1 INFO-VAX	Tue, 13 May 2003	Volume 2003 : Issue 264       Contents: Advanced Server: file locking ! Re: Advanced Server: file locking ! Re: Advanced Server: file locking ! Re: Advanced Server: file locking ! Re: Advanced Server: file locking ! Re: Advanced Server: file locking ! Re: Advanced Server: file locking ! Re: Advanced Server: file locking , Re: Are Bridgeworks and Multinet compatible? Re: Best practices for VMS Re: Best practices for VMS Re: Best practices for VMS Re: Best practices for VMS Re: Best practices for VMS Re: Best practices for VMS7 BLASTed directory locks, timing windows & endless loops ; Re: BLASTed directory locks, timing windows & endless loops ' Re: creating licenses for my own demos? < Re: How Alpha will save Itanium - must reading for Bill Todd< Re: How Alpha will save Itanium - must reading for Bill Todd How do i analyse TCPDump ?. How many Intel Itaniums have been sold so far?1 how to use smg$create_menu in vax macro language? 5 Re: how to use smg$create_menu in vax macro language? 5 Re: how to use smg$create_menu in vax macro language? 5 Re: how to use smg$create_menu in vax macro language? 5 Re: how to use smg$create_menu in vax macro language?  INIT/SHADOW  Re: INIT/SHADOW  Re: INIT/SHADOW  Re: Java "Unknown host" problem ; Re: MicroVAX and VAX models EOSL list (end of service life) P Re: Not fixed yet but still trying: Problems changing from serial port to DecSer9 OpenVMS Pearl - OpenVMS Article on InTech Brazil Magazine  Re: simh emulator and cluster  Re: Spamfilter for OpenVMS?  Re: Spamfilter for OpenVMS?  Re: Spamfilter for OpenVMS?  Re: Spamfilter for OpenVMS?  Re: Spamfilter for OpenVMS?  Re: Spamfilter for OpenVMS?  Re: Spamfilter for OpenVMS?  Re: Spamfilter for OpenVMS?  Re: Spamfilter for OpenVMS?  Re: Spamfilter for OpenVMS?  Re: Spamfilter for OpenVMS?  RE: Spamfilter for OpenVMS?  Re: Spamfilter for VMS? 3 Re: Sparky is losing the race Andrew ... and badly! 3 Re: Sparky is losing the race Andrew ... and badly! 3 Re: Sparky is losing the race Andrew ... and badly! 3 Re: Sparky is losing the race Andrew ... and badly!   Re: Status of Commserver product) Re: Stopping a que other then a print que 7 Re: What is the schedule for the DII COE certification? 7 Re: What is the schedule for the DII COE certification? 7 Re: What is the schedule for the DII COE certification? 7 Re: What is the schedule for the DII COE certification? 7 Re: What is the schedule for the DII COE certification? ' X-windows: XtAppSetExitFlag() missing ?   F ----------------------------------------------------------------------  % Date: Tue, 13 May 2003 10:25:43 +0300 - From: "Kari Keronen" <kari.keronen@digita.fi> & Subject: Advanced Server: file locking; Message-ID: <761wa.2549$M8.1920442@reader1.news.jippii.net>   L We recently upgradet our VMS cluster from 7.2-1 to 7.3-1 and Advanced ServerL from 6.0C to 7.3A. Now when creating a file through Advanced Server the fileG stays locked much longer than before. This is causing problems with one 1 application that tries to move the created files. K Is there any way to configure Advanced Server so that the file locking time 2 is shorter or must I reconfigure the application ?   -Kari-   ------------------------------  % Date: Tue, 13 May 2003 10:07:57 +0200 : From: Karl Rohwedder <extern.karl.rohwedder@volkswagen.de>* Subject: Re: Advanced Server: file locking/ Message-ID: <b9q916$nis2@doiweb4.volkswagen.de>    Kari Keronen schrieb: N > We recently upgradet our VMS cluster from 7.2-1 to 7.3-1 and Advanced ServerN > from 6.0C to 7.3A. Now when creating a file through Advanced Server the fileI > stays locked much longer than before. This is causing problems with one 3 > application that tries to move the created files. M > Is there any way to configure Advanced Server so that the file locking time 4 > is shorter or must I reconfigure the application ? >  > -Kari- >  > E The Openfile Caching can be configured voa ADMIN/CONFIG -> ADVANCED,  - pls. note that an Eco1 is available for 7.3A.    --    + mit freundlichen Gren | with best regards     Karl Rohwedder		iT-Ingenieurteam$ Ellernbruch 11		D-38112 Braunschweig  E mailto:it-ingteam(at)t-online.de | mailto:rohwedder(at)decus.decus.de - mailto:extern.karl.rohwedder(at)volkswagen.de    ------------------------------  % Date: Tue, 13 May 2003 09:33:12 -0400 * From: "Syltrem" <syltremzulu@videotron.ca>* Subject: Re: Advanced Server: file locking4 Message-ID: <Ww6wa.6813$945.20143@tor-nn1.netcom.ca>   From the release notes:   $ 3.11 Disabling Opportunistic Locking4 Opportunistic locking is enabled by default. Under a2 few circumstances, customers might want to disable: opportunistic locking, such as when they require immediate7 access to files at the OpenVMS system level. Typically, 8 these customers would also disable open file caching. To: disable this feature so that files are not locked, add the5 following two lines to the PWRK$COMMON:PWRK.INI file:    [PLM]  ENABLE_OPLOCKING = 0  L You probably have to restart the pwrk server for this change to take effect.   --   OpenVMS 7.2-1, Oracle 8.1.6.0    Syltrem I http://pages.infinit.net/syltrem (OpenVMS related web site - en franais) 8 To reply to myself directly, remove zulu from my address  H "Kari Keronen" <kari.keronen@digita.fi> a crit dans le message de news:0 761wa.2549$M8.1920442@reader1.news.jippii.net...G > We recently upgradet our VMS cluster from 7.2-1 to 7.3-1 and Advanced  ServerI > from 6.0C to 7.3A. Now when creating a file through Advanced Server the  fileI > stays locked much longer than before. This is causing problems with one 3 > application that tries to move the created files. H > Is there any way to configure Advanced Server so that the file locking time4 > is shorter or must I reconfigure the application ? >  > -Kari- >  >    ------------------------------  % Date: Tue, 13 May 2003 08:49:58 -0500 ( From: brandon@dalsemi.com (John Brandon)* Subject: Re: Advanced Server: file locking1 Message-ID: <03051308495885@dscis6-0.dalsemi.com>   I > > We recently upgradet our VMS cluster from 7.2-1 to 7.3-1 and Advanced  > ServerK > > from 6.0C to 7.3A. Now when creating a file through Advanced Server the  > fileK > > stays locked much longer than before. This is causing problems with one 5 > > application that tries to move the created files. J > > Is there any way to configure Advanced Server so that the file locking > time6 > > is shorter or must I reconfigure the application ? > > 
 > > -Kari-  O I have seen this also - I went from V5.0 to V7.2, however it did not impact our D environment as yours.  I noted the lock delay to be about 5 seconds.  L I have also had problems with a single W2K client accessing files.  The userJ would end up locking the file for use and even after that user had closed,O exited, and shutdown her PC the file was still locked.  This was never observed O with other Win clients (95, 98, NT).  I have other W2K clients that do not have 
 this problem.   L There was a recent thread about file locking you may be intersted in reading up on.   > From the release notes:  > & > 3.11 Disabling Opportunistic Locking6 > Opportunistic locking is enabled by default. Under a4 > few circumstances, customers might want to disable< > opportunistic locking, such as when they require immediate9 > access to files at the OpenVMS system level. Typically, : > these customers would also disable open file caching. To< > disable this feature so that files are not locked, add the7 > following two lines to the PWRK$COMMON:PWRK.INI file:  >  > [PLM]  > ENABLE_OPLOCKING = 0  N Not sure I would like doing this... looks like the fix is worse than the bite.A I would pursue a resolution in the application before I did this.      John Brandon VMS Systems Administrator  Dallas Semiconductor first.last@dalsemi.com 972.371.4172 wk    ------------------------------  % Date: Tue, 13 May 2003 10:07:10 -0400 * From: "Syltrem" <syltremzulu@videotron.ca>* Subject: Re: Advanced Server: file locking4 Message-ID: <M07wa.6817$945.20192@tor-nn1.netcom.ca>  E "John Brandon" <brandon@dalsemi.com> a crit dans le message de news: & 03051308495885@dscis6-0.dalsemi.com...   <snip>   > > 	 > > [PLM]  > > ENABLE_OPLOCKING = 0 > J > Not sure I would like doing this... looks like the fix is worse than the bite. C > I would pursue a resolution in the application before I did this.  >     4 And how do you propose to fix a closed application ? :-)   L The file remains opened by advanced server, even after the PC is turned off.C The fix has to be on the advanced server side, not the application.   5 Works great for us. We are back to pre-v7.3 behavior.   K If you use the files stored on OpenVMS only for PC apps you don't need this K setting. But if you use the same files on the VMS side too, you'll want it.    --   OpenVMS 7.2-1, Oracle 8.1.6.0    Syltrem I http://pages.infinit.net/syltrem (OpenVMS related web site - en franais) 8 To reply to myself directly, remove zulu from my address   ------------------------------  % Date: Tue, 13 May 2003 09:18:04 -0500 ( From: brandon@dalsemi.com (John Brandon)* Subject: Re: Advanced Server: file locking1 Message-ID: <03051309180473@dscis6-0.dalsemi.com>    > > > [PLM]  > > > ENABLE_OPLOCKING = 0 > > L > > Not sure I would like doing this... looks like the fix is worse than the > bite. E > > I would pursue a resolution in the application before I did this.  > > 6 > And how do you propose to fix a closed application ? > :-)    BFH   M Actually, if the application was attempting to access a locked file, then the E resolution would be to skip it and come back later.  But that too has 
 implications.   N > The file remains opened by advanced server, even after the PC is turned off.E > The fix has to be on the advanced server side, not the application.  > 7 > Works great for us. We are back to pre-v7.3 behavior.   1 I am still on V7.2A - wanting to upgrade to V7.3.   M > If you use the files stored on OpenVMS only for PC apps you don't need this M > setting. But if you use the same files on the VMS side too, you'll want it.   J My problem was with one client and that client was a Win 2000.  We had notK other problems with other Win 2000 so we opted to downgrade the Win 2000 to N Win/NT (or was that an upgrade???).  That resolved the problem and we have not
