com : Subject: Mac patches 31 os x critical flaws - more needed!B Message-ID: <1164808057.512117.157630@14g2000cws.googlegroups.com>  $ when did vms ever need that many ...  5 http://www.theinquirer.net/default.aspx?article=36033    ------------------------------  % Date: Wed, 29 Nov 2006 09:37:23 +0100 / From: Paul Sture <paul.sture.nospam@hispeed.ch> ( Subject: Re: OpenVMS Clustering QuestionJ Message-ID: <paul.sture.nospam-61B997.09372329112006@mac.sture.homeip.net>  ) In article <ekf31b$18h4$1@pyrite.mv.net>, :  Stephen Hoffman <Hoff@HoffmanLabs-RemoveThis-.Org> wrote:  G >    Make no mistake -- this is a big project -- the aggregate process  ? > state is scattered all over system and process address space.  >   1 If I may add to that, it is indeed a big project.   D >    There's a version of this checkpoint-restart mechanism already H > available, I might add.  It's called a transactional database.  Works J > nicely.  If you're not working on a database, you can implement most of > > what you need here using locks and existing RMS flush calls.  G It was available on VMS in the early to mid eighties. Unfortunately it  D wasn't marketed properly. RMS based, with pre and post journalling, ? before Digital implemented that. The key library routines were  H Transaction Start and the Commit or Rollback, and naturally there was a + fair bit of record locking stuff in there.    I For example when adding a record to an indexed file within a transaction  > which could get rolled back, on subsequent operations in that H transaction which deal with that record, do you return record locked or  record doesn't exist yet?   D I can tell you that I had quite a fight with the ex-IBM crowd about H record locking. They seemed to think that you should lock a record when B printing a price list, but when you know that the prices are only I updated once a year, and you do it with all users logged off, use a read   regardless...   > Incidentally, that brings us back to the argument about using D BACKUP/IGNORE=INTERLOCK. If you know your applications well and can : quiesce them at the right time, it really isn't a problem.   --  
 Paul Sture   ------------------------------  % Date: Wed, 29 Nov 2006 02:16:40 -0500 ' From: Dave Froble <davef@tsoft-inc.com> # Subject: Re: OpenVMS Support Issues 9 Message-ID: <y7ednVbuP4x6rvDYnZ2dnUVZ_uGdnZ2d@libcom.com>    Michael D. Ober wrote:; > "Beach Runner" <Bob4Health@hotmail.com> wrote in message  ? > news:1164750217.021917.208620@l12g2000cwl.googlegroups.com... G >> A company I know complained about their support.  HP said, so get it 
 >> elsewhere. J >> They had large purchases for many products on order. They put a hold on >> all of them. / >> Lots of people with suits from HP showed up.  >>8 >> THAT made a real change. Not theory. Not maybe. FACT. >> > I > Unfortunately this is the only way to get most corporations to listen.  ? > Those of us in small companies don't have that type of clout.  >  > Mike Ober. >  >   I The total from all small companies probably is quite large.  The problem  @ is, HP won't gag on a few of them canceling, or threatening to, H contracts.  But when enough of them do, it surely (shirley) will appear 0 on the bottom line, and will then get attention.  2 But there is a problem with that, at least for HP.  B Several years ago when the stock market took a dive, my insurance G company, Nationwide, significantly increased all premiums.  I bitched.  G "So what, I was told".  I dropped them and found other insurers.  Now,  G I'm getting inundated by mailers and other offers, pleading with me to  0 come back, and they'll save me a bunch of money.  F The bastards must have figured that most customers would just swallow I the increase.  Bitching doesn't hurt their bank account.  Must be that a  F bunch of people dumped them.  That does hurt their bank account.  But 8 they're not doing so well at getting the customers back.  / Only difference is, I'm not sure that HP cares.    --  4 David Froble                       Tel: 724-529-0450> Dave Froble Enterprises, Inc.      E-Mail: davef@tsoft-inc.com DFE Ultralights, Inc.  170 Grimplin Road  Vanderbilt, PA  15486    ------------------------------    Date: 29 Nov 2006 08:16:34 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) # Subject: Re: OpenVMS Support Issues 3 Message-ID: <IKss1X1tsR2j@eisner.encompasserve.org>   h In article <490b3$456baf37$cef8887a$17585@TEKSAVVY.COM>, JF Mezei <jfmezei.spamnot@teksavvy.com> writes: > L > No, the TOS people lived or died depending on the colour of their shirts. / > The red ones died. The other ones didn't die.   B    Although TOS generally picked on red shirts for death, many redD    shirts lived.  And some times others died.  If your going to makeA    an interesting show you have to give some of the guest stars a     death scene.   