1 INFO-VAX	Sun, 17 Jun 2001	Volume 2001 : Issue 334       Contents: Re: backup help  config tcpip on 7.3 : Re: config tcpip on 7.3 - BIND resolvers *are* nameservers" Re: help installing 7.3 in cluster" Re: help installing 7.3 in cluster Intrusion & Password Checking  Re: modern day x-terminals Re: Mozilla 0.9.1 ) Re: Mozilla 0.9.1 - faster but no helpers ) Re: Mozilla 0.9.1 - faster but no helpers 7 Re: openVMS concurrent user licences (v7.2 on MicroVAX) = Re: OT: I'm just a "pup"!  Was:Re: DEC vs IBM, story of Moses  Re: POSTs via SSL to Apache % QIO Socket Problem - Buffer Overflows # Re: Setting up Mail, WEB page, etc. 6 Re: Shadowed system disk upgrade procedure - VMS 5.5-2  F ----------------------------------------------------------------------  + Date: Sun, 17 Jun 2001 10:23:48 +0000 (UTC) ' From: david20@alpha2.mdx.ac.uk (D.Webb)  Subject: Re: backup help+ Message-ID: <9gi0fj$7pf$2@aquila.mdx.ac.uk>   [ In article <9lolitghkr7jt7mq86v5r3juee2jite1nf@4ax.com>, mike <cdunevant@yahoo.com> writes: C >Grad student working an internship...totally new to VMS (as are my $ >full-time co-workers, appearently). > C >I'm using VMS 7.2 on an Alpha server and trying to backup the data F >disc to a sequence of tapes.  After reading the manual several times,8 >I came up with the following (which works pretty well). > D >$ BACKUP/IMAGE/IGNORE=label_processing  DUA1:  MUB6:data.sav/rewind > G >The only problem is...this doesn't back up files currently in use.  To 9 >capture these, it looks like I need to use the qualifier  >IGNORE=interlock. > C >I tried to specify both IGNORE qualifiers at the same time several E >ways, but always ended up with a syntax error.  Can someone point me ) >in the right direction?  Thanks a bunch!  >  >Mike Dunevant     Try     D $ BACKUP/IMAGE/IGNORE=(label,interlock)  DUA1:  MUB6:data.sav/rewind      
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------  % Date: Sun, 17 Jun 2001 09:20:20 -0700 ! From: Tom Linden <tom@kednos.com>  Subject: config tcpip on 7.39 Message-ID: <CIEJLCMNHNNDLLOOGNJIAEFJCMAA.tom@kednos.com>   3 Just installed 7.3 and am struggling through tcpip.   L "Before configuring your system as a BIND resolver, you should first be sureE that ther is at least on system on the network configured either as a  primary $ or secondary server for this domain"  J What does this mean?  Nameservers?  If so, then they must mean defined and not configured?   - Documentation is of little help for a novice.    ------------------------------    Date: 17 Jun 2001 12:56:53 -05007 From: hamilton@encompasserve.org (Bradford J. Hamilton) C Subject: Re: config tcpip on 7.3 - BIND resolvers *are* nameservers 3 Message-ID: <YUXkd1QhOtGy@eisner.encompasserve.org>    Hi Tom,    You are correct in your assumption regarding Nameservers.  If you know of reliable namesevers in your network, then define them asb your "BIND resolvers".  You must specify them with an FQDN (Fully Qualified Domain Name), as well.   --Brad  ^ >In article <CIEJLCMNHNNDLLOOGNJIAEFJCMAA.tom@kednos.com>, Tom Linden <tom@kednos.com> writes:5 > Just installed 7.3 and am struggling through tcpip.  > N > "Before configuring your system as a BIND resolver, you should first be sureG > that ther is at least on system on the network configured either as a 	 > primary & > or secondary server for this domain" > L > What does this mean?  Nameservers?  If so, then they must mean defined and > not configured?  > / > Documentation is of little help for a novice.  >  >    ------------------------------  % Date: Sun, 17 Jun 2001 09:38:11 +0100   From: Paul Sture <paul@sture.ch>+ Subject: Re: help installing 7.3 in cluster + Message-ID: <VA.000003c7.6fc10edb@sture.ch>   M In article <rdeininger-1606011614150001@user-2ivechd.dialup.mindspring.com>,   Robert Deininger wrote:  > F > In article <CIEJLCMNHNNDLLOOGNJICEFBCMAA.tom@kednos.com>, Tom Linden > <tom@kednos.com> wrote:  > K > > First time at it, installing 7.3 on virgin HW, this is connect on a lan  > > which had a cluster.L > > I went through first step and rebooted, and am now stuck on the question > > & > > Enter this cluster's group number: > > I > > I assume that this is not arbitrary but is a number know to the other  > > members of the cluster.  > >  > > How do I determine this? > I > If you are making a new cluster, you can pick anything you want for the H > cluster group number and the cluster password.  But make sure the pairJ > does not match any other VMS cluster on the same LAN.  These days, there@ > are rarely multiple clusters on the LAN, so that part is easy. > N I beg to differ on that last sentence. At work we have multiple clusters (and N IBMs, Suns, PCs etc) on our LAN. Some are 20 or so miles away, so should that 	 read WAN?   R Last year I had a phone call from someone trying to track down a system which was P trying to join one of their production clusters. It was failing to join because P the password was wrong, but it was giving that production cluster a very severe G performance hit, just because someone had picked the same group number.   I The message that gave me was very clear - cluster group numbers should be F allocated with the same care one would give to DECnet or IP addresses. ___ 
 Paul Sture Switzerland    ------------------------------  # Date: Sun, 17 Jun 2001 16:26:28 GMT - From: "Maria Hanson" <hansonkm@earthlink.net> + Subject: Re: help installing 7.3 in cluster F Message-ID: <US4X6.13165$Kq4.1289913@newsread2.prod.itd.earthlink.net>  G Paul is correct and from experience I can tell you that a DEV cluster 2  floors up gave some hartache to * my system admin group for a few weeks. :-)  
 -Maria Hanson   - "Paul Sture" <paul@sture.ch> wrote in message % news:VA.000003c7.6fc10edb@sture.ch...  > In articleA <rdeininger-1606011614150001@user-2ivechd.dialup.mindspring.com>,  > Robert Deininger wrote:  > > H > > In article <CIEJLCMNHNNDLLOOGNJICEFBCMAA.tom@kednos.com>, Tom Linden > > <tom@kednos.com> wrote:  > > I > > > First time at it, installing 7.3 on virgin HW, this is connect on a  lan  > > > which had a cluster.E > > > I went through first step and rebooted, and am now stuck on the  question > > > ( > > > Enter this cluster's group number: > > > K > > > I assume that this is not arbitrary but is a number know to the other  > > > members of the cluster.  > > >  > > > How do I determine this? > > K > > If you are making a new cluster, you can pick anything you want for the J > > cluster group number and the cluster password.  But make sure the pairL > > does not match any other VMS cluster on the same LAN.  These days, thereB > > are rarely multiple clusters on the LAN, so that part is easy. > > J > I beg to differ on that last sentence. At work we have multiple clusters (andJ > IBMs, Suns, PCs etc) on our LAN. Some are 20 or so miles away, so should that > read WAN?  > I > Last year I had a phone call from someone trying to track down a system 	 which was I > trying to join one of their production clusters. It was failing to join  because J > the password was wrong, but it was giving that production cluster a very severeI > performance hit, just because someone had picked the same group number.  > K > The message that gave me was very clear - cluster group numbers should be H > allocated with the same care one would give to DECnet or IP addresses. > ___  > Paul Sture
