INFO-VAX Sun, 28 Jan 2007 Volume 2007 : Issue 55 Contents: Re: DECUS Seminars in Toronto and =?ISO-8859-1?Q?Montr=E9al?= Re: How long to really setup a VMS system ? Re: Nonstop UNIX takes another loss Re: OpenSSL & OSU Re: PL/I for Itanium Re: PPPD, TCPIP Services and incomplete negotiations Re: PPPD, TCPIP Services and incomplete negotiations Re: SHOW TERMINAL glitch (PPP) Re: SHOW TERMINAL glitch (PPP) VAMP is back! Re: VAMP is back! Re: VAMP is back! VMS installation trick Re: VMS installation trick ---------------------------------------------------------------------- Date: Sat, 27 Jan 2007 19:03:29 -0800 From: Ken Fairfield Subject: Re: DECUS Seminars in Toronto and =?ISO-8859-1?Q?Montr=E9al?= Message-ID: <522i49F1mlg5dU1@mid.individual.net> Sue wrote: > Well how about a hug, I am not real big on beating anyone. Though > for comedy we could do a duct tape routine. I can not be the only Red > Green fan in the world. You're not! The local PBS station (Portland, OR) carries it and we're quite the fans. :-) -Ken -- Ken & Ann Fairfield What: Ken dot And dot Ann Where: Gmail dot Com ------------------------------ Date: Sat, 27 Jan 2007 23:04:12 -0500 From: =?ISO-8859-1?Q?Arne_Vajh=F8j?= Subject: Re: How long to really setup a VMS system ? Message-ID: <45bc20bb$0$49205$14726298@news.sunsite.dk> Dave Weatherall wrote: > This is one of the attitudes that has led to te demise of VMS and the > rise of the PC. It encourages the view that VMS , despite name-change, > is 'closed' and the Win/Mac PC is open. 'They won't let me do it on > there, I'll do it on the PC' , whether it's a mortgage repayment > program or a behavioural model is what I've seen happen. > > If we believe that VMS is secure system, that the privelege structure > works so that 'you can only do the harm you've been granted > privelegefor', then we should be/have been encouraging people to use > the system. Are their exceptions? Sure. The discussion was about production systems. You don't just do things on those. And it does not matter if it VMS, Solaris, Linux or Windows. If it is for mixed usage, then it is another discussion. Arne ------------------------------ Date: Sat, 27 Jan 2007 18:16:56 -0500 From: JF Mezei Subject: Re: Nonstop UNIX takes another loss Message-ID: <45bbddc9$0$9160$c3e8da3@news.astraweb.com> jreederhi@nac.net wrote: > Unix" at the Toronto Stock Exchange. The Exchange is replacing its > NonStop trading system with a Linux system call Leapfrog. Similar to > replacing a tractor with 10,000 chickens. www.availabilitydigest.com As a sideline, the Toronto Stock exchange has done a lot bad things with regars to IT in its history. Since the 1980s, the TSE was sure the Montreal exchange would die and made sure it had the capacity to handle it. Quebec tax incentives delayed the death of the montreal exchange by 10 years. By then the TSE had not kept up with the times, probably thinking its upgrades in the 1980s would have been enough and had to shut down for a number of afternoons because its systems couldn't cope with the volume. (The montreal exchange had already migrated from VMS to Solaris before it was shutdown). After they upgraded to Tandem, the TSE stopped making the news about its IT system failures. I guess they have forgotten about their 1990s follies already and are taking the big risk by going to an intested platform for such a high visibility application where downtime makes the top of the national TV news. ------------------------------ Date: Sun, 28 Jan 2007 00:27:14 GMT From: "Malcolm Dunnett" Subject: Re: OpenSSL & OSU Message-ID: wrote in message news:1169650222.301368.141740@v33g2000cwv.googlegroups.com... | | Although it isn't a direct answer to you question, I've also had | problems getting OSU and SSL working. Part of the problem is that in | order for it to work with current web clients, it requires using the | SSL_TASK method. I've built the OSU 3.10 server on Itanium against the HP supplied SSL library (based on SSL 97e) using the SSL MST and it appears to work fine with "current" web browsers ( Firefox and IE 7 ). Maybe I just haven't found the problems yet, what sort of issues should I be looking for. I use the server on Alpha extensively with the SSL MST and haven't seen any problems with current web clients but that's with an older version of the OpenSSL library. ------------------------------ Date: Sat, 27 Jan 2007 23:13:46 -0500 From: =?ISO-8859-1?Q?Arne_Vajh=F8j?