1 INFO-VAX	Mon, 22 Mar 2004	Volume 2004 : Issue 161       Contents: Re: Gartner  Re: Gartner 3 Re: Help - Pathworks & Windows 2000 Advanced Server  Re: HSV110 replacement question  NFS mount on VMS RE: NFS mount on VMS Re: NFS mount on VMS Re: NFS mount on VMS Re: NFS mount on VMS& Re: OpenVMS Cluster  Oracle Dataguard?G RE: OpenVMS virus proof while other disks get zapped by latest viruses!  Re: Quorum Disk Question RE: Quorum Disk Question Re: Securing files under VMS Re: Securing files under VMS Re: Securing files under VMS Re: Server hang - compute bound & Re: Someone tried to crack my Alpha...& Re: Someone tried to crack my Alpha... Spam form kerry  Re: Spam form kerry  Re: Spam form kerry $ Re: VAX clock cycles per instruction$ Re: VAX clock cycles per instruction$ Re: VAX clock cycles per instruction* Re: Wants to run autogen evertime it boots* Re: Wants to run autogen evertime it boots* Re: Wants to run autogen evertime it boots8 Re: why isn't there room on this disk for the swap file?8 Re: why isn't there room on this disk for the swap file?8 Re: why isn't there room on this disk for the swap file?8 Re: why isn't there room on this disk for the swap file?8 Re: why isn't there room on this disk for the swap file?8 Re: why isn't there room on this disk for the swap file?8 Re: why isn't there room on this disk for the swap file?8 Re: why isn't there room on this disk for the swap file?8 Re: why isn't there room on this disk for the swap file?8 Re: why isn't there room on this disk for the swap file?  F ----------------------------------------------------------------------    Date: 22 Mar 2004 07:36:09 -0600+ From: young_r@encompasserve.org (Rob Young)  Subject: Re: Gartner3 Message-ID: <cRfL4t8J+4Vz@eisner.encompasserve.org>   V In article <405E62B5.DF2CCEC5@istop.com>, JF Mezei <jfmezei.spamnot@istop.com> writes:  N > Do Gartner customers actually use Gartner reports to set directions, or just2 > help justify decisions they have already taken ?  F 	Decisions already taken?  That would be poor strategy, and backwards.9 	I'm sure some dysfunctional organizations work that way.    				Rob    ------------------------------  # Date: Mon, 22 Mar 2004 13:56:26 GMT # From: "John Smith" <a@nonymous.com>  Subject: Re: GartnerF Message-ID: <egC7c.405$uz5.89@twister01.bloor.is.net.cable.rogers.com>  8 "Rob Young" <young_r@encompasserve.org> wrote in message- news:cRfL4t8J+4Vz@eisner.encompasserve.org... 4 > In article <405E62B5.DF2CCEC5@istop.com>, JF Mezei# <jfmezei.spamnot@istop.com> writes:  > K > > Do Gartner customers actually use Gartner reports to set directions, or  just4 > > help justify decisions they have already taken ? > G > Decisions already taken?  That would be poor strategy, and backwards. : > I'm sure some dysfunctional organizations work that way.  K I've seen major banks work that way in the past. They'd shop the street for G opinions from the consulting arms of the major auditing firms and other K industry consultants until they got the opinion they wanted in writing, and ) toss the remaining opinions in the trash.    ------------------------------  % Date: Mon, 22 Mar 2004 12:59:48 +0000 % From: David Gray <police@spamcop.net> < Subject: Re: Help - Pathworks & Windows 2000 Advanced Server8 Message-ID: <vcot50d6e6ri75sote3qmqao9lupej1jo5@4ax.com>   Hi Roy,   G >Are you quite sure that DATABASE01 had the correct address for Jerry ? 0 >What does "nslookup jerry" give on DATABASE01 ?  E All I get from NSLOOKUP is can't find host error messages, even after . a seemingly successful SET  DOMAIN=extenza-w1    Server:  localhost Address:  127.0.0.1   3 *** localhost can't find jerry: Non-existent domain     E >If it gave a different address than Jerry's *real* address, then you : >might have been (successfully) pinging the wrong machine.  5 I'm pretty sure the IP address was correct for JERRY.   = >> Anyone got any ideas on what might have been the problem?   >  >In a word: Windows !   8 Yeah :-)   All works ok now but what a faffing around.     Cheers,  	Dave.     ------------------------------  % Date: Mon, 22 Mar 2004 09:44:03 -0600 ( From: David Harrold <DHarrold@wi.rr.com>( Subject: Re: HSV110 replacement question8 Message-ID: <k82u50tvoo5p7vossmih8qca1o4456oo96@4ax.com>  L On Sat, 20 Mar 2004 11:55:34 -0700, John Nebel <john.nebel@csdco.com> wrote:   Hi John,  C >Is replacing one of an HSV pair as simple as shutting it down and  > >connecting another controller?, then powering up the new one? > C >I haven't been able to find instructions for this anywhere.  Just  + >thinking ahead, in case that is necessary.  >   L I've actually had to do this.  Yes it is possible.  It can even be done withH I/O still happening on the EVA.  The caution I received was to have your! environment as quiet as possible.   F If I can find the document HP sent me at the time, I'll send it to youN offline, but yes, you just shutdown the controller to be replaced, swap it for% the new one and start the new one up.    >John Nebel    Hope that helps.   Dave Harrold    N ..............................................................................N David Harrold                              E-Mail: David.Harrold at aurora.orgJ Lead Software Systems Engineer              Phone:          (414) 647-6204J                                             Pager:          (414) 941-4634J Aurora Health Care                            Fax:          (414) 647-4999 3031 W. Montana Street Milwaukee, WI 53215    ------------------------------    Date: 22 Mar 2004 01:15:41 -0800- From: rajarao1106@yahoo.com (Raja Rao Sindhe)  Subject: NFS mount on VMS = Message-ID: <de8f9c73.0403220115.3dd1169a@posting.google.com>    Hi,   C    We exported files from clearcase to OpenVMS, mounted files using & NFS mount ($tcpip mount dnfs2:[ab.src]0 /path="/view/abc_view/ab/src"/host="<hostname>",F able to see files from Unix system on OpenVMS system, but problem I amE facing is OpenVMS adding '$' sign beyond and after Capital letters in E the filenames, I tried on ODS2 disk, How to avoid '$ in the filename. & later I tried on ODS5 also by setting   $ set proc /parse_style=extended $ set proc /case_look=sensitive   # Do we need set some more variables. ) Any input on this, is greatly appreciated    Waiting for reply, Raja rao sindhe    ------------------------------  % Date: Mon, 22 Mar 2004 11:15:27 -0500 * From: "Brian Tillman" <tillmabg@yahoo.com> Subject: RE: NFS mount on VMS ' Message-ID: <012FFD7F.C22236@yahoo.com>    Raja Rao Sindhe wrote:  E >    We exported files from clearcase to OpenVMS, mounted files using ( > NFS mount ($tcpip mount dnfs2:[ab.src]2 > /path="/view/abc_view/ab/src"/host="<hostname>",H > able to see files from Unix system on OpenVMS system, but problem I amG > facing is OpenVMS adding '$' sign beyond and after Capital letters in G > the filenames, I tried on ODS2 disk, How to avoid '$ in the filename.   + The dollar sign signifies case change.  See H http://h71000.www7.hp.com/doc/732final/6526/6526pro_048.html#index_x_963 --  