 > Switzerland  >  >    ------------------------------  % Date: Sun, 17 Jun 2001 13:47:19 +0200 2 From: martin@radiogaga.harz.de (Martin Vorlaender)& Subject: Intrusion & Password Checking; Message-ID: <3b2c98c7.524144494f47414741@radiogaga.harz.de>    Hi all,   C for a web-based password change utility I'm writing (yes, I know of I Richard Levitte's and Tom Wade's programs), I'm a little lost for details 8 of intrusion and password dictionary / history checking.  E Intrusion: Is it possible to use $SCAN_INTRUSION for checking without F having the service insert a suspect record (as I read it, the entry inE the System Services Reference suggests this is possible)? Or do I use  $SHOW_INTRUSION for that?   J Dictionary: Does one use a key of the actual length of the password-to-be,C or fill it up with NULs to the record size (32)? The first approach @ would find "xyz" even though only "xyzzy" is listed in the file.  B History: Is the record format documented somewhere? I believe it's8   struct { char username[32]; quadword hashed_pw[...]; }H where username is the key and the number of hashed PWs can be calculatedG from rab$w_rsz returned for a record. How does one compare a PW against G the hashed PWs? I.e. where does one get the algorithm and the salt from % to $HASH_PASSWORD the password-to-be?   % Many thanks in advance for any clues.    cu,    Martin  L P.S.: Although I'm implementing this for use at my employer, I'm programmingK this in my spare time. When the utility is finished, I'm going to put it in  the public domain. --  G  Your mouse has moved.     | Martin Vorlaender  |  VMS & WNT programmer 4  Windows must be restarted | work: mv@pdv-systeme.deH  for the change to take    |    http://www.pdv-systeme.de/users/martinv/;  effect. Reboot now? [OK]  | home: martin@radiogaga.harz.de    ------------------------------  % Date: Sun, 17 Jun 2001 11:47:41 -0500 % From: Keith Brown <kbrown780@isd.net> # Subject: Re: modern day x-terminals / Message-ID: <tipnhuhvgp3fe4@corp.supernews.com>    Malcolm Dunnett wrote:  0 > In article <M7cyShwo3vNd@tachxxsoftxxconsult>,D >    wayne@tachysoft.xxx.563109.killspam.015d (Wayne Sewell) writes: >  >>  D >> Does anybody know of a currently supported x-terminal?  Basically >> somethingH >> like the vxt 2000+, but with current hardware and software.  I have a >> xvt, but H >> I don't really use it.  It's slow and doesn't have that much memory. L >> Also, most of my network is 100 megabit, but of course the ancient vxt is >> only 10.  >>   >>   >>  L >> Before anybody mentions emulation on PCs, I do *not* want to use a PC for >> this. >>  K >     What specifically don't you like about using a PC. I ask this because  >     itK > seems that there's not a lot in the way of dedicated X terminals anymore. F > The company we used to get ours from ( NeoStations from Neoware ) noH > longer seems to build a "classic X terminal" - but they have a product: > based on embedded linux which includes the X11 protocol. > H >    I looked at an earlier version about a year ago and while it mostly >    worked I > it wasn't quite "ready for prime time". I'd suspect that in the ensuing L > year they've probably improved it. I suspect this may be the product we'll( > go with next time we need X terminals. > > >    More info at :  http://www.neoware.com/products/4000.html >  > K I would also recommend trying Linux. I use it from home to connect with my  L VMS systems at work.  It does native X11 and works very well, and of course  the price is free. --   Keith Brown  kbrown780@isd.net    ------------------------------   Date: 17 Jun 2001 07:17:10 CDT= From: wayne@tachysoft.xxx.563109.killspam.015d (Wayne Sewell)  Subject: Re: Mozilla 0.9.1. Message-ID: <GkLehKAvR5c+@tachxxsoftxxconsult>   In article <rdeininger-1606011301140001@user-2ive6ri.dialup.mindspring.com>, rdeininger@mindspring.com (Robert Deininger) writes: 0 > In article <SaHi3RGc9HXz@tachxxsoftxxconsult>,@ > wayne@tachysoft.xxx.563109.killspam.015d (Wayne Sewell) wrote: > J >> In article <3b2a4383.39153439@news.process.com>, goathunter@goatley.com > (Hunter Goatley) writes:E >> > On Fri, 15 Jun 2001 16:49:07 GMT, system@SendSpamHere.ORG (Brian  > Schenkenberger,  >> > VAXman-) wrote: >> >  3 >> >>>You need Multinet V4.2A with a patch or V4.3.  >> >>>E >> >>>As of 0.9.1, Mozilla on OpenVMS is using BSD 4.4 style sockets.  >> >>   >> >>What was it using before??? >> >> P >> > It was using BSD 4.3-style sockets before.  MultiNet and TCPware do not yetN >> > have full support for BSD 4.4-style sockets.  Until they do, Mozilla will0 >> > not work correctly on MultiNet and TCPware. >>  Q >> What a stupid thing to do, making such a major change at the last minute. They Q >> should be concentrating on fixing bugs in preparation for the released version N >> instead of adding new features/functionality.   They certainly shouldn't beH >> doing stuff that causes it not to work on people's existing stacks.   > Influence ) >> from the billyworld mindset, no doubt.  >>  I >> It looks like we're back to not having a decent browser for vms again.  > G > I don't know any details of this socket stuff, but it's possible they J > needed some functionality that requires BSD 4.4.  Remember how everybodyJ > has been complaining about problems with Mozilla?  Maybe this fixes some
 > of them. > I > If it was easy, maybe you or I would have already done it.  The Mozilla D > folks are making progess; I'll give them the benefit of the doubt.    N If they made the change because they *had* to, I concede your point.  Only if.     --  O =============================================================================== M Wayne Sewell, Tachyon Software Consulting  (281)812-0738  wayne@tachysoft.xxx : http://www.tachysoft.xxx/www/tachyon.html and wayne.html  K change .xxx to .com in addresses above, assuming you are not a spambot  :-) O =============================================================================== K Hotel guy (after bed demolition):  That bed goes back to Henry the eighth!! O    Curly: That's nothin'!  We had a bed go back to Sears and Roebuck the fifth!    ------------------------------    Date: 17 Jun 2001 12:34:41 +0100O From: pmoreau@dev.ath.cena.fr (Patrick MOREAU, CENA Athis, Tel: 01.69.57.64.40) 2 Subject: Re: Mozilla 0.9.1 - faster but no helpers  Message-ID: <uf40AIkho0d1@sable>  4 In article <9uZ0US8xCzf4@eisner.encompasserve.org>, 9 hamilton@encompasserve.org (Bradford J. Hamilton) writes:  > N > Mozilla 0.9.1 is considerably faster than 0.9 on my DPW 433au.  I have foundM > only one issue so far: the inability to type an address into the URL window P > without crashing the browser with some kind of error having to do with threadsS > (no details - the box is at work, and I haven't had time to document the problem. P > The box is 7.2-1 with most of the level-1 ECO's.  I applied the latest THREADS* > patch, but the problem keeps happening).  G Yes its faster, thanks to a better management of files into the cache.    H I have no problems with the url window (DS10 VMS 7.2-1, Motif 1.2-5) butL helpers applications don't work now. I've deleted all helpers i've added and% re-created them without any success.     Patrick  --O =============================================================================== O pmoreau@cena.dgac.fr  (CENA)     ______      ___   _           (Patrick MOREAU) 4 moreau_p@decus.fr (DECUS)       / /   /     / /|  /|J CENA/Athis-Mons FRANCE         / /___/     / / | / |   __   __   __   __  N BP 205                        / /         / /  |/  |  |  | |__| |__  |__| |  |N 94542 ORLY AEROGARE CEDEX    / /   ::    / /       |  |__| | \  |__  |  | |__|N http://www.ath.cena.fr/~pmoreau/            http://www.multimania.com/pmoreau/O ===============================================================================    ------------------------------  # Date: Sun, 17 Jun 2001 13:42:32 GMT = From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) 2 Subject: Re: Mozilla 0.9.1 - faster but no helpers0 Message-ID: <009FDA9A.4B86AF85@SendSpamHere.ORG>  r In article <uf40AIkho0d1@sable>, pmoreau@dev.ath.cena.fr (Patrick MOREAU, CENA Athis, Tel: 01.69.57.64.40) writes:5 >In article <9uZ0US8xCzf4@eisner.encompasserve.org>,  : >hamilton@encompasserve.org (Bradford J. Hamilton) writes: >>  O >> Mozilla 0.9.1 is considerably faster than 0.9 on my DPW 433au.  I have found N >> only one issue so far: the inability to type an address into the URL windowQ >> without crashing the browser with some kind of error having to do with threads T >> (no details - the box is at work, and I haven't had time to document the problem.Q >> The box is 7.2-1 with most of the level-1 ECO's.  I applied the latest THREADS + >> patch, but the problem keeps happening).  > H >Yes its faster, thanks to a better management of files into the cache.  >XI >I have no problems with the url window (DS10 VMS 7.2-1, Motif 1.2-5) but M >helpers applications don't work now. I've deleted all helpers i've added ande& >re-created them without any success.   3 Yeap!  It's much faster at crashing the X server...d  K When will this thing actually run?  How about adding some debugging so that J this server crasher can be made to function without crashing the server???   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMC            yO city, n., 1. a place where trees are cut down and streets are named after them.    ------------------------------  % Date: Sun, 17 Jun 2001 14:10:52 +0200e" From: "Hans Vlems" <hvlems@iae.nl>@ Subject: Re: openVMS concurrent user licences (v7.2 on MicroVAX)( Message-ID: <9gi6l6$2r4$1@news.IAEhv.nl>  G Your system has two VMS licenses loaded: a type A license, which allows 
 unlimited useqH as well as a type C license for 12 users. But the type C license in this case is useless,= assuming that the type A license is valid/legal for the 3185.a   Hans  . Jon Pickup <jpickup@bocs.com> wrote in message9 news:EQ1W6.20568$jd1.1064127@monolith.news.easynet.net...gG > Can anyone help me out with how to establish how many concurrent usern > licences I have on my system?t >h > SHOW LIC /CHARGE shows...t >e+ > VMS/LMF Charge Information for node BELF1s5 > This is a MicroVAX 3100-85, hardware model type 490 G > Type: A, Units Required: 60 (VAX/VMS Capacity or OpenVMS Unlimited ort Base)l1 > Type: B, * Not Permitted * (VAX/VMS F&A Server). > C >    ---->   Type: C, Units Required: 100 (VAX/VMS Concurrent User)  >w2 > Type: D, * Not Permitted * (VAX/VMS Workstation)C > Type: E, Units Required: 230 (VAX/VMS System Integrated Products)k4 > Type: F, Units Required: 20 (VAX Layered Products)' > Type: G, * Not Permitted * (Reserved) 5 > Type: H, * Not Permitted * (Alpha Layered Products)e0 > Type: I, Units Required: 20 (Layered Products) >a > ...and SHOW LIC shows....  >0  > Active licenses on node BELF1:= > ------- Product ID -------- ---- Rating ----- -- Version --O@ > Product Producer Units Avail Activ Version Release Termination. > BASE-VMS-250136 DEC 60 0 A 0.0 (none) (none)/ > NET-APP-SUP-200 DEC 200 F 0 0.0 (none) (none)u0 > PWLMDOSCC05.01 DEC 100 0 100 0.0 (none) (none)0 > PWNWXXXFP05.00 DEC 100 0 100 0.0 (none) (none) >e4 >   ---->  VMS-USER DEC 1200 0 100 0.0 (none) (none) > K > Compaq reckoned that I divide the 1200 by the 100 to get 12, then add oneoH > for the base user, to get 13 as an upper limit. But then they couldn'tJ > explain how come we seemed to be routinely exceeding 13. Which leaves me > nowhere again... >n3 > OpenVMS User Processes at 12-JUN-2001 12:25:25.79 6 > Total number of users = 16, number of processes = 17' > Username Interactive Subprocess Batchp > AINE 1 > BOCSMANAGER 1  > CAROLYN 1o > GOFIONA 2m	 > GOJOE 1y > GOJOSEPH 1 > IAN 1e