 had it since.   : The 5-second delay does not impact our environment at all.  L That said, I seem to recall that AS V7.2 was the first step towards Win 2000L compatibility and that AS V7.3 was the solution (outside of bug fixes, etc.) Does that ring a bell?       John Brandon VMS Systems Administrator  Dallas Semiconductor first.last@dalsemi.com 972.371.4172 wk    ------------------------------  % Date: Tue, 13 May 2003 11:45:11 -0400 * From: "Syltrem" <syltremzulu@videotron.ca>* Subject: Re: Advanced Server: file locking4 Message-ID: <Fs8wa.6835$945.20212@tor-nn1.netcom.ca>  F AS 7.3 supports Win 2000 domains and can be a passive member of ActiveH Directory (that is, knows it but cannot alter it). 7.4 will be an active
 member of AD. K I'm not sure how 7.2 was dealing with W2k domains, we didn't have any then.   F We were experiencing file locking problems (on the VMS side) after theL migration to 7.3. That was for all clients including those running on Win98.J The 5-second delay before the file is closed by AS is still there (as withJ 7.2) but it does not impact us. More than 1 PC user can open the file, butG once it's open then you cannot use it from a VMS application until it's + closed (PWRK$LMSRV has it and won't share).    --   OpenVMS 7.2-1, Oracle 8.1.6.0    Syltrem I http://pages.infinit.net/syltrem (OpenVMS related web site - en franais) 8 To reply to myself directly, remove zulu from my address  E "John Brandon" <brandon@dalsemi.com> a crit dans le message de news: & 03051309180473@dscis6-0.dalsemi.com...
 > > > > [PLM]  > > > > ENABLE_OPLOCKING = 0 > > > J > > > Not sure I would like doing this... looks like the fix is worse than the 	 > > bite. G > > > I would pursue a resolution in the application before I did this.  > > > 8 > > And how do you propose to fix a closed application ? > > :-)  >  > BFH  > K > Actually, if the application was attempting to access a locked file, then  the G > resolution would be to skip it and come back later.  But that too has  > implications.  > K > > The file remains opened by advanced server, even after the PC is turned  off.G > > The fix has to be on the advanced server side, not the application.  > > 9 > > Works great for us. We are back to pre-v7.3 behavior.  > 3 > I am still on V7.2A - wanting to upgrade to V7.3.  > J > > If you use the files stored on OpenVMS only for PC apps you don't need thisK > > setting. But if you use the same files on the VMS side too, you'll want  it.  > L > My problem was with one client and that client was a Win 2000.  We had notJ > other problems with other Win 2000 so we opted to downgrade the Win 2000 toL > Win/NT (or was that an upgrade???).  That resolved the problem and we have not  > had it since.  > < > The 5-second delay does not impact our environment at all. > I > That said, I seem to recall that AS V7.2 was the first step towards Win  2000H > compatibility and that AS V7.3 was the solution (outside of bug fixes, etc.)  > Does that ring a bell? >  >  > John Brandon > VMS Systems Administrator  > Dallas Semiconductor > first.last@dalsemi.com > 972.371.4172 wk    ------------------------------  % Date: Tue, 13 May 2003 11:07:47 -0500 ( From: brandon@dalsemi.com (John Brandon)* Subject: Re: Advanced Server: file locking1 Message-ID: <03051311074699@dscis6-0.dalsemi.com>   H > AS 7.3 supports Win 2000 domains and can be a passive member of ActiveJ > Directory (that is, knows it but cannot alter it). 7.4 will be an active > member of AD. M > I'm not sure how 7.2 was dealing with W2k domains, we didn't have any then.   O V7.2 is Win 2000 aware and have some compliancy with it, however I do not think 3 it can even be a passive member.  Not sure on that.   H > We were experiencing file locking problems (on the VMS side) after theN > migration to 7.3. That was for all clients including those running on Win98.L > The 5-second delay before the file is closed by AS is still there (as withL > 7.2) but it does not impact us. More than 1 PC user can open the file, butI > once it's open then you cannot use it from a VMS application until it's - > closed (PWRK$LMSRV has it and won't share).   K Then I would assume that the locking is a design feature.  And since A/S is G emulating a Win DOMAIN (or AD) then I would believe that this is proper 6 behavior.  OK, Win has no proper behavior... *snicker*  , OK, the 5-second delay is not at issue then.  O You say that more than one PC user can open the file.  I believe that the first G user has exclusive rights to the file, subsequent users have read only.   K What happens - when you flip the two parameters (as below) - to the PC?  On N VMS it is no longer locked, correct?  What about from another PC, is it lockedK or unlocked?  No effect?  Does the parameter only affect VMS?  Reading your 8 HELP snip I believe that only VMS is affected.  Curious.  