 Brian Tillman    ------------------------------    Date: 22 Mar 2004 17:55:01 +0100' From: huber@mppmu.mpg.de (Joseph Huber)  Subject: Re: NFS mount on VMS + Message-ID: <tFupsBKtP$GX@vms.mppmu.mpg.de>   m In article <de8f9c73.0403220115.3dd1169a@posting.google.com>, rajarao1106@yahoo.com (Raja Rao Sindhe) writes: E >    We exported files from clearcase to OpenVMS, mounted files using ( > NFS mount ($tcpip mount dnfs2:[ab.src]2 > /path="/view/abc_view/ab/src"/host="<hostname>",H > able to see files from Unix system on OpenVMS system, but problem I amG > facing is OpenVMS adding '$' sign beyond and after Capital letters in G > the filenames, I tried on ODS2 disk, How to avoid '$ in the filename. ( > later I tried on ODS5 also by setting " > $ set proc /parse_style=extended! > $ set proc /case_look=sensitive  > % > Do we need set some more variables. + > Any input on this, is greatly appreciated   ? The mounted disk itself must be ODS5 to avoid this translation: ,   put /struct=5  in the TCPIP MOUNT command.A (if the software is HP TCPIP services, don't know for other TCPIP 	  stacks).       --  >    Joseph "Sepp" Huber, Muenchen   http://www.huber-joseph.de/   ------------------------------  % Date: Mon, 22 Mar 2004 10:44:50 -0500 4 From: David R. Beatty <QWDavidER.TYBeattyUI@sas.com> Subject: Re: NFS mount on VMS 8 Message-ID: <ic2u50l1v1269pnfo0shf36pfs95hgrqgg@4ax.com>  F On 22 Mar 2004 01:15:41 -0800, rajarao1106@yahoo.com (Raja Rao Sindhe) wrote:   >Hi, > D >   We exported files from clearcase to OpenVMS, mounted files using' >NFS mount ($tcpip mount dnfs2:[ab.src] 1 >/path="/view/abc_view/ab/src"/host="<hostname>", G >able to see files from Unix system on OpenVMS system, but problem I am F >facing is OpenVMS adding '$' sign beyond and after Capital letters inF >the filenames, I tried on ODS2 disk, How to avoid '$ in the filename.' >later I tried on ODS5 also by setting  ! >$ set proc /parse_style=extended   >$ set proc /case_look=sensitive > $ >Do we need set some more variables.* >Any input on this, is greatly appreciated >  >Waiting for reply,  >Raja rao sindhe  F   What version of TCP/IP Services are you running?  ODS-5 is supportedF in V5.3 and V5.4 by using the /Stucture=5 qualifier on the TCPIP MOUNT command.   David R. Beatty    ------------------------------  % Date: Mon, 22 Mar 2004 12:27:03 -0500 4 From: David R. Beatty <QWDavidER.TYBeattyUI@sas.com> Subject: Re: NFS mount on VMS 8 Message-ID: <he8u50p3pmnjnbllo2b9gsen9ibv7r6070@4ax.com>  3 On Mon, 22 Mar 2004 10:44:50 -0500, David R. Beatty % <QWDavidER.TYBeattyUI@sas.com> wrote:    > G >On 22 Mar 2004 01:15:41 -0800, rajarao1106@yahoo.com (Raja Rao Sindhe)  >wrote:  >  >>Hi,  >>E >>   We exported files from clearcase to OpenVMS, mounted files using ( >>NFS mount ($tcpip mount dnfs2:[ab.src]2 >>/path="/view/abc_view/ab/src"/host="<hostname>",H >>able to see files from Unix system on OpenVMS system, but problem I amG >>facing is OpenVMS adding '$' sign beyond and after Capital letters in G >>the filenames, I tried on ODS2 disk, How to avoid '$ in the filename. ( >>later I tried on ODS5 also by setting " >>$ set proc /parse_style=extended! >>$ set proc /case_look=sensitive  >>% >>Do we need set some more variables. + >>Any input on this, is greatly appreciated  >> >>Waiting for reply, >>Raja rao sindhe  > G >  What version of TCP/IP Services are you running?  ODS-5 is supported G >in V5.3 and V5.4 by using the /Stucture=5 qualifier on the TCPIP MOUNT 	 >command.  >  >David R. Beatty  = ... make that /Structure=5 ... preview is my friend ... David    ------------------------------  % Date: Mon, 22 Mar 2004 09:47:19 -0500 * From: "Syltrem" <syltremzulu@videotron.ca>/ Subject: Re: OpenVMS Cluster  Oracle Dataguard? 3 Message-ID: <1bE7c.1512$Xy3.5514@tor-nn1.netcom.ca>   K I don`t use dataguard because there is a cost associated to it, but in case L of a crash (or scheduled shutdown), I just restart the db on the other node.H Downtime < 1 minute (or whatever it takes to bring the instance down and restart it).H Users only have to reconnect, nothing to change in the config whatsoever (using IP aliases).   D Even with dataguard, users would have to reconnect unless the app is failover-aware.    HTH    --   Syltrem    OpenVMS 7.3-1 + Oracle 8.1.7.4H http://pages.infinit.net/syltrem (OpenVMS related web site, en franais)% ---zulu is not in my email address--- 7 "Hal Kuff" <kuff@tessco.com> a crit dans le message de ' news:c3ka9s$p06@library1.airnews.net...  >  >  >  > B >     We would like to hear from anyone using Oracle on an OpenVMS
 cluster...F > How does the dataguad feature wotk in a cluster to provide failover? >  >  >  >  >    ------------------------------  % Date: Mon, 22 Mar 2004 09:48:18 -0500 * From: "Brian Tillman" <tillmabg@yahoo.com>P Subject: RE: OpenVMS virus proof while other disks get zapped by latest viruses!' Message-ID: <012FF57E.C22236@yahoo.com>    JF Mezei wrote:   ? > Palm is to PDAs what SUN is to servers. It rode the .com boom  > and everybody ; > got one because it was so trendy. But now, they aren't so  > healthy because > > people don't buy anymore either because they doN't need one, > or they already  > have one.   H The iPAQ has become the most popular PDA, and people are buying a lot of" them.  But, thanks for playing.-- 