 > JACKIE 1
 > JANINE 1 > LYCAROL 1 	 > PIPPA 1y	 > ROBIN 1a > ROSALEEN 1
 > SPACETIME 1n	 > SUSAN 1w > VALERIE 1d >  > Many thanks in advance,O >J > Jon. >o >e   ------------------------------  # Date: Sun, 17 Jun 2001 13:13:19 GMTa= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)sF Subject: Re: OT: I'm just a "pup"!  Was:Re: DEC vs IBM, story of Moses0 Message-ID: <009FDA96.36A338DC@SendSpamHere.ORG>  i In article <lsAlm9XrpE+r@eisner.encompasserve.org>, malmberg@encompasserve.org (John E. Malmberg) writes:h1 >In article <009FD904.6B325A13@SendSpamHere.ORG>, @ >system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) writes:4 >> In article <3B28E7DC.9515F5DF@BlueBubble.UK.Com>,4 >Roy Omond <Roy.Omond@BlueBubble.UK.Company> writes:) >>>"Brian Schenkenberger, VAXman-" wrote:g >>>> >>M >> Presently, due to accrued mass storage needs, this final microVAX is now avL >> lofty 33 lbs.  This model is presently capable of interfacing with a veryM >> rudimentary communication protocol but has occasional bursts of error thateL >> makes it very difficult to comprehend the intended program function.  TheN >> reams of output generated by this model still require constant operator in- >> tervention for disposal.n > L >It can be important to try to comprehend the communication protocol at that >age.r  L I have a teenage daughter too.  Believe me, it is far more difficult to com-' prehend and decipher that protocol!  ;)e    O >When my youngest was that age, if occurred to me that the phrase "push button"rI >was being repeated, and was being accompanied by a slight clicking soundhE >near the reset and power buttons of the computer I was working on...l  K He is trying.  He's still having problems with a few basic speech sounds.  sM For example, the "sw" at the beginning of swing.  He's been saying "Trrr"ing,AK with quite a roll on the "r" for swing.  It's not a high level protocol butaL at least it's not in some encrypted and coded protocol like that used my theL teenager.  In 25 days she's off on a three week tour of Europe as part of a L student ambassador program.  I can only hope that whilst spending 10 days in0 the UK, a bit of the King's English is absorbed.   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMr             O city, n., 1. a place where trees are cut down and streets are named after them.l   ------------------------------  % Date: Sun, 17 Jun 2001 14:08:45 +0200 2 From: martin@radiogaga.harz.de (Martin Vorlaender)$ Subject: Re: POSTs via SSL to Apache; Message-ID: <3b2c9dcd.524144494f47414741@radiogaga.harz.de>n  , Marty Kuhrt (kuhrt@encompasserve.org) wrote:@ > Does anyone have an example 3GL program that can emulate POSTsD > via SSL to the Apache (CSWS 1.0-1) webserver?   I'm trying to testA > out some forms I have created to see if they respond correctly.gD > I can do them manually by calling up the cgi page and filling out C > the blanks, but I'd like to automate it to do regression testing.i >rG > I'm running CSWS 1.0-1 (Apache 1.3.12, mod_ssl/2.6.2, OpenSSL/0.9.5),o= > Multinet 4.3A-X, and VMS 7.2-1 on an Alphaserver 200 4/233.   E A perl program would probably be the easiest to write. You'll need to E install the libwww-perl (LWP) and Crypt::SSLeay modules on top of thenD standard perl installation (available on CPAN, http://www.cpan.org).   Hope it helps,   Martin -- uD                         | Martin Vorlaender  |  VMS & WNT programmer1  VMS is today what      | work: mv@pdv-systeme.dedE  Microsoft wants        |    http://www.pdv-systeme.de/users/martinv/ 8  Windows NT 8.0 to be!  | home: martin@radiogaga.harz.de   ------------------------------  # Date: Sun, 17 Jun 2001 12:58:26 GMTr- From: rsmith@arel.com.dot-a-u (Russell Smith)e. Subject: QIO Socket Problem - Buffer Overflows: Message-ID: <SP1X6.66290$hV3.121356@newsfeeds.bigpond.com>  N I am currently having problems with losing bytes under load with socket under : TCPIP Services using a BASIC program.  Versions as follows  A   DIGITAL TCP/IP Services for OpenVMS Alpha Version V5.0A - ECO 2o=   on a Digital Personal WorkStation  running OpenVMS V7.2-1  i  0   ... also tried on a machine with V5.0A - ECO 3     BASIC V1.3-006  O The problem seems to occur when the buffer fills.  The code is using $QIOW and  M uses INET$C_MSG_NBIO to return immediately regardless of whether data can be  P written.  If the byte count in word 0 of the IOSB is less than the message size P I submit a IO$_SETMODE|IO$M_WRTATTN and wait for an AST to fire letting me know  to send again.  P After some playing around I noticed that it appeared that each time the problem O was occuring the returned status was SS$_SUSPENDED and the sent byte count was sO 35.  This appeared to be wrong, as the other sice of the link received none of -O the 35 bytes.  A change to resend ALL in the case of a SS$_SUSPENDED seemed to   help.e  G Is there any significance in the fact that the C Socket error code for tP SS$_SUSPENDED in the online TCP docs is 35, and this is the byte count returned " when SS$_SUSPENDED is returned ???  H As I said earlier, the check for suspended helped, previously I treated J SS$_NORMAL and SS$_SUSPENDED the same way.  I am still experiencing (less N frequent) problems though with bytes going missing, when sending under load.  L Are there any gotchas that I am missing, or am making incorrect assumptions  about the return statuses ???z  F Is is possible to use IO$_SENSEMODE with sockets, in a similar way to O mailboxes, to check the amount of space in the socket buffer, before doing the   $QIO ?  # Any help would be much appreciated.h   Regards,  
 Russell Smithm Arel Asia Pacific    ------------------------------  + Date: Sun, 17 Jun 2001 09:55:02 +0000 (UTC) ' From: david20@alpha2.mdx.ac.uk (D.Webb) , Subject: Re: Setting up Mail, WEB page, etc.+ Message-ID: <9ghupl$7pf$1@aquila.mdx.ac.uk>   \ In article <3B2A5849.B0012B56@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes: >Tom Linden wrote: >> tJ >> Installed  a T1 yesterday and as a result will need to do POP mail, Web
 >> server. >. > K >When you get the TCPIP V5.x software for VMS, it comes with a POP and SMTPw >servers for email.e >f  L Earlier versions of this software eg UCX 4.2 also support both SMTP and POP.    L >You can then get Apache web server, or the better OSU DECthreads web serverI >was was designed for VMS and has been in operationf or a long time (very0! >stable software). Both are free.  >e  B You might also want to look at a web based interface for VMS mail.K The public domain YAHMAIL program works with most (if not all)  of the VMS e3 based webservers. See http://wasd.vsm.com.au/wasd/.i         I >The one thing that doesn't come out of the box with VMS is IMAP support.  >sL >TCPIP Services also gives you a DHCP server, NFS server and client, TELNET,
 >FTP etc etc.u    1 For IMAP support you will need to look at either v   1) Purchasing PMDF      F which is an extremely good piece of mailhub software. It also includesA POP and IMAP servers together with an improved VMS MAIL and PINE.n  K However although relatively inexpensive compared to most other similar MTA l6 products it may well be overkill for what you require.   or  L 2) Use one of the other VMS TCPIP stacks I believe both TCPWARE and Multinet provide IMAP servers.     = These products are all available from http://www.process.com/     E Unfortunately I am unaware of any public domain IMAP4 server for VMS.   