= Subject: Re: PL/I for Itanium Message-ID: <45bc22fb$0$49198$14726298@news.sunsite.dk> Michael Kraemer wrote: > Richard B. Gilbert schrieb: >> That doesn't mean that there aren't any but it suggests that PL/I is >> not widely used. > > That's partially true, but can be said from > any language but C/C++. Java and C# have both passed C++/C for new development. There are also still significant VB6/VB.NET even though C/C++ is probably bigger. > Too bad that PL/I was an overly complex language > and for that reason not easy to implement on > anything but mainframes and VAXen. C was simply > "faster on the floor" for most other platforms > in use nowadays. It is not obvious to me that it is technical reasons that makes PL/I a mainframe and VMS/VAX thing. I suspect it is more a history/people/educational thing. Another complex language is Ada and it is AFAIK widely used on Unix and embedded platforms. Arne ------------------------------ Date: Sat, 27 Jan 2007 22:03:38 GMT From: Forrest Kenney Subject: Re: PPPD, TCPIP Services and incomplete negotiations Message-ID: <45BBC3BB.AD7A612E@hp.com> 1) You should not need to start anything. The trace code is stated one we do the line conversion. 2) Right off the top of my head I do not know why you would have gotten a privilege error on the command. Are you sure you have privileges to create a detached process, and the privileges to change attributes on another serial port. 3) There is not sufficient data to know exactly what you are trying to do here. What terminal are you logged in on. What terminal port are you tring to switch. The complete dialog of the PPPD command you issued would really help. After the PPPD connect do you see lots of what looks like line noise on your terminal. Forrest JF Mezei wrote: > > Forrest Kenney wrote: > > There is an undocumented tracing capability built into PPPD. To > > enable it use the command set ttxx/trace/full. The /full is optional > > it just tracks more packet types. > > Sorry for delay in followups, but have been quite busy checking everything, > including resoldering serial connectors to ensure full modem control. (lack > of carrier detect caused the VMS terminla driver to complain). > > Now however, I cannot get any PPP connection going. > > Here is a trace at the serial level (between a MAC and the Alpha. This > starts just after the PPPD CONNECT command was manually issued. > > The short story, the MAC sends configuration requests, but no data comes > back from the alpha after it has issued the "CONNECTTERM" message. > > R denotes characters sent from the Mac. > L denotes characters sent by the Alpha. > followed by minutes,seconds,hundreds of seconds > followed by the dash, followed by number of characters in that buffer. > > R:56:48:92- 1 ° > 0 > D > > L:56:48:99- 96 °°°%PPPD-I-CONNECTTERM, converting connection on device _VTA > 000255542424444445545422666767766626666667666266266766625554 > DAD50004D9D3FEE534452DC03FE65249E703FEE5349FE0FE04569350F641 > > 139: to a Point-to-Point connection° > 333327626256667276256667266666676660 > 139A04F0100F9E4D4FD0F9E403FEE5349FED > > R:56:51:35- 45 ~°}#À!}!%} }4}"}&} } } } }%}&ÇÀüÎ}'}"}(}"¥}<~ > 7F72C272272737272727272727272CCFC72727272A737 > EFD301D15D0D4D2D6D0D0D0D0D5D670CED7D2D8D25DCE > > R:56:53:26- 45 ~°}#À!