 Brian Tillman    ------------------------------    Date: 22 Mar 2004 06:43:11 -0800- From: phillip_thayer@hotmail.com (PhilThayer) ! Subject: Re: Quorum Disk Question = Message-ID: <5ee1d1b7.0403220643.70a75ee7@posting.google.com>    > !> > ! J > !We have over 60 2-node VMS clusters using the system disk as the quorumL > !disk.  Management did not want to waste space/spindles for a quorum disk.D > !Now that we are using EVA's for new clusters, I can make an extraA > !pagefile/quorum disk as an extra LUN without using unnecessary  > !spindles/space. > O > I was told that "best practice" suggested not to use the system disk, (or any P > other disk) but to use a separate one.  I don't know the precise reason(s) forL > the suggestion, but I suppose it might have to do with accessing the disk;H > imagine the havoc that might be visited on a cluster deleteing 4+GB of@ > highly-fragmented files from a disk which contains QUORUM.DAT. > N > I haven't had the pleasure of using EVA for more than a year now, but I feltF > great creating a quorum disk of minimal size - no more wasted space! >  >   E It makes sense to use the system disk as the quorum disk.  After all, A if the system disk disappears you have no system so no need for a C quorum disk on that system.  If each system uses the system disk as C the quorum disk, and the expected votes is set to 1 then you have a D cluster that will not hang when the other nodes in the system crash.   PT   ------------------------------  % Date: Mon, 22 Mar 2004 09:52:11 -0500 * From: "Brian Tillman" <tillmabg@yahoo.com>! Subject: RE: Quorum Disk Question ' Message-ID: <012FF5CB.C22236@yahoo.com>    Bradford J. Hamilton wrote:   > > It must be mounted explicitly - the only cluster I currently > have managerG > access to mounts the disk in SYLOGICALS.  ISTR the mount can exist in @ > SYPAGSWPFILES, as well, but my recollection could be faulty...  J I have my page ans swap files on my quorum disk, they being the only filesK on it.  Naturally, it gets mounted in SYPAGSWPFILES.COM.  I have most of my L administrative files on a disk other than SYS$SYSDEVICE, so I mount that one in SYLOGICALS.COM. --  
 Brian Tillman    ------------------------------  % Date: Mon, 22 Mar 2004 08:55:17 +0000 * From: Nic Clews <sendspamhere@[127.0.0.1]>% Subject: Re: Securing files under VMS ' Message-ID: <c3m9nf$j95$1@lore.csc.com>    John Smith wrote:  > L > I'd be interested in hearing about methods other than ACL's which are used, > in VMS shops for securing access to files. > J > There are many applications where really secure access to information isI > required...military, intelligence, medical, financial, etc..... and I'm F > looking for thoughts on approaches to security beyond ACL's in a VMS > environment.  H The practice of keeping systems in protected (locked) screened rooms andC private networks, if a network at all. Data transfer very carefully F regulated, a distinct lack of interchangeable media devices, a need toH know / need to access policy, systems and hard drives in physical lockedC enclosures, and we're not talking that generic rack type enclosure.    --  ? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciences  nclews at csc dot com    ------------------------------  # Date: Mon, 22 Mar 2004 14:19:00 GMT # From: "John Smith" <a@nonymous.com> % Subject: Re: Securing files under VMS G Message-ID: <oBC7c.514$uz5.507@twister01.bloor.is.net.cable.rogers.com>   7 "Nic Clews" <sendspamhere@[127.0.0.1]> wrote in message ! news:c3m9nf$j95$1@lore.csc.com...  > John Smith wrote:  > > I > > I'd be interested in hearing about methods other than ACL's which are  used. > > in VMS shops for securing access to files. > > L > > There are many applications where really secure access to information isK > > required...military, intelligence, medical, financial, etc..... and I'm H > > looking for thoughts on approaches to security beyond ACL's in a VMS > > environment. > J > The practice of keeping systems in protected (locked) screened rooms andE > private networks, if a network at all. Data transfer very carefully H > regulated, a distinct lack of interchangeable media devices, a need toJ > know / need to access policy, systems and hard drives in physical lockedE > enclosures, and we're not talking that generic rack type enclosure.    Undoubtedly true.   G Unfortunately I didn't make my first post clear enough. What I'm really L looking at are mandatory access controls and audit trails on file access and modifications.  J For example, if files are kept in the native file system under VMS we haveJ ACL's and audit alarms which can be enabled. We could, from an applicationL standpoint, create a 'service' which manages the CRUD of files and could logD these actions into a database (UID, PID, timestamps, ip address, MACK address, etc...), and under VMS we can even capture the file version number H so actioned. We have dba's who own the database in a separate group fromL system managers and in some senses they can keep watch on one another (trustH but verify). But system managers do have pretty much complete control of what's outside of a database.   K My biggest problem with the application I'm thinking about this for is that K the nature of the data is not 'operational' but more referential in nature. K With operational data there are usually many processes and eyeballs looking E at it in a short period of time that anything obviously wrong will be I spotted almost immediately, whereas with the referential data/files it is K imperative that the data not be altered through time whilst the data is not H archived off-line (not for at least 10 years). I can't rely on long-termD people knowledge of what to look for as people quit and take perhapsI undocumented 'processes' with them when they leave, os the 'controls' and 0 audits have to be automated and post exceptions.  H I suppose that once created/received, the data could be stored on DVD orD other write-once media, or even hosted in a different facility underH different 'ownership'. The environment is not one where loss/altering ofK data is life or security threatening, but rather one where loss of trust in A the accuracy of the information would be extremely harmful to the  organization tending the data.  E So in a broad sense I guess what I'm hoping to hear are other general J strategies for securing files and data in a production environment for the