      [PLM]      ENABLE_OPLOCKING = 0     H This also disables caching.  What kind of impact does this have?  Slower response time?   John Brandon VMS Systems Administrator  Dallas Semiconductor first.last@dalsemi.com 972.371.4172 wk    ------------------------------  # Date: Tue, 13 May 2003 17:37:16 GMT , From: "Phil Hudson" <phil.hudson@compaq.com>5 Subject: Re: Are Bridgeworks and Multinet compatible? / Message-ID: <g3awa.761$_11.45@news.cpqcorp.net>    Warren, J     Note that I mentioned that a problem 'may' exist.  We don't personallyF     certify BridgeWorks against any other TCP/IP stack except for UCX.K     At the same time, we have never had a customer report a problem against G     BridgeWorks, stating that it doesn't work with some specific TCP/IP  stack.K     It is very possible that the FTP & REXEC clients built into BridgeWorks D     will parse the output from another FTP & REXEC server just fine.  -     So, there is nothing to fix at this time.   J     With that said, we can & will look into certifying the next release of,     BridgeWorks against other TCP/IP stacks.  G     In the meantime, please give it a try and let me know your results.  NoteE     that once you define & build the connection, everything else will E     work.  This is because the DCE/RPC team has certified against the      different TCP/IP stacks. Phil      : "Warren Spencer" <wspencer@ap.nospam.org> wrote in message/ news:937960576warrenspencer1977@216.168.3.30... / > phil.hudson@compaq.com (Phil Hudson) wrote in ' > <VSOua.590$SM3.584@news.cpqcorp.net>:  >  > Phil,  > H > Many thanks for the response. Is there fix planned for the FTP & REXEC > issue?  Time frame?  >  > ws > 
 > >Warren, > > H > >I apologize for the delay in responding to this note. I have been laxJ > >in monitoring this user group, so your question had to be brought to my
 > >attention.  > > , > >Anyway, the answer to your question is...J > >    Yes, Multinet will work just fine in place of TCP/IP  Services(UCX)F > >    at runtime. BridgeWorks generates a runtime connection based onI > >    DCE/RPC, which in turn has successfully tested against the various 3 > >    TCP/IP stacks out there, including MultiNet.  > > A > >The only problem that you may run into is at development time. @ > >The BridgeWorks UI uses FTP & REXEC to provide the integrated? > >build feature, and the automatic importing of the ANA & STDL ? > >files. The FTP & REXEC clients built into the BridgeWorks UI ? > >were written with the expectation that UCX based FTP & REXEC C > >servers would be on the other end.   It parses the returned data ; > >based on how these UCX utilities return the information.  > > C > >If this problem does exists, it only means that you will need to C > >manually FTP the generated files over to OpenVMS, and build them / > >by hand using the generated build procedure.  > >  > >Phil Hudson > >(HP BridgeWorks Engineering)    ------------------------------    Date: 13 May 2003 08:06:26 -0000= From: Doc.Cypher <Use-Author-Supplied-Address-Header@[127.1]> # Subject: Re: Best practices for VMS 5 Message-ID: <20030513080626.9632.qmail@nym.alias.net>   < On 12 May 2003, c00per11242001@yahoo.ca (Vic Mendham) wrote:  F >Specifically, I'm looking for which files to be reset, and how often.F >I realize this my depend on the adminsitrator or on the system usage. >But does anyone have anything?   M What policy or legislation impacts the logs you are required to keep?  That's M probably the main issue.  Beyond that, you want to keep enough data online to , track down problems that people may report.   0 I keep 7 days of logs, and clean out as follows:  , $ retain        = f$cvtime("-7-","ABSOLUTE")* $ DELETE/ERASE <file>;*/BEFORE="''retain'"  N This ensures you keep the correct files even if new versions have been created because of a reboot.    P I deal with the following (if I've missed anything, would someone point it out).  ' 1. TCP/IP Logs (In my case, Multinet).   2. Web logs (WASD). + 3. Performance monitoring logs (Hyperspi).   4. Accounting. 5. Audit. (Code below).  6. Operator log.  D >Also does anyone have a script which would create a new copy of theE >security audit file once a day or week and rename it to -1 or old or  >with the date?   O Here's how I handle the Audit log, if there's a better way I'd like to know it.   < $!      Check running processes to see old AUDIT_SERVER PID. $!