}!%} }4}"}&} } } } }%}&ÇÀüÎ}'}"}(}"¥}<~ > 7F72C272272737272727272727272CCFC72727272A737 > EFD301D15D0D4D2D6D0D0D0D0D5D670CED7D2D8D25DCE > > R:56:55:18- 45 ~°}#À!}!%} }4}"}&} } } } }%}&ÇÀüÎ}'}"}(}"¥}<~ > 7F72C272272737272727272727272CCFC72727272A737 > EFD301D15D0D4D2D6D0D0D0D0D5D670CED7D2D8D25DCE > > R:56:57:09- 45 ~°}#À!}!%} }4}"}&} } } } }%}&ÇÀüÎ}'}"}(}"¥}<~ > 7F72C272272737272727272727272CCFC72727272A737 > EFD301D15D0D4D2D6D0D0D0D0D5D670CED7D2D8D25DCE > > R:56:59:00- 45 ~°}#À!}!%} }4}"}&} } } } }%}&ÇÀüÎ}'}"}(}"¥}<~ > 7F72C272272737272727272727272CCFC72727272A737 > EFD301D15D0D4D2D6D0D0D0D0D5D670CED7D2D8D25DCE > > R:57:00:92- 45 ~°}#À!}!%} }4}"}&} } } } }%}&ÇÀüÎ}'}"}(}"¥}<~ > 7F72C272272737272727272727272CCFC72727272A737 > EFD301D15D0D4D2D6D0D0D0D0D5D670CED7D2D8D25DCE > > R:57:02:83- 45 ~°}#À!}!%} }4}"}&} } } } }%}&ÇÀüÎ}'}"}(}"¥}<~ > 7F72C272272737272727272727272CCFC72727272A737 > EFD301D15D0D4D2D6D0D0D0D0D5D670CED7D2D8D25DCE > > R:57:04:74- 45 ~°}#À!}!%} }4}"}&} } } } }%}&ÇÀüÎ}'}"}(}"¥}<~ > 7F72C272272737272727272727272CCFC72727272A737 > EFD301D15D0D4D2D6D0D0D0D0D5D670CED7D2D8D25DCE > > RE: PPPD SHOW TTA0: > > This correctly picks up the speed from the terminal device. But it defaulst > to XON/XOFF flow control. But even if I get the /FLOW=HARD in the PPPD, it > doesn't seem to make a difference. Note that I can succesfully login > manually and issue the PPPD command interactively and there are no issues > with flow control at that time. > > RE: PPPD SET TRACE /FULL > > If I do SET TRACE=trace.log /FULL, I get a nasty insufficient privilege > error after it has converted the line. No trace log is created. > > If I don't specify a file, it seems to go to MBA163: But which process > consumes the mailbox data ? > > Must I manually start the logger utility from another terminal to capture > the data ? > > Also, I realise that PPPD doesn't support authentication. Will it fail if a > client tries to authenticate ? Or will it just respond with a success > status no matter what credentials are presented ? ------------------------------ Date: Sat, 27 Jan 2007 22:05:23 GMT From: Forrest Kenney Subject: Re: PPPD, TCPIP Services and incomplete negotiations Message-ID: <45BBC424.835240D@hp.com> I don't see any IPCP messages it looks like the PPP line data got negotiated but nothing ever asked for IPCP negotiation to configure the TCP/IP parts. Forrest JF Mezei wrote: > > I managed to get a log from the MAC (sorry it is in french, but you should > still understand it) > > (envoi = sending ) > > 27/01/2007 04:31 Connexion PPP établie. > 27/01/2007 04:31 LCP : envoi d¹une demande de configuration. > 27/01/2007 04:31 LCP: >> Async Map = 0x00000000. > 27/01/2007 04:31 LCP: >> Magic Number = 0x65D43949. > 27/01/2007 04:31 LCP: >> Compression de protocole > 27/01/2007 04:31 LCP: >> Contrôle/Compression d¹adresse > 27/01/2007 04:31 LCP : envoi d¹une demande de configuration. > 27/01/2007 04:31 LCP: >> Async Map = 0x00000000. > 27/01/2007 04:31 LCP: >> Magic Number = 0x65D43949. > 27/01/2007 04:31 LCP: >> Compression de protocole > 27/01/2007 04:31 LCP: >> Contrôle/Compression d¹adresse > 27/01/2007 04:31 LCP : envoi d¹une demande de configuration. > 27/01/2007 04:31 LCP: >> Async Map = 0x00000000. > 27/01/2007 04:31 LCP: >> Magic Number = 0x65D43949. > 27/01/2007 04:31 LCP: >> Compression de protocole > 27/01/2007 04:31 LCP: >> Contrôle/Compression d¹adresse > 27/01/2007 04:31 LCP : envoi d¹une demande de configuration. > 27/01/2007 04:31 LCP: >> Async Map = 0x00000000. > 27/01/2007 04:31 LCP: >> Magic Number = 0x65D43949. > 27/01/2007 04:31 LCP: >> Compression de protocole > 27/01/2007 04:31 LCP: >> Contrôle/Compression d¹adresse > 27/01/2007 04:31 LCP : envoi d¹une demande de configuration. > 27/01/2007 04:31 LCP: >> Async Map = 0x00000000. > 27/01/2007 04:31 LCP: >> Magic Number = 0x65D43949. > 27/01/2007 04:31 LCP: >> Compression de protocole > 27/01/2007 04:31 LCP: >> Contrôle/Compression d¹adresse > 27/01/2007 04:31 LCP : envoi d¹une demande de configuration. > 27/01/2007 04:31 LCP: >> Async Map = 0x00000000. > 27/01/2007 04:31 LCP: >> Magic Number = 0x65D43949. > 27/01/2007 04:31 LCP: >> Compression de protocole > 27/01/2007 04:31 LCP: >> Contrôle/Compression d¹adresse > 27/01/2007 04:31 LCP : envoi d¹une demande de configuration. > 27/01/2007 04:31 LCP: >> Async Map = 0x00000000. > 27/01/2007 04:31 LCP: >> Magic Number = 0x65D43949. > 27/01/2007 04:31 LCP: >> Compression de protocole > 27/01/2007 04:31 LCP: >> Contrôle/Compression d¹adresse > 27/01/2007 04:31 LCP : envoi d¹une demande de configuration. > 27/01/2007 04:31 LCP: >> Async Map = 0x00000000. > 27/01/2007 04:31 LCP: >> Magic Number = 0x65D43949. > 27/01/2007 04:31 LCP: >> Compression de protocole > 27/01/2007 04:31 LCP: >> Contrôle/Compression d¹adresse > 27/01/2007 04:31 Aucune réponse de l¹ordinateur distant. ------------------------------ Date: Sat, 27 Jan 2007 21:58:35 GMT From: Forrest Kenney Subject: Re: SHOW TERMINAL glitch (PPP) Message-ID: <45BBC28C.633B15D5@hp.com> Once more, show terminal used $GETDVI to mine most of the data from the terminal line. The rest are gotten using a sense mode or sense char $QIO. When ASN driver is in the picture it does not support the sense operations done by SHOW TERMINAL and thus the error. The reason the $GETDVI stuff looks reasonable it that ASNDRIVER clones much of the terminals UCB before the line is converted over. In a perfect world SHOW TERMINAL would be smart enough to say Invalid device when done against ASNx:. When we did the PPP work in the mid 90's it just did not seem worth it to go back and teach SHOW TERMINAL to do that either way you get an error. Forrest Ps. This is not guess work I wrote ASNDRIVER. Michael Moroney wrote: > > JF Mezei writes: > > >$ show term tta0 > >Terminal: _ASN10: Device_Type: Unknown Owner: No Owner > >Physical terminal: _TTA0: > > >Terminal Characteristics: > ... > >%SYSTEM-F-BADPARAM, bad parameter value > > >Why the error message at the bottom ? > > I played with PPP long (6+ years) ago and remember seeing this. I > looked into it enough to see $ show terminal issued 2 QIOs against > the terminal port, one (must have been a SENSEMODE) worked, the > other failed with the BADPARAM status. I don't remember what the other > function was or what information it was trying to retrieve. ------------------------------ Date: Sat, 27 Jan 2007 18:19:08 -0500 From: JF Mezei Subject: Re: SHOW TERMINAL glitch (PPP) Message-ID: <45bbde4b$0$9160$c3e8da3@news.astraweb.com> Forrest Kenney wrote: > a perfect world SHOW TERMINAL would be smart enough to say Invalid > device > when done against ASNx:. Note that on VAX, when I swicthed a line to SLIP, it would mark the terminal as a network device and SHOW TERMINAL would fail. Same with DDCMP lines (Decnet over serial lines). ------------------------------ Date: 27 Jan 2007 13:41:30 -0800 From: "issinoho" Subject: VAMP is back! Message-ID: <1169934090.937632.69740@s48g2000cws.googlegroups.com> Right. I finally got off my lazy ass and got this site running again. Why was it down? Well, following some OS patches last year, things started to go badly wrong on my Alpha. Apache processes were locking out in COLPG state (what the hell is that anyway?); TCPWare SSH was locking up; CDE wouldn't let me login, etc, etc. Anyway, it was sick. An attempted uninstall of the patches failed and I got fed up and buttoned it. What I did today was install OpenVMS in PRESERVE mode which forced out the patches and on first glance it seems to have done the trick. On first glance? Hmmmm. I already see a COLPG process which isn't good and SSH is still shot. The site does seem to be up and healthy though. This leaves me in a bit of a dilemma. On one hand the site