 long term.   John   ------------------------------    Date: 22 Mar 2004 11:39:15 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) % Subject: Re: Securing files under VMS 3 Message-ID: <wB8jYSrWMNZE@eisner.encompasserve.org>   m In article <oBC7c.514$uz5.507@twister01.bloor.is.net.cable.rogers.com>, "John Smith" <a@nonymous.com> writes:  >  > I > Unfortunately I didn't make my first post clear enough. What I'm really N > looking at are mandatory access controls and audit trails on file access and > modifications.  7    "mandatory access controls" sounds like SEVMS to me.    ------------------------------    Date: 22 Mar 2004 08:05:46 -0800' From: ultrajoe@spamcop.net (Joe Sewell) ( Subject: Re: Server hang - compute bound= Message-ID: <a55b951e.0403220805.5b16b4d0@posting.google.com>   r martinkirby12@yahoo.co.uk (Martin Kirby) wrote in message news:<224291b.0403182359.4461f8ab@posting.google.com>... > Joe, > B > Do you have the graphics update kit installed? It is pretty well > necessary. > C > We don't have any reports of that. If you have a service contract H > please raise a call, if not please give me some details on your system@ > - graphics cards, cpus etc. and see if you can gather some cpu4 > addresses and stack traces for the looping server. >  > Martin Kirby  F Graphics update 200 is installed on one machine.  I don't know if thatB machine has exhibited the problem or not; it's rather difficult to
 reproduce.  C I know some of the systems exhibiting the problem are using 4D20's; D I'm not sure if any of them are using PowerStorm 350's.  The systemsF themselves are usually XP1000's, I believe, although there may be some% Alpha 500-series machines in the mix.   D I haven't been able to get any cpu addresses or stack traces for theB looping server, again due to the difficulty to reproduce.  I'm notC sure, in all honesty, how to get a stack trace on a running process , (unless that's in ANALYZE/SYSTEM somewhere).   ------------------------------  % Date: Mon, 22 Mar 2004 14:37:37 +0100 3 From: Michael Unger <spam.to.unger@spamgourmet.com> / Subject: Re: Someone tried to crack my Alpha... : Message-ID: <c3msja$29q1o8$1@ID-152801.news.uni-berlin.de>  , On 2004-03-21 19:47, "Didier Morandi" wrote:   > Didier Morandi wrote:  > 
 >> Hi all, >>  & >> Finally, some life in my system :-)A >> I found today a file in my default anonymous directory, named  I >> BIGDICT.ZIP size 8106 blocks. Unzipped, it gave a CRACKLIB.TXT, 29820  
 >> blocks. >  > Here is the author's log: 6 > (the TAGGED dir is empty, and I *did not* create it) > R > 22-FEB-2004 14:13:37.88 User:anonymous logged in ident:anonymous@on.the.net from# >   Host:pD9EE15D5.dip.t-dialin.net #           ^^^^^^^^ ^^^^^^^^^^^^^^^^   $ - IP address in hexadecimal notation - "dip" = "dial-in pool"* - "t-dialin.net" is equal to "t-online.de"   >  > [...]   D Having an (accurate?) time stamp and the originator's IP address you< could ask the abuse desk of t-online.de for further support.   Michael    --  ; Real names enhance the probability of getting real answers. @ Please do *not* send "Security Patch Notifications" or "SecurityA Updates"; this system isn't running a Micro$oft operating system. 5 My e-mail account at DECUS Munich is no longer valid.    ------------------------------  % Date: Mon, 22 Mar 2004 08:22:21 -0700 + From: "Barry Treahy, Jr." <Treahy@MMaz.com> / Subject: Re: Someone tried to crack my Alpha... ' Message-ID: <405F04AD.9010000@MMaz.com>    Michael Unger wrote:  - >On 2004-03-21 19:47, "Didier Morandi" wrote:  >  >    >  >>Didier Morandi wrote:  >> >>     >>
 >>>Hi all, >>> & >>>Finally, some life in my system :-)A >>>I found today a file in my default anonymous directory, named  I >>>BIGDICT.ZIP size 8106 blocks. Unzipped, it gave a CRACKLIB.TXT, 29820  
 >>>blocks.	 >>>        >>>  >>Here is the author's log: 6 >>(the TAGGED dir is empty, and I *did not* create it) >>R >>22-FEB-2004 14:13:37.88 User:anonymous logged in ident:anonymous@on.the.net from# >>  Host:pD9EE15D5.dip.t-dialin.net  >>     >>$ >          ^^^^^^^^ ^^^^^^^^^^^^^^^^ > % >- IP address in hexadecimal notation  >- "dip" = "dial-in pool" + >- "t-dialin.net" is equal to "t-online.de"  >  >    >  >>[...]  >>     >> > E >Having an (accurate?) time stamp and the originator's IP address you = >could ask the abuse desk of t-online.de for further support.  >  >    > I I wouldn't bother, they are not receptive or responsive to SPAM or ABUSE  
 complaints...      Barry    --    > Barry Treahy, Jr                       E-mail: Treahy@MMaz.com> Midwest Microwave                          Phone: 480/314-1320> Vice President & CIO                         FAX: 480/661-7028                            ------------------------------  % Date: Mon, 22 Mar 2004 07:00:45 -0800 # From: "Tom Linden" <tom@kednos.com>  Subject: Spam form kerry9 Message-ID: <NDEMLKKEBOIFBMJLCECIEELJCPAA.tom@kednos.com>   8 Yes I know that this didn't come from him, but I have of6 late seen a lot of messages whose subject is typically3 Re:  something.  Anybody have experience with this?  The IP actually belongs to RIPE    Tom   '     #7          22-MAR-2004 06:41:37.15  MAIL  From:   SMTP%"kerry.main@hp.com" To:     SMTP%"tom@kednos.com"  CC:  Subj:   Re: Re: corrected     Return-Path: <kerry.main@hp.com>L Received: from kednos.com (195.193.79.62) by FREJA.KEDNOS.COM (MX V5.3 An1f)J           with ESMTP for <tom@kednos.com>; Mon, 22 Mar 2004 06:41:35 -0800 From: kerry.main@hp.com  To: tom@kednos.com Subject: Re: Re: corrected% Date: Mon, 22 Mar 2004 15:40:12 +0100  MIME-Version: 1.0  Content-Type: multipart/mixed;G               boundary="----=_NextPart_000_0016----=_NextPart_000_0016" 