F    After years of watching TOS reruns on a B&W TV, I was caught by theE    bold red color of the bridge rails the first time I tuned it in on     a color TV.  I    I just saw the opening of the first pilot and realised that the bridge B    rails, etc., were not the same color as they were in productionH    episodes.  I wonder if they were changed between pilots, or later on?  I    I'll have to dig the second pilot and some early production shows out  &    of my tape library and have a look.   ------------------------------    Date: 29 Nov 2006 07:50:28 -0800$ From: "AEF" <spamsink2001@yahoo.com># Subject: Re: OpenVMS Support Issues C Message-ID: <1164814209.890261.240830@n67g2000cwd.googlegroups.com>    Beach Runner wrote:  > Bill Gunshannon wrote:G > > In article <1164520070.551042.321500@f16g2000cwb.googlegroups.com>, + > > 	"AEF" <spamsink2001@yahoo.com> writes:  [...]  > > > L > > > Cancelling your support contract (or not renewing it, or whatever) mayJ > > > also help. And who knows? It may not. They might not care even about< > > > that! But when you've got nothing to lose...go for it! > > I > > The only option that will have any effect.  It will give them another K > > yawn, but it wil give you the money to spend elsewhere on real support.  > >  > > bill > >  > > --F > A company I know complained about their support.  HP said, so get it > elsewhere.I > They had large purchases for many products on order. They put a hold on  > all of them.. > Lots of people with suits from HP showed up. > 7 > THAT made a real change. Not theory. Not maybe. FACT.   G So what did the suits do? Beg for them not to cancel their many orders? A Re-hire fired knowledgeable support people just for this company? F Apologize in person? Offer a month's worth of toner for free? What did they do?   AEF    ------------------------------    Date: 29 Nov 2006 03:38:00 -0800/ From: "Volker Halle" <volker_halle@hotmail.com> L Subject: Re: OpenVMS/Alpha V8.3 + Motif V1.6 + Password locked console crashB Message-ID: <1164800280.920949.35350@n67g2000cwd.googlegroups.com>  	 Eberhard,   B could you enable image accounting (SET ACC/ENA=IMAGE) and look for8 image termination records with an unusual status value ?   Volker.    ------------------------------  % Date: Wed, 29 Nov 2006 22:51:39 +0800  From: prep@prep.synonet.com H Subject: Re: Reloading Alpha drivers, was: Re: Mouse: thumbwheel support0 Message-ID: <8764cys3us.fsf@k9.prep.synonet.com>  1 Paul Sture <paul.sture.nospam@hispeed.ch> writes:   E > That is something which has always been missing on VMS. RT-11 had a B > means to unload a device driver and it was a very handy feature.  F So did RSX, but neither unloaded the driver/device data structures. IfF you wanted to reload a driver with identical data structures, fine and@ it worked. But no changing the data, esp as who knows could have@ salted away addresses to all sorts of bits, and the system woudl really like them to stay there.    --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------   Date: 29 Nov 2006 07:09:33 GMT2 From: "Dave Weatherall" <djw-nothere@nospam.nohow>= Subject: Re: Thoughts on the book: DEC is dead, long live DEC ? Message-ID: <DTiotGxQ0bj6-pn2-3uGf8dYIy3Tz@dave2_os2.home.ours>   D On Tue, 28 Nov 2006 18:31:40 UTC, ChrisQuayle <nospam@devnul.co.uk>  wrote:   > Stephen Hoffman wrote:   <Snip>  C > 1987.  Major Internet backbones were T1 lines.  Though yes, once  M > > something reaches critical mass, it's catch-up time -- which is itself a  J > > real problem -- if you're not already apace in the game.  OpenVMS was N > > obviously late to the game, first having TCP/IP (UCX) in the fall of 1988. > >   K > That may be true, but the only place I ever used Decnet seriously was at  K > Dec, where the corporate network was said to be the largest in the world  J > at the time. The functionality was amazing - copy a file on the command J > line from anywhere in the world. Vax notes etc, so, so good. Everywhere K > else I worked ran tcp/ip networking. Experimented with decnet at home on  I > an ancient 730, vms 4.3, flaky R80 etc, but never got comfortable with  H > all the concepts (terminology ?) and setup in the same way that I did K > later with ip networking. It just seemed overcomplex and the real killer  H > was that it didn't talk to anything else in the lab, other than other H > Vaxen, or via kermit and serial lines. By that time, I had a scrapped  <Snip>  6 ISTR hat Linux does DecNet - haven't tried it yet tho'   --   Cheers - Dave W.   ------------------------------    Date: 29 Nov 2006 08:08:56 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) = Subject: Re: Thoughts on the book: DEC is dead, long live DEC 3 Message-ID: <NmBLeskUBcdX@eisner.encompasserve.org>   t In article <DTiotGxQ0bj6-pn2-3uGf8dYIy3Tz@dave2_os2.home.ours>, "Dave Weatherall" <djw-nothere@nospam.nohow> writes: > 8 > ISTR hat Linux does DecNet - haven't tried it yet tho'  D    Linux is one of many non-DEC OS that DECnet is available for.  