is up & good, and if I treat it as a black-box then I can just leave it to do its stuff. However... broken things annoy me and as a an IT professional I really want to get the Alpha back to its best. So... I plan to migrate VAMP onto my Vista box (will be seamless to you guys), back things up and then reinstall my Alpha afresh. In the time that VAMP has been down I've missed multitudes of upgrades from VMS to Apache, MySQL to phpBB. I will take the opportunity to upgrade these items as well. I will of course move VAMP back onto VMS as soon as I can. Thanks for your patience and polite enquiries recently. I hope you continue to find this resource useful. As ever... http://vamp.issinoho.com ------------------------------ Date: 27 Jan 2007 14:27:03 -0800 From: "issinoho" Subject: Re: VAMP is back! Message-ID: <1169936823.109931.175120@s48g2000cws.googlegroups.com> On Jan 27, 10:21 pm, Arne Vajh=F8j wrote: > issinoho wrote: > > Apache processes were locking > > out in COLPG state (what the hell is that anyway?)http://h71000.www7.hp= .com/doc/73final/6552/6552pro_009.html > > Arne "Collided Page Wait, involuntary wait state; likely to indicate a=20 memory shortage, waiting for hard page faults" Well, it's definitely not memory, I've got 1Gb inside my PWS 433au. ------------------------------ Date: 27 Jan 2007 17:44:39 -0600 From: cornelius@encompasserve.org (George Cornelius) Subject: Re: VAMP is back! Message-ID: In article <1169934090.937632.69740@s48g2000cws.googlegroups.com>, "issinoho" writes: > I finally [...] got this site running again. > > Why was it down? Well, following some OS patches last year, things > started to go badly wrong on my Alpha. Apache processes were locking > out in COLPG state (what the hell is that anyway?); TCPWare SSH was > locking up; CDE wouldn't let me login, etc, etc. Memory issues? Someone else will be able to explain it, no doubt, but I have always thought of Collided Page Wait as a transient state, not something you got stuck in. I almost never see it, but then my systems all have plenty of memory, and paging and swapping don't happen much. I would expect that you need to do some memory tuning. Do you have enough page file space? Is it possible you need a swap file? Have you used pagefile for a system dump and left a lot of space allocated to the last dump? Do you need to buy more memory? What happens if you turn off the Extended File Cache? Also, I would like to know what $ MON IO and $ MON PAGE show, and whether you have done $ AUTOGEN SAVPARAMS and then $ AUTOGEN GETDATA SETPARAMS to let your system self-tune (If you do a DIFF on SYS$SYSTEM:SETPARAMS .DAT after the AUTOGEN run, you can see if there are any parameters it decided to drastically alter). -- George Cornelius cornelius PREPOSITION eisner.decus.org cornelius PREPOSITION mayo.edu ------------------------------ Date: Sun, 28 Jan 2007 00:29:57 +0000 (UTC) From: moroney@world.std.spaamtrap.com (Michael Moroney) Subject: VMS installation trick Message-ID: I knew how to do this for a while, but perhaps many here will find this trick useful. Let's say you have a "new" Alpha box with a severely limited hardware configuration, such as a DS10L from Island Computer or something. It has exactly one (blank) hard drive, maybe a CD-ROM, or maybe not. This is common hobbyist situation, due to the cheap (or even free!) DS10Ls available from Island or elsewhere. Let's say that you also have a VMScluster already (can be a 1 node cluster) but you want the new node to run a different version of VMS than the cluster or something. (say the cluster is running V8.2 and you want to install V8.3 on the new node) You have the installation kit on a spare drive in the cluster, but you don't have a second spare drive there. You don't want the new node to be part of the cluster, at least not permanently. The spare disk with the installation