 X-Priority: 3  X-MSMail-Priority: Normal   , This is a multi-part message in MIME format.  0 ------=_NextPart_000_0016----=_NextPart_000_0016 Content-Type: text/plain;          charset="Windows-1252" Content-Transfer-Encoding: 7bit    Please read the document.     0 ------=_NextPart_000_0016----=_NextPart_000_0016' Content-Type: application/octet-stream;          name="data.pif" ! Content-Transfer-Encoding: base64   Content-Disposition: attachment;         filename="data.pif"      --- & Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).@ Version: 6.0.593 / Virus Database: 376 - Release Date: 2/20/2004   ------------------------------  + Date: Mon, 22 Mar 2004 15:56:12 +0000 (UTC) , From: lewis@PROBE.mitre.org (Keith A. Lewis) Subject: Re: Spam form kerry. Message-ID: <c3n2as$9m9$1@newslocal.mitre.org>   "Tom Linden" <tom@kednos.com> writes in article <NDEMLKKEBOIFBMJLCECIEELJCPAA.tom@kednos.com> dated Mon, 22 Mar 2004 07:00:45 -0800:9 >Yes I know that this didn't come from him, but I have of 7 >late seen a lot of messages whose subject is typically 4 >Re:  something.  Anybody have experience with this?  >The IP actually belongs to RIPE  1 >------=_NextPart_000_0016----=_NextPart_000_0016 ( >Content-Type: application/octet-stream; >        name="data.pif"" >Content-Transfer-Encoding: base64! >Content-Disposition: attachment;  >        filename="data.pif" >   J When you get a .pif, .scr, .exe, .bat, or other type of executable file inL e-mail, it is generally not spam but a virus.  In this case it probably cameJ from somebody who had both your e-mail address and Kerry's in MS-Outlook. / You can trace the IP and try to figure out who.   J The other day my wife said Norton Antivirus prevented her from downloadingL an attachment onto my PC.  Good catch on Norton's part but since my attemptsE at education have obviously failed, I took the preventative action of K leaving Warcraft 3 running on it, to discourage use of the toy computer for F e-mail.  Now she's using her own profile on Mozilla in her own VMS CDEK workspace instead.  And Explorer on MacOS.  MacOS has the cool feature that J even though it's "logged in" to a privileged account you have to enter the) password when you make sensitive changes.   0 --Keith Lewis              klewis {at} mitre.org> The above may not (yet) represent the opinions of my employer.   ------------------------------  % Date: Mon, 22 Mar 2004 15:45:46 -0000 * From: "Richard Brodie" <R.Brodie@rl.ac.uk> Subject: Re: Spam form kerry, Message-ID: <c3n1nd$12k4@newton.cc.rl.ac.uk>  b "Tom Linden" <tom@kednos.com> wrote in message news:NDEMLKKEBOIFBMJLCECIEELJCPAA.tom@kednos.com...  : > Yes I know that this didn't come from him, but I have of8 > late seen a lot of messages whose subject is typically5 > Re:  something.  Anybody have experience with this?   : It's just a virus that forges From addresses; probably one of the Netsky family.    ------------------------------    Date: 22 Mar 2004 07:54:56 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) - Subject: Re: VAX clock cycles per instruction 3 Message-ID: <mMPDdmV0wfyB@eisner.encompasserve.org>   ] In article <4bc97a76.0403211939.36ace829@posting.google.com>, tollan83@yahoo.com (et) writes: 
 > hey all, > D > i am working on trying to calculate the number of clock cycles for> > instructions in the IS for the VAX, and i wanted to run someH > information by everyone to see if i am going about this the right way.; > the point of this project is to learn, so here it goes...  >   ?    Clock cycles per instruction is a complex question on a VAX. D    Clock cylcles will vary for a given instruction depending on suchG    things as cache hits.  You have to answer questions like whether the C    overhead in resolving faults like page faults should be counted.   G    Clock cycles will also vary due to implementation.  Some VAXen don't E    implement all the instructions in hardware, there's quite a bit of H    overhead to "execute" those which aren't.  Some VAXen will impelementG    instructions differently on the same model depending on options like F    floating point accelerators (which do a lot more than just floating
    point).  F    Some VAX instructions are so complex that the clock cycles could beA    data dependent.  Some of the more complex instructions are not =    atomic, they can be insterrupted mid-instruction (look for F    instructions like MOVCx which store intermediate results in general    purpose registers.)  B    Some instructions interlock, which means they could vary on SMPH    VAXen depending on whether another CPU is holding the inter-processor    lock.  C    And some instructions depend on instruction mix on some models.  G    Memory write from one instruction might complete while the following ?    register to register instruction executes, but if the second E    instruction did memory access it would have to hold off a cycle or     two.   F    As I think you've already seen, the number of cycles will depend on0    the complexity of addressing of the operands.  C    When actually measuring the performance of an 11/780 years ago I G    found it easy to see the effect of many of these issues.  I was able G    to get variation from 0.75 to 2.1 MIPS on an 11/780 in a single test (    all using fairly simple instructions.  L    And yes, there are two-byte opcodes.  There is no conceptual upper limit J    to the number of bytes in an opcode, but not all two-byte opcodes were G    ever used, and none larger were used by DEC.  On some systems it was E    fairly easy for customers to add opcodes, but I assume you're only     interested in what DEC did.  G    As far as the size of the instruction, IIRC they varied from 1 to 12 I    bytes, but that's only if you don't count the displacement table for a G    CASEL.  If you count that as an operand then the upper limit is 4GB.           ------------------------------  % Date: Mon, 22 Mar 2004 15:45:06 +0000 ( From: John Travell <john@travell.uk.net>- Subject: Re: VAX clock cycles per instruction : Message-ID: <c3n1m3$29j7g9$1@ID-120847.news.uni-berlin.de>   Bob Koehler wrote:_ > In article <4bc97a76.0403211939.36ace829@posting.google.com>, tollan83@yahoo.com (et) writes:  > 