MyF    son runs it.  There was a bug in the first version he loaded but it    was soon fixed.   ------------------------------  % Date: Wed, 29 Nov 2006 16:37:00 +0000 - From: Roy Omond <Roy.Omond@BlueBubble.UK.Com> = Subject: Re: Thoughts on the book: DEC is dead, long live DEC 4 Message-ID: <ekkcvc$knm$1$8300dec7@news.demon.co.uk>   prep@prep.synonet.com wrote:+ > ChrisQuayle <nospam@devnul.co.uk> writes:  >  > C >>That may be true, but the only place I ever used Decnet seriously D >>was at Dec, where the corporate network was said to be the largestB >>in the world at the time. The functionality was amazing - copy aE >>file on the command line from anywhere in the world. Vax notes etc,  >>so, so good. >  > B > There was also HEPnet and the NASA internal net. If you wanted aF > serious, reliable network, and could live with the size limit, there > was only one that worked.   < Ah yes, HEPnet/SPAN.  Are you referring to the size limit in, DECnet of 63 areas and 1024 nodes per area ?  @ HEPnet/SPAN got round that by setting the backbone nodes to haveA a maximum area number of 36 (pretty sure I got the number right). ? Areas above that were reserved for "local" usage, with the only B downside being that you had to use "poor man's routing" from nodes in the high area numbers.   B If I remember correctly, HEPnet/SPAN was the second biggest world-$ wide DECnet after DEC's own network.   Those were the days ;-)   	 Roy Omond  Blue Bubble Ltd.   ------------------------------  % Date: Wed, 29 Nov 2006 02:29:18 -0500 ' From: Dave Froble <davef@tsoft-inc.com> * Subject: Re: VAX 11/730 SABACKUP questions9 Message-ID: <TvGdnRr-PedAq_DYnZ2dnUVZ_qCdnZ2d@libcom.com>    Chris Scheers wrote:A > I need some help from the way back machine.  (Old farts unite!)  > J > I have a client with an IDC 11/730 (R80/RL02/TU80).  They don't want to 
 > replace it.  > I > The system disk (R80) died and has been replaced.  Now they can't find   > their SABACKUP media.  > H > What are my options here?  I know how to boot SABACKUP from a TU58 or M > RL02, but I don't have a machine that I can use to generate media for them.  > K > Is there a way to boot a 730 from the TU80?  If so, could some kind soul  0 > provide me with the relevant console commands? > H > With a DEUNA, is there a way to boot a 730 as a satellite?  How about  > from an Infoserver?  > D > If, as I suspect, I am out of luck on these alternate methods, is I > someone set up to generate SABACKUP media on TU58 or RL02?  The target  C > OS is VMS 5.4, but I think any 5.2 or later SABACKUP should work.  >  > Thanx! >    Chris,  C VMS at one time was distributed on 9-track tape.  The only thing I  < cannot remember is whether SA BACKUP was on a separate tape.  G Regardless, if you have access to another 9-track drive, you should be  E able to copy or build SA BACKUP on a tape.  (It has been a very long   time, so I could be wrong.)   H To boot the tape, for example, MTA0:, I believe you just do that at the  console prompt, BOOT MTA0    Remember STABACKIT   --  4 David Froble                       Tel: 724-529-0450> Dave Froble Enterprises, Inc.      E-Mail: davef@tsoft-inc.com DFE Ultralights, Inc.  170 Grimplin Road  Vanderbilt, PA  15486    ------------------------------  # Date: Wed, 29 Nov 2006 14:10:48 GMT ' From: ChrisQuayle <nospam@devnul.co.uk> * Subject: Re: VAX 11/730 SABACKUP questions2 Message-ID: <Ixgbh.70$GF5.41@newsfe2-gui.ntli.net>   Dave Froble wrote:   >  > Chris, > E > VMS at one time was distributed on 9-track tape.  The only thing I  > > cannot remember is whether SA BACKUP was on a separate tape. > I > Regardless, if you have access to another 9-track drive, you should be  G > able to copy or build SA BACKUP on a tape.  (It has been a very long   > time, so I could be wrong.)  > J > To boot the tape, for example, MTA0:, I believe you just do that at the  > console prompt, BOOT MTA0  >  > Remember STABACKIT >   I The 730 used an msdos like batch / command file on a tu58 normally, with  E the various cpu registers loaded to reflect boot mode and controller  H that you wished to boot from. At the terminal, you can boot from either B tu58 drive 0, or 1,  but not from a disk or other device. The cpu B microcode must be loaded from tu58 before the main os boot can be E started. I have a 725 (730, deskside) which could have built console  C tapes, but the console is dead and haven't had time yet to fix it.  3 Anyone have schematics for the console subsystem ?.   G I still have a collection of tu58 tapes for 730 in store, which the op  G is quite welcome to borrow, but will need to check if a stabackup tape  ; is among them. I think the os was 4.3 or 4.7, but not sure.   C Contact off group at: syseng at gfsys dot co dot uk, if it helps...    Regards,   Chris    ------------------------------   End of INFO-VAX 2006.657 ************************