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------  % Date: Sun, 17 Jun 2001 02:05:32 -0400   From: John Santos <JOHN@egh.com>? Subject: Re: Shadowed system disk upgrade procedure - VMS 5.5-2.6 Message-ID: <1010617015415.48727A-100000@Ives.egh.com>  ' On 15 Jun 2001, John E. Malmberg wrote:n  5 > In article <wUbW6.120$kQ2.22703@news-west.eli.net>, @ > "frank brown" <frank.brown@ci.seattle.wa.unitedstates> writes:D > > I want to upgrade our shadowed VMS 5.5-2 system disk from a pairD > > of RF31s to RF72s.  There is no room for additional disks on ourD > > dual DSSI-bus 2-VAX4300 cluster (it's maxed out with disks) so I( > > plan to perform the following steps: > E > I would recommend if you have not done so already to make sure thatL@ > you have standalone backup installed on one of your non-systemF > data disks, or have some way to boot your system from an OpenVMS VAX	 > CD-ROM.i > D > For this type of move having standalone backup on your system diskF > will also work.  The recomendation of having it on a non-system diskK > is so that you can quickly boot it if you need to restore the system disko3 > something unlikely with a shadowed configuration., > > > > 1. Dismount one of the existing system disk shadow members4 > > 2. Shut everything down (CPUs and disk cabinets). > > 3. Replace the dismounted RF31 with a RF72 > E > If you have shutdown both your systems, the fastest method would bei > to:r > E >   4. Boot standalone backup from a hard disk (preferred) or CD-ROM.  > P >   5. Use STANDALONE BACKUP to backup the remaining system disk to the new RF72  E If S/A backup is on the system disk, you can boot s/a backup and copyWD the system disk (i.e. the one you booted from) to another disk.  TheB disk can be write-locked (if it has an external write-lock switch)E through this whole process.  S/A backup does not need write access totG the boot disk.  You can even RESTORE to the boot disk using S/A backup,rC though you may need to reboot S/A backup afterwards, if you want todH do more backup operations.  (This is essential to doing a backup/restoreD of a single-disk system to/from tape.  You can install S/A backup onF the system disk, boot it, backup the disk to tape, restore the tape toE the same disk (defragmenting it in the process) and reboot the disk.)p  C Since the original poster said he didn't have any spare DSSI slots,a@ backing up the S/A backup disk is the only way to accomplish theB task unless he temporarily replaces one of the data disks with the? new disk (temporarily).  In a process like this, I would try touD minimize moving disks around.  Each time you do it is an opportunity6 to bend a pin in a connector, or cause other problems.  B Standalone backup requires only about 850 blocks when installed on@ the system disk, but requires a whole (minimal) VMS installation? when installed on a non-system disk, since it shares 99% of the  files with the regular O/S.=  ! > > 6. Shut everything down again=K >   7. Install the other RF72 to the slot the first RF31 initially occupiedy >   8. Power up all the boxess9 >   9. Reboot a VAX to use the RF72 that you restored to.f9 >  10. Restore your system disk shadow set configuration. H >  11. You may now boot the second VAX to the RF72 that you restored to. > N > > If the VAX boots OK then once the system is up, check to see if the systemG > > disk thinks it's part of a shadow set.  (Will it?)  If not, perform.O > > MOUNT/SYSTEM/SHADOW=(...) to mount the other new disk, to initiate a shadow 	 > > copy.n > O > IIRC: When you have manually broken the system disk shadow, you must manuallyeI > put it back.  Of course the state of shadowing for your disks should beb; > checked after every boot as part of your standard checks.n > O > > Am I missing anything here?  Do I need to perform device node operations ath- > > the console (>>>) level on the new disks?  > L > As with all DSSI disks, you must make sure that (The spelling of these are > from memory):tE > FORCEUNI, ALLCLASS, and UNITNUM are set correctly on the new disks.w >  > -Johni > wb8tyw@qsl.network > Personal Opinion Only  >  >    -- n John SantosI Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------   End of INFO-VAX 2001.334 ************************