 $ SHOW SYSTEM  $!, $ set audit/journal=security/thresh=warn=200 $ set audit/server=exit O $ ren SYS$MANAGER:security.audit$journal SYS$MANAGER:security_old.audit$journal J $ DELETE/ERASE SYS$MANAGER:security_old.audit$journal;*/BEFORE="''retain'" $ set audit/server=start $ wait 00:00:05  $ set audit/server=init  $!   HTH.     Doc. --  K OpenVMS:  Eight out of ten hackers                    http://vmsbox.cjb.net K           prefer *other* operating systems.        http://althacker.cjb.net    ------------------------------   Date: 13 May 03 13:05:13 +0200) From: p_sture@elias.decus.ch (Paul Sture) # Subject: Re: Best practices for VMS ) Message-ID: <yxsSyPfQ29$c@elias.decus.ch>   u In article <20030513080626.9632.qmail@nym.alias.net>, Doc.Cypher <Use-Author-Supplied-Address-Header@[127.1]> writes: > > On 12 May 2003, c00per11242001@yahoo.ca (Vic Mendham) wrote: > G >>Specifically, I'm looking for which files to be reset, and how often. G >>I realize this my depend on the adminsitrator or on the system usage.   >>But does anyone have anything? > O > What policy or legislation impacts the logs you are required to keep?  That's O > probably the main issue.  Beyond that, you want to keep enough data online to . > track down problems that people may report.  > 2 > I keep 7 days of logs, and clean out as follows: > . > $ retain        = f$cvtime("-7-","ABSOLUTE"), > $ DELETE/ERASE <file>;*/BEFORE="''retain'" > P > This ensures you keep the correct files even if new versions have been created > because of a reboot.   > R > I deal with the following (if I've missed anything, would someone point it out). > ) > 1. TCP/IP Logs (In my case, Multinet).   > 2. Web logs (WASD). - > 3. Performance monitoring logs (Hyperspi).   > 4. Accounting. > 5. Audit. (Code below).  > 6. Operator log. > E >>Also does anyone have a script which would create a new copy of the F >>security audit file once a day or week and rename it to -1 or old or >>with the date? > Q > Here's how I handle the Audit log, if there's a better way I'd like to know it.  > > > $!      Check running processes to see old AUDIT_SERVER PID. > $! > $ SHOW SYSTEM  > $!. > $ set audit/journal=security/thresh=warn=200 > $ set audit/server=exit Q > $ ren SYS$MANAGER:security.audit$journal SYS$MANAGER:security_old.audit$journal L > $ DELETE/ERASE SYS$MANAGER:security_old.audit$journal;*/BEFORE="''retain'" > $ set audit/server=start > $ wait 00:00:05  > $ set audit/server=init  > $! >  > HTH. >   % I'm perhaps being picky, but wouldn't    $ set audit /server=new_log   4 eliminate the time window when auditing is inactive?   From the help:   SET      AUDIT        /SERVER              /SERVER=keyword[,...]  ... A        INITIATE           Enables auditing during system startup. >                           Ordinarily, auditing is started fromC                           VMS$LPBEGIN in STARTUP.COM but, if a site F                           redefines the logical name SYS$AUDIT_SERVER_E                           INHIBIT, the OpenVMS system waits for a SET G                           AUDIT/SERVER=INITIATE command before enabling  ... @        START              Starts the audit server process on theG                           system. In order to fully enable the auditing B                           subsystem, the SET AUDIT/SERVER=INITIATE<                           command must be used after the SETC                           AUDIT/SERVER=START command has completed.   G                           Compaq recommends using the following command >                           procedure to start the audit server:  9                           SYS$SYSTEM:STARTUP AUDIT_SERVER   D So, there's the delay of at least 5 seconds in your procedure beforeC auditing is re-enabled, and your rename/delete combination will not ! be included in the audit journal.   B Using SET AUDIT /SERVER=NEW_LOG, I would either rename the journalB beforehand, or use F$SEARCH to get the full name including versionC number and do the rename/delete afterwards. Probably the former, as  it's less lines of code.            --  
 Paul Sture   ------------------------------  % Date: Tue, 13 May 2003 06:39:52 -0500 ( From: brandon@dalsemi.com (John Brandon)# Subject: Re: Best practices for VMS 1 Message-ID: <03051306395225@dscis6-0.dalsemi.com>   B > Active development environments can have the same directory treeI > repeated many times under different upper-level directories on the same H > spindle or volume-set. Sometimes they need to compare results based on' > multiple variations of the same code.   N As I previously stated, this is not a problem.  We have many duplicates.  FromE time to time we have went back and "reminded" development to cleanup  E directory trees.  We have from time to time, went back and found that L we had duplication left over from a move.  It happens.  The intent is NOT to6 keep one directory but to keep an eye on what we have.   >  DFU does this, also.    Does not DFU require a license?      John Brandon VMS Systems Administrator  Dallas Semiconductor first.last@dalsemi.com 972.371.4172 wk    ------------------------------   Date: 13 May 2003 08:20 CDT ' From: carl@gerg.tamu.edu (Carl Perkins) # Subject: Re: Best practices for VMS - Message-ID: <13MAY200308205494@gerg.tamu.edu>   , brandon@dalsemi.com (John Brandon) writes...  }Does not DFU require a license? }  } 