 >>hey all, >>D >>i am working on trying to calculate the number of clock cycles for> >>instructions in the IS for the VAX, and i wanted to run someH >>information by everyone to see if i am going about this the right way.; >>the point of this project is to learn, so here it goes...  >> >  > A >    Clock cycles per instruction is a complex question on a VAX. F >    Clock cylcles will vary for a given instruction depending on suchI >    things as cache hits.  You have to answer questions like whether the E >    overhead in resolving faults like page faults should be counted.  > I >    Clock cycles will also vary due to implementation.  Some VAXen don't G >    implement all the instructions in hardware, there's quite a bit of J >    overhead to "execute" those which aren't.  Some VAXen will impelementI >    instructions differently on the same model depending on options like H >    floating point accelerators (which do a lot more than just floating >    point). > H >    Some VAX instructions are so complex that the clock cycles could beC >    data dependent.  Some of the more complex instructions are not ? >    atomic, they can be insterrupted mid-instruction (look for H >    instructions like MOVCx which store intermediate results in general >    purpose registers.) > D >    Some instructions interlock, which means they could vary on SMPJ >    VAXen depending on whether another CPU is holding the inter-processor
 >    lock. > E >    And some instructions depend on instruction mix on some models.  I >    Memory write from one instruction might complete while the following A >    register to register instruction executes, but if the second G >    instruction did memory access it would have to hold off a cycle or 	 >    two.V > H >    As I think you've already seen, the number of cycles will depend on2 >    the complexity of addressing of the operands. > E >    When actually measuring the performance of an 11/780 years ago IoI >    found it easy to see the effect of many of these issues.  I was ableSI >    to get variation from 0.75 to 2.1 MIPS on an 11/780 in a single teste* >    all using fairly simple instructions. > N >    And yes, there are two-byte opcodes.  There is no conceptual upper limit L >    to the number of bytes in an opcode, but not all two-byte opcodes were I >    ever used, and none larger were used by DEC.  On some systems it waskG >    fairly easy for customers to add opcodes, but I assume you're only   >    interested in what DEC did. > I >    As far as the size of the instruction, IIRC they varied from 1 to 12nK >    bytes, but that's only if you don't count the displacement table for amI >    CASEL.  If you count that as an operand then the upper limit is 4GB.  >  >    s^ Actually, a POLYG instruction can occupy 56 bytes, counted from one OPcode to the next Opcode.G You also have to remember that VAX code carries a LOT of I-stream data. _ e.g. opcode, operand-specifier-1, up-to-4-bytes-Istream-data, op-spec-2, 4-more-Istream-bytes, s
 op-spec-3.  Y such as: ADDL3 @#1000[r1],@#2000[r2],@r3 is (iirc) 12 bytes, 8 of which are Istream data.      -- s John Travell* Independent VMS and Alpha Support analyst. john- at - jomatech - dot - comh http://www.jomatech.com/   ------------------------------    Date: 22 Mar 2004 11:38:18 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)M- Subject: Re: VAX clock cycles per instructione3 Message-ID: <A4kOpbjC69Tq@eisner.encompasserve.org>D  e In article <c3n1m3$29j7g9$1@ID-120847.news.uni-berlin.de>, John Travell <john@travell.uk.net> writes:. > Bob Koehler wrote:  ` > Actually, a POLYG instruction can occupy 56 bytes, counted from one OPcode to the next Opcode.I > You also have to remember that VAX code carries a LOT of I-stream data.oa > e.g. opcode, operand-specifier-1, up-to-4-bytes-Istream-data, op-spec-2, 4-more-Istream-bytes,   > op-spec-3. >   G    Exactly why I mention the CASEL displacement table.  The next opcodeh    could be far, far away.   ------------------------------  % Date: Mon, 22 Mar 2004 08:30:39 +0100c* From: Paul Sture <nospam@sture.homeip.net>3 Subject: Re: Wants to run autogen evertime it bootsi: Message-ID: <c3m4mv$29kk0l$1@ID-132135.news.uni-berlin.de>   G Henry wrote: > All: > C > I am tring to get my DPW600a set up on VMS and while I worked forsC > Digital for 14+ years I was NOT focused on VMS and I had a lot ofsD > people around me that I could ask for help when I had a question.  > F > My DPW600a has had Free BSD and Windows NT on it and now it is VMS'sG > turn (Damn that Bob Palmer :-)) It has been probally 10 years and VMS.D > 5.1 since I last played with VMS and while I am having fun with it > again I NEED some help now...  > E > I have succesfully installed 7.2-2 (I did this verison on purpose I @ > need it for a piece of software that I ultimetly want to load) > G > I had the system working well but I just rebooted it and now eveytimeiE > it boots it askes to run autogen.  It is complaining that there are E > not enough GBLSectons to start DECwindows (it says that it only hasR1 > 250 when it needs something like 280 to start).   C I had this problem, and solved it by putting the following line in A MODPARAMS.DAT.   MIN_GBLSECTIONS=1000   <snip>   >  I also get some kindeG > of OPCOM message (it srcolls off the console and I cannot capture it) @ > It sayting something about "Too few servers detected from node > Local:.AS09X DTSS" g    F If it doesn't already exist, create SYS$MANAGER:NET$EVENT_LOCAL.NCL by) copying it from NET$EVENT_LOCAL.TEMPLATE.n   Now add the following lines:   !w ! block dtss too few servers !n8 block event dispatcher outbound stream * global filter -0          ((node,dtss), Too Few Servers Detected)  H You may also want to add the following if you see the relevant messages:  C BLOCK EVENT DISPATCHER OUTBOUND STREAM local_stream GLOBAL FILTER -a1          ((NODE,DTSS), Synchronization Completed)l  = To find out more about DTSS, download and read the VMS FAQ at@1 http://h71000.www7.hp.com/wizard/openvms_faq.htmln   ------------------------------  % Date: Mon, 22 Mar 2004 08:55:23 +0100 - From: "Martin Vorlaender" <mv@pdv-systeme.de>c3 Subject: Re: Wants to run autogen evertime it bootse9 Message-ID: <c3m65c$29hre1$1@ID-56200.news.uni-berlin.de>h   Paul Sture wrote:  > G Henry wrote: >>  I also get some kindH >> of OPCOM message (it srcolls off the console and I cannot capture it)A >> It sayting something about "Too few servers detected from node  >> Local:.AS09X DTSS"o > H > If it doesn't already exist, create SYS$MANAGER:NET$EVENT_LOCAL.NCL by+ > copying it from NET$EVENT_LOCAL.TEMPLATE.o ...[blocking DTSS messages]...  D Or, if you don't want to sync time using DTSS, disable it altogether by  $   $ define/system net$disable_dtss 1   in SYS$MANAGER:SYLOGICALS.COM.   cu,.   Martin -- .