kit is a SCSI disk, but you can't connect the SCSI drive to the DS10L either because the one PCI slot is occupied, or you simply don't have a spare PCI SCSI adapter. How to get VMS onto the new drive? Well one way is to boot as a satellite and clone the cluster's system disk onto the new system's drive, but like I said, let's say you want to install a new version from scratch. To do that you boot from the installation kit disk. But it's in the wrong spot. Also, the VMS on the installation kit is configured standalone. Well, here's a trick, you can actually boot a cluster version of the installation kit by temporarily booting the system from the installation kit as a cluster node. Unsupported, of course. Here's how. Get the system's Ethernet address. >>> SHOW CONFIGURATION at the console. Use CLUSTER_CONFIG.COM on the cluster to add the node as a satellite. If you plan to add the node in the cluster, perhaps sometimes, use the SCSNODE/SCSSYSTEMID that you plan to use, or you can create a one-time use pair. Set the system to not have a local pagefile. Don't boot the new system yet. From now on, assume CLUSTER_CONFIG creates the root as , and I'll use as the example, but if it comes up with something else, use it instead. The installation kit disk must be MSCP-served. Mount the installation kit disk privately. (Replace KIT: with the name of the drive the kit is on) $ MOUNT KIT: label Change the sysgen parameters as follows: $ MC SYSGEN SYSGEN> USE KIT:[SYS10.SYSEXE]ALPHAVMSSYS.PAR SYSGEN> SET VAXCLUSTER 1 ! boot as a cluster, not standalone SYSGEN> SET VOTES 0 ! don't disturb quorum in the cluster SYSGEN> SET EXPECTED_VOTES n ! use the same as what the cluster uses SYSGEN> SET SCSNODE name ! same name as used with CLUSTER_CONFIG.COM SYSGEN> SET SCSSYSTEMID number ! same as CLUSTER_CONFIG.COM SYSGEN> WRITE KIT:[SYS10.SYSEXE]ALPHAVMSSYS.PAR You need to allow the system into the cluster. $ COPY /LOG SYS$SYSTEM:CLUSTER_AUTHORIZE.DAT KIT:[VMS$COMMON.SYSEXE] If the installation kit isn't mounted /NOWRITE, the procedure hangs later $ DISM KIT: $ MOUNT/SYSTEM/NOWRITE KIT: label Get into LANCP and tell it to boot from the kit: drive. MC LANCP LANCP> SET NODE /ROOT=KIT: Now you can boot the system. >>> BOOT EWA0 The standard installation procedure runs, but the system is part of the cluster, not standalone. One restriction is, when it asks for a volume label for the new system disk, it must be unique in the cluster. Normally, it could be anything. When done, and you shut the system down, you can delete the root if you wish. You can now boot the new system from its own disk. ------------------------------ Date: Sat, 27 Jan 2007 21:03:28 -0500 From: JF Mezei Subject: Re: VMS installation trick Message-ID: <45bc04d1$0$6025$c3e8da3@news.astraweb.com> Michael Moroney wrote: > SYSGEN> USE KIT:[SYS10.SYSEXE]ALPHAVMSSYS.PAR > SYSGEN> SET VAXCLUSTER 1 ! boot as a cluster, not standalone This is actually not necessary. As much as it may not be logical. The boot node's lancp will happily serve the boot file and config. And MSCP will be setup to give the satellite access to its boot disk/root. The satellite node will simply not make a request to join the cluster. (But I think that you still want the cluster_authorize.dat file located in the satellite node's sys$system). In my case, the only reason to turn off vaxcluster was a switch problem that didn't quite support vaxcluster stuff. That is how I found out it was possible. I had the CD image onto a real drive on the vax. And from the VAX: > MC LANCP SHOW NODE: BIKE (08-00-2B-87-3D-XX): MOP DLL: Load file: APB.EXE Load root: $3$DIA0: Boot type: Alpha satellite Once booted, it would execute the CD's menu and you can then install the software on the local drive which is totally "naked" and unused. ------------------------------ End of INFO-VAX 2007.055 ************************