 }John Brandon   C It does not require one. There is a freeware version. It is spiffy.    --- Carl   ------------------------------  % Date: Tue, 13 May 2003 06:58:46 -0700 . From: "David D Miller" <ddmiller@raytheon.com># Subject: Re: Best practices for VMS F Message-ID: <OFBF9F8FB2.5D13F8E5-ON07256D25.004C259D@rsc.raytheon.com>   John:   I Yep I took a crack at listing all the files I knew about.  Not only that, K but I included a COM file to reset the version numbers.  As to the specific I commands to restart each file ... I can't remember if I included that off I hand.  That would be a nice addition.  Contributions gratefully accepted.   2 "Miller's rev of Hoff's update".  I like that.  :)   dave.        John Santos writes ...  " On 12 May 2003, Vic Mendham wrote:  : > Just wondering if anyone has any best practices for VMS?H > Things like reset the operator.log once a day or once a week or month./ > Reset account.dat once a day, week, or Month. 4 > Reset the security logs once a day, week or month., > Disk space not to be allowed over 80% usedD > Create a backup copy of the sysuaf file, incase the production one > gets corrupted. H > always (if budgets allow) carry a spare system disk (offline), and use > backup to  > copy once a week.  > G > Specifically, I'm looking for which files to be reset, and how often. G > I realize this my depend on the adminsitrator or on the system usage.   > But does anyone have anything? > E > Also does anyone have a script which would create a new copy of the F > security audit file once a day or week and rename it to -1 or old or > with the date? > 	 > Regards   B I wonder if Dave Miller's rev of Hoff's update of the Baldwin book will cover this?   -- John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------    Date: 13 May 2003 15:58:04 -00004 From: Doc.Cypher <Use-Author-Address-Header@[127.1]># Subject: Re: Best practices for VMS 5 Message-ID: <20030513155804.4242.qmail@gacracker.org>   8 On 13 May 03, p_sture@elias.decus.ch (Paul Sture) wrote:A >In article <20030513080626.9632.qmail@nym.alias.net>, Doc.Cypher 5 ><Use-Author-Supplied-Address-Header@[127.1]> writes:   / >> $ set audit/journal=security/thresh=warn=200  >> $ set audit/server=exit+ >> $ ren SYS$MANAGER:security.audit$journal ) >> SYS$MANAGER:security_old.audit$journal M >> $ DELETE/ERASE SYS$MANAGER:security_old.audit$journal;*/BEFORE="''retain'"  >> $ set audit/server=start  >> $ wait 00:00:05 >> $ set audit/server=init  & >I'm perhaps being picky, but wouldn't >  >$ set audit /server=new_log > 5 >eliminate the time window when auditing is inactive?   # That would be much better, thanks.     I've ended up using   $ $ SET AUDIT /SERVER=(FLUSH, NEW_LOG)  3 I really don't know if the flush is needed though.       Doc. --  6 The bigger the humbug, the better people will like it.K ~ Phineas Taylor Barnum.                             https://vmsbox.cjb.net K                                                    http://althacker.cjb.net    ------------------------------  + Date: Tue, 13 May 2003 05:59:58 +0000 (UTC) 3 From: "Richard Maher" <maher_rj@hotspamnotmail.com> @ Subject: BLASTed directory locks, timing windows & endless loops2 Message-ID: <b9q1ku$r1a$1@hercules.btinternet.com>   Hi,   J I have the serialization lock on a directory with a BLAST to check for theK creation of a file that I'm waiting for. Before I do a $search for the file H I re-queue the lock to eliminate the timing window where the target fileG could be created after the search but before I take out the lock again.   G The problem is that my $search (or $qio io$_access or DIR) triggers the J BLAST because it wants the directory lock. So My program just loops around beautifully. Any ideas?   I Couldn't see any directory file attributes for revision that tell me when % the directory *contents* has changed.    Regards Richard Maher    ------------------------------   Date: 13 May 2003 06:02 CDT ' From: carl@gerg.tamu.edu (Carl Perkins) D Subject: Re: BLASTed directory locks, timing windows & endless loops- Message-ID: <13MAY200306023241@gerg.tamu.edu>   7 "Richard Maher" <maher_rj@hotspamnotmail.com> writes... K }I have the serialization lock on a directory with a BLAST to check for the L }creation of a file that I'm waiting for. Before I do a $search for the fileI }I re-queue the lock to eliminate the timing window where the target file H }could be created after the search but before I take out the lock again. } H }The problem is that my $search (or $qio io$_access or DIR) triggers theK }BLAST because it wants the directory lock. So My program just loops around  }beautifully. Any ideas? } J }Couldn't see any directory file attributes for revision that tell me when& }the directory *contents* has changed. }  }Regards Richard Maher  K As an aside, you should consider converting your lock to NL mode instead of G dequeing it, and then just convert it back. This is more efficient (and G faster) than a dequeue and enqueue, which is what the above sounds like G you are doing. (Using a PR mode lock instead of EX makes it so that the E conversion to NL won't write to the lock value block which the XQP is 
 maintaining.)   	 Anyhow...   I If you are sufficiently sneaky, it may be possible to do it such that the H blocking AST doesn't directly trigger the directory check, but only doesH so indirectly (allowing for some checkig to see if it is needed) or only# if some other indicator is not set.   A Example: (This is one way to do things. I thought of what I think @ is a better way, given in the paragraph after this one.) Say youG convert your lock to NL mode and then convert it back to a higher mode, D doing the former in the blocking AST and the latter right before youE do the lookup. This should let the other queued lock get the resource H when you go to NL and then put your lock in the conversion queue waitingA to lock the resource when the other thing is done with it (so the C conversion from NL would be synchronous to fire off your check when G they are done doing whatever it is they were doing). When the AST runs, C you set a flag (or whatever) to tell your main process to check the F directory. If it is already checking, don't set the indicator. InsteadI the lock would be put in the conversion queue to go from NL to the higher J mode right there inside the blocking AST. Then the more or less extraneousI firing of your blocking AST when you do the check wouldn't itself set off H another check, instead it would just convert the lock to NL and then putH it in the conversion queue to wait on your checking to finish so that itH will be converted to lock the resource again. I hope. There may still beK a window where strange things could happen in there after your main process M has done the lookup but before it can reset the indicator (it may be possible E to use something for the indicator that minimizes, or eliminates, the E window - such as the value of the status part of the directory lookup E QIO's IOSB which you initialize to a non-zero value and then set to 0 G to indicate that you are doing the lookup, and which the IO will set to F something else for you itself when it is done since there is no status code of 0).   G Another possible way to eliminate the extraneous looping (which is also G quite sneaky) would be for your blocking AST to enqueue an EX mode lock D on some other resource of your own invention (possibly involving theG name of the directory). This $ENQ would have an AST specified (AST, notTG blocking AST - parameter 7 rather than 9). This AST fires when the lock:H is granted, triggering the actual directory lookup. After the lookup youE can just cancel one waiting but not yet granted lock request and thenfG dequeue the granted lock. That one waiting lock that is canceled is thelE one that your own directory lookup caused to happen. If anything else G touches the directory and triggers the blocking AST after you check butVF before you do your cleanup, there will be one or more additional locksH queued up on your resource after your cancel just one extraneous one, soG when you dequeue your lock another one will be granted and it will fire G off its AST and trigger a directory check again. I don't see any timingiE hole in this method, although you could get some extra lookups if thetE blocking AST is fired off more than once before you manage to do youryG actual directory lookup. You could reduce this problem by canceling all E enqueued but ungranted lock requests, if any, waiting on the resource  before you do your lookup.  E The more I think about this, the more I like this additional resource F lock AST method. It seems to be relatively simple and free from timing	 problems.e  F Something you may need to keep in mind: if your lock is in EX mode youF can only have one process waiting for a file (or files) to be added toG any given directory. If you try this with two different processes, then E each process will try to have a lock on the serialization resource innI EX mode and they will alternate causing the other's blocking ASTs to fire I in an endless (or it would probably seem that way, anyhow) system slowingOG lock swapping loop. A fine reason to go with the PR mode lock for this, B as long as the XQP uses EX mode locks - PR mode locks are mutuallyD compatable and will not block each other, and each will fire its ownF blocking AST when the XQP comes along and tries to get an EX mode lockG (assuming it does). At that point you only have to be careful with your 7 trigger resource, so that it is unique to each process.r   --- Carl   ------------------------------  # Date: Tue, 13 May 2003 09:02:18 GMTt* From: "Mark Buda" <buda_NO@SPAM.yahoo.com>0 Subject: Re: creating licenses for my own demos?< Message-ID: <uw2wa.17534$Jf.8741797@news1.news.adelphia.net>  . "Beyonder" <beyonder@vrx.net> wrote in message2 news:bvtgbv8uqqqg9b83eg29hf0f9iecu8gvbm@4ax.com...  K > >So, If you have V7.2 or later of PAKGEN and a PAKGEN PAK, you can create:6 > >PAKs for the producer that matches the TOKEN field.& > nope. It doesn't. I have 7.2 openvms > and I *DON'T* have PAKGEN @ > wasn't installed. and I looked everywhere for it. ain't there. > ( > lovely english, but you get the point.< > of course this is on an ALPHA that I'm testing, not a vax.6 > perhaps I should be less stupid and check the vax ;) >d > curiouser and curiouser   ; Excuse my "missing" implied VMS.  Please accept my apology.i  G PAKGEN is built into VMS from V7.2 and thereafter, much like SIPP's are  (RMSJNL, etc.)  I Everyone has it on V7.2 after - Yep everyone.  It is enabled by getting aS) pack called PAKGEN from the DSPP program.o  K The directions I posted specify how to use it.  Out of it comes a real live-J PAK with your producer/issuer that you are allowed to use, after obtaining it from the DSPP program.g  < After reading the documentation, let us know what you think.   ------------------------------    Date: 12 May 2003 23:12:46 -0700' From: icerq4a@spray.se (David Svensson)-E Subject: Re: How Alpha will save Itanium - must reading for Bill Todda= Message-ID: <734da31c.0305122212.20239f33@posting.google.com>g   Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote in message news:<b9ojiu$444$1@new-usenet.uk.sun.com>...d > David Svensson wrote: < > > "Bill Todd" <billtodd@metrocast.net> wrote in message >  > > O > >>The new Madison TPC-C result *is* on the next-gen box, Rob:  it's using theIP > >>new Pinnacles chipset, and you'll be able to buy one this summer (though whyM > >>you'd want to is not clear, given this first glimpse of its performance).e > >> > >  > > J > > The Superdome is a nice machine with very good performance, and I knowG > > many who would like to buy these. You don't have to be number 1 all-G > > the time! Companies which replace a 4 year old high-end server willl: > > get a good increase in performance with the Superdome. >  > Humm > E > Except that the new shiny SuperDome with Itanium III in is probablyaE > about the same speed as the SuperDome with the next HP-PA processorn > in.a > C > Now given that the 1 GHz HP-PA processor has things like softwareb@ > available for it and you don't need to migrate off anything toC > get to it why would you bother with Itanium III based SuperDomes.r > B > HP-PA even runs 32 bits apps well and you can even do it without0 > having to run them through a binary converter. > D > Now if HP had come out with a replacement for the Dome with a halfC > decent backplane, I/O etc and put Itanium III in it then that mayf) > have been interesting, but they didn't.  > B > Currently the F15K has ~2x the bisectional bandwidth of the Dome? > the P690 slightly higher but half the CPU's. Putting the Domey > in the rather tired category.k > @ > In fact HP's execution on Enterprise IA-64 based systems is so@ > bad that despite the fact that they helped design IA-64 it was? > after all their high end replacement for HP-PA they have been3> > beaten to the market by Unisys, NEC and SGI with large scale > IA-64 systems. > 	 > Regardse > Andrew Harrisonn  D My point was on Superdome in general, HP-PA or IA-64 doesn't matter.C Those who have bought Superdome or will buy Superdome are satisfiedlE with it's performance and price/performance. Most customers don't see C a big difference in reality between high-end Sun/IBM/HP boxes. WhenaC they buy one of them, they are often happy, it doesn't matter whichj
 one it is.   ------------------------------  % Date: Tue, 13 May 2003 02:28:02 -0400m* From: "Bill Todd" <billtodd@metrocast.net>E Subject: Re: How Alpha will save Itanium - must reading for Bill ToddN2 Message-ID: <2sWdneSddevvDV2jXTWcow@metrocast.net>  4 "David Svensson" <icerq4a@spray.se> wrote in message7 news:734da31c.0305122212.20239f33@posting.google.com...s   ...e  F > My point was on Superdome in general, HP-PA or IA-64 doesn't matter.E > Those who have bought Superdome or will buy Superdome are satisfiedeG > with it's performance and price/performance. Most customers don't see5E > a big difference in reality between high-end Sun/IBM/HP boxes. WhentE > they buy one of them, they are often happy, it doesn't matter whicht > one it is.  K People who have directed the acquisition of high-priced systems tend not tovI be dissatisfied with their purchases, since it can get them fired.  (Yes,iK some actually *are* happy with them as well.)  And since most people making H high-end purchasing decisions really aren't that sharp technically, theyJ indeed do not 'see' big differences among Sun/IBM/HP offerings even though" such differences do in fact exist.  J This particular forum, however, *is* pretty sensitive to such differences:K there's a decided disdain for Sun systems (or at least Sun processors) heresJ because of their relatively poor performance (though less recognition thatF they also consume commensurately less power and chip area and that SunK systems seem to leverage what performance they have quite effectively), anddF if SuperDomes are similarly hobbled (in their case by a poorly-scalingI system architecture) then they aren't likely to be viewed very positivelyiL either, at least while POWERx systems (and Alphas) prove that the job can be done far better.   - bill   ------------------------------  # Date: Tue, 13 May 2003 12:08:07 GMTa" From: "ravi" <hm.ravi@digital.com># Subject: How do i analyse TCPDump ?e/ Message-ID: <He5wa.744$8H.427@news.cpqcorp.net>m   Hi All,uE     We are getting some TCPDUMP from the our client which needs to beDL analysed. We found a tool called TCPTRACE in Tcptrace website. But this toolK expects dumps in specific formats only . If then how do I analyse TCPDump ?.? The dumps were produced using the TCPTRACE utility of OpenVMS .o   regardsu ravi   ------------------------------    Date: 13 May 2003 10:38:50 -0700) From: baby_p_nut2@yahoo.com (Baby Peanut)f7 Subject: How many Intel Itaniums have been sold so far? = Message-ID: <96d83290.0305130938.2db78d21@posting.google.com>e  & http://www.chipzilla.org/?article=9467   ------------------------------    Date: 13 May 2003 08:04:26 -0700  From: chnzw@yahoo.com.cn (jeams): Subject: how to use smg$create_menu in vax macro language?= Message-ID: <f7e545c0.0305130704.3dcbe616@posting.google.com>o  8 .macro CREATION_MENU menu_id, menu, nb_choice, menu_type pushal nb_choice	 pushl  #0  pushal menu_type pushal menug pushal menu_id calls  #5, g^smg$create_menu e .endm CREATION_MENUr   and i call it like this:   menu .ascid / test /  / CREATION_MENU menu_id, menu, #3, SMG$K_VERTICAL   
 is it right ?    ------------------------------  % Date: Tue, 13 May 2003 17:10:22 +0200v% From: "Jeams ZHI" <chnzw@hotmail.com>1> Subject: Re: how to use smg$create_menu in vax macro language?3 Message-ID: <b9r1ro$17q$1@news-reader13.wanadoo.fr>.   ERROR: invalide arguments ???   "jeams" <chnzw@yahoo.com.cn>9 ??????:f7e545c0.0305130704.3dcbe616@posting.google.com... : > .macro CREATION_MENU menu_id, menu, nb_choice, menu_type > pushal nb_choice > pushl  #0a > pushal menu_type
 > pushal menu  > pushal menu_id > calls