@   OpenVMS:                | Martin Vorlaender  |  OpenVMS rules!3    The operating system   | work: mv@pdv-systeme.dewF    God runs the           |   http://www.pdv-systeme.de/users/martinv/:    earth simulation on.   | home: martin@radiogaga.harz.de   ------------------------------  % Date: Mon, 22 Mar 2004 09:50:15 -0000 * From: "Richard Brodie" <R.Brodie@rl.ac.uk>3 Subject: Re: Wants to run autogen evertime it boots + Message-ID: <c3mcsp$mic@newton.cc.rl.ac.uk>@  3 "G Henry" <newsgroup@gnshenry.com> wrote in message 2 news:7des50pf9nkmsd38io72qqjarqe5dlmfq7@4ax.com...  G > TO solve the autogen problem I have tried slect YES to run autogen itnE > just asks me to run it again.  I have slected NO and telnetted intoaE > the system and gone into modparams and set MIN_GBLSECTION = 600 andoB > then run autogen (autogen genparams reboot) and it did not work.  8 AUTOGEN is picky about exact names of SYSGEN parameters.? If you have MIN_GBLSECTION = 600 (without the S) it won't work.   J > I have gone into sysgen and set min and default values of GBLSECTIONS to > 600s  H I don't know how you could go about changing the defaults. AFAIK you can only change the actual values.  F As John says, skimming the AUTOGEN report is probably the thing to do.   ------------------------------  % Date: Mon, 22 Mar 2004 08:20:38 +0000c From: Roy Omond <Roy@Omond.net>-A Subject: Re: why isn't there room on this disk for the swap file?W4 Message-ID: <c3m7kn$3g4$1$8302bc10@news.demon.co.uk>   John Laird wrote:nK > On Sun, 21 Mar 2004 22:31:05 +0000 (UTC), helbig@astro.multiCLOTHESvax.dem3 > (Phillip Helbig---remove CLOTHES to reply) wrote:  >  > J >>In article <c3kf40$84f$1@pcls4.std.com>, moroney@world.std.spaamtrap.com >>(Michael Moroney) writes:  >> >>G >>>The primitive boot drivers can't deal with extension headers, so thenK >>>drive can't be so fragmented that all fragments are too small to get the_K >>>allocated space in a single header.  Defragment the drive (easiest is to F >>>do a BACKUP/IMAGE and restore the backup /IMAGE if you don't have a& >>>defragmenter) or use another drive. >>G >>I started out with a freshly initialized drive, but maybe even one or.H >>two screw-ups was enough to through it out of kilter.  Since I plan to= >>use the drive only for secondary swap and page files, I can * >>re-initialize it and give it another go. >  > N > You may want to $ INIT/INDEX=BEGIN then, and get the largest contiguous areaH > possible.  I never let autogen create or modify page/swap files - muchM > easier to use sysgen interactively.  To be honest, I wouldn't bother with aDJ > swap file at all, unless you are forced to have at least one (I forget).N > The days when memory was limited and/or expensive and there were performanceM > gains to be had from encouraging swapping over paging were a very long timep > ago.  F Agreed re using SYSGEN interactively - I *never* let Autogen mess with page/swap files.  G Minor nitpick:  instead of placing the Index file at the *beginning* in D order to get the largest contiguous area, place it at the *end* i.e.   	$ init/index=ende  E This way, the outer tracks (and hence the "fastest") will be used for1@ the real work, paging.  The index file will pretty much never be	 accessed.:  , Only when you're in search of perfection :-)  	 Roy Omondf Blue Bubble Ltd.   ------------------------------  + Date: Mon, 22 Mar 2004 08:06:15 +0000 (UTC):P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)A Subject: Re: why isn't there room on this disk for the swap file?a$ Message-ID: <c3m6pm$oqr$1@online.de>  2 In article <405E0CE5.B96B4B05@istop.com>, JF Mezei$ <jfmezei.spamnot@istop.com> writes:   P > As I recall, autogen only deals with system drive page/swap files. If you haveP > secondary ones on a spearate disk (which is what I have), then you need to useK > the PAGEFILE = 0 in modparams otherwise autogen will think that the small D > pagefile on your system drive is inadequate and needs to be grown.  H AUTOGEN will certainly change the sizes of secondary files, or at least  create them.   ------------------------------  + Date: Mon, 22 Mar 2004 08:18:30 +0000 (UTC)nP From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)A Subject: Re: why isn't there room on this disk for the swap file?D$ Message-ID: <c3m7gm$oqr$2@online.de>  C In article <gm9s5054moq7sp3nijhno0v4st1e7tra6i@4ax.com>, John Lairda% <nospam@laird-towers.org.uk> writes: -  I > >> The primitive boot drivers can't deal with extension headers, so theMM > >> drive can't be so fragmented that all fragments are too small to get the M > >> allocated space in a single header.  Defragment the drive (easiest is tonH > >> do a BACKUP/IMAGE and restore the backup /IMAGE if you don't have a( > >> defragmenter) or use another drive. > >mH > >I started out with a freshly initialized drive, but maybe even one orI > >two screw-ups was enough to through it out of kilter.  Since I plan too> > >use the drive only for secondary swap and page files, I can+ > >re-initialize it and give it another go._ > N > You may want to $ INIT/INDEX=BEGIN then, and get the largest contiguous area
 > possible.  n   That makes sense.r  = > I never let autogen create or modify page/swap files - much M > easier to use sysgen interactively.  To be honest, I wouldn't bother with aiJ > swap file at all, unless you are forced to have at least one (I forget).N > The days when memory was limited and/or expensive and there were performanceM > gains to be had from encouraging swapping over paging were a very long time7 > ago.  F The idea is to use this drive, which is a solid-state-disk, for large I page files for a memory-starved ALPHA (either a 255/233 with 64 MB, or a  H 3000/300 LX with 48 MB, which is in the cluster now until I get the 255 G fixed).  It IS much faster than a normal disk.  (When I have some more o: money in the bank, I'll buy some memory for the machines!)  C Back to AUTOGEN: one can set the size to 0 to prevent AUTOGEN from oG messing with the files.  What about hard-coding the value to something tE other than 0?  Presumably, AUTOGEN will do its calculations, but not aI actually change anything since the value is hard-coded; the best of both p worlds?n   ------------------------------  % Date: Mon, 22 Mar 2004 10:03:19 +0100y* From: Paul Sture <nospam@sture.homeip.net>A Subject: Re: why isn't there room on this disk for the swap file?n: Message-ID: <c3ma4n$21lnbs$1@ID-132135.news.uni-berlin.de>  / Phillip Helbig---remove CLOTHES to reply wrote:oO >>>You may want to $ INIT/INDEX=BEGIN then, and get the largest contiguous area  >>>possible.   >> >>That makes sense.e >  > $ > I re-initialized the disk as aboveI > (INIT/SYSTEM/INDEX=BEGIN/HEADERS=1000) and started over.  Same problem.h> > There must be some magic to the 99999 free blocks after all. >   F Could it be a "feature" of your solid state disk? Do you seem similar B behaviour when you give the pagefile the entire disk, for example?   ------------------------------  + Date: Mon, 22 Mar 2004 08:50:32 +0000 (UTC)yP From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)A Subject: Re: why isn't there room on this disk for the swap file?f$ Message-ID: <c3m9co$qo5$1@online.de>  P > > You may want to $ INIT/INDEX=BEGIN then, and get the largest contiguous area > > possible.  h >  > That makes sense.y  " I re-initialized the disk as aboveG (INIT/SYSTEM/INDEX=BEGIN/HEADERS=1000) and started over.  Same problem.>< There must be some magic to the 99999 free blocks after all.   ------------------------------  + Date: Mon, 22 Mar 2004 08:52:21 +0000 (UTC)oP From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)A Subject: Re: why isn't there room on this disk for the swap file?s$ Message-ID: <c3m9g5$qo5$2@online.de>  > In article <c3m7kn$3g4$1$8302bc10@news.demon.co.uk>, Roy Omond <Roy@Omond.net> writes:   I > Minor nitpick:  instead of placing the Index file at the *beginning* iniF > order to get the largest contiguous area, place it at the *end* i.e. >  > 	$ init/index=end  > G > This way, the outer tracks (and hence the "fastest") will be used for B > the real work, paging.  The index file will pretty much never be > accessed.e  > OK, the size of the contiguous area will be the same, but the  performance might be better.  E However, with a solid-state disk, I don't think there are really any   "outer tracks".  :-)  . > Only when you're in search of perfection :-)   Always.u   ------------------------------  + Date: Mon, 22 Mar 2004 08:58:15 +0000 (UTC)sP From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)A Subject: Re: why isn't there room on this disk for the swap file?-$ Message-ID: <c3m9r6$qo5$3@online.de>  D In article <c3m7gm$oqr$2@online.de>, helbig@astro.multiCLOTHESvax.de3 (Phillip Helbig---remove CLOTHES to reply) writes: 2  P > > You may want to $ INIT/INDEX=BEGIN then, and get the largest contiguous area > > possible.  I   Tried that, same problem.   I Maybe this will throw some light on the issue.  DEFRAGMENT/MONITOR shows A@ three roughly equal blocks of free space located roughly at the I beginning, middle and end of the volume.  That would explain things, but iD a) why is this the case and b) the 400000 block page file which was $ created is obviously not contiguous.   ------------------------------  % Date: Mon, 22 Mar 2004 09:55:55 -0000 * From: "Richard Brodie" <R.Brodie@rl.ac.uk>A Subject: Re: why isn't there room on this disk for the swap file?h+ Message-ID: <c3md7d$os6@newton.cc.rl.ac.uk>   ] "Phillip Helbig---remove CLOTHES to reply" <helbig@astro.multiCLOTHESvax.de> wrote in messagee news:c3m9co$qo5$1@online.de...  $ > I re-initialized the disk as aboveI > (INIT/SYSTEM/INDEX=BEGIN/HEADERS=1000) and started over.  Same problem.e> > There must be some magic to the 99999 free blocks after all.  J If you look at AUTOGEN.COM around the  "insufficient", there do seem to beT a lot of 100000 magic numbers. Analyzing the exact algorithm is left as an exercise.   ------------------------------  % Date: Mon, 22 Mar 2004 12:34:42 +0000t- From: John Laird <nospam@laird-towers.org.uk>uA Subject: Re: why isn't there room on this disk for the swap file? 8 Message-ID: <ofmt50p8vtme2cfsh84mso29bk0qec886g@4ax.com>  I On Mon, 22 Mar 2004 08:58:15 +0000 (UTC), helbig@astro.multiCLOTHESvax.deo1 (Phillip Helbig---remove CLOTHES to reply) wrote:a  E >In article <c3m7gm$oqr$2@online.de>, helbig@astro.multiCLOTHESvax.de-4 >(Phillip Helbig---remove CLOTHES to reply) writes:  >mQ >> > You may want to $ INIT/INDEX=BEGIN then, and get the largest contiguous areaD >> > possible.   >t >Tried that, same problem. >eJ >Maybe this will throw some light on the issue.  DEFRAGMENT/MONITOR shows A >three roughly equal blocks of free space located roughly at the eJ >beginning, middle and end of the volume.  That would explain things, but E >a) why is this the case and b) the 400000 block page file which was p% >created is obviously not contiguous.   I It used to be the case that INIT would allocate most of INDEXF.SYS in thecF middle of the volume, so you would find two large extents either side.  E Per your autogen problems - what happens when you try sysgen's create D command ?  I have not seen this fail to allocate an entire disk to aI pagefile.  Even the restriction of having one header still allows for far = more fragments  than your newly-initialised disk should have.s   -- y@ Never mind what road we're on...  keep your eyes on the carrot!    Mail john rather than nospam...y   ------------------------------  % Date: Mon, 22 Mar 2004 10:09:03 -0400t* From: JF Mezei <jfmezei.spamnot@istop.com>A Subject: Re: why isn't there room on this disk for the swap file?a) Message-ID: <405EF377.E98888C3@istop.com>   N Ok, if you have  RAM DISK, what sort of drive does it fake ? Does it appear asJ a SCSI drive on VMS ?  Or does it use a special driver ? (remember the one, used to build standalone backups on VAX ?) ?  N Also, have you considered tweaking the cluster size on the drive ? Perhaps youN are trying to create a file that may not exceed the total disk block size, butK might exceed some RMS structures which are used by default for this type ofr drive ?s    F Also, aree you using SYSGEN to create the swapfiles ? Or are you stillM attempting to do it through AUTOGEN ? It is possible that AUTOGEN may barf ateI your request before even feeding it to SYSGEN even though SYSGEN might bes perfectly happy.   ------------------------------   End of INFO-VAX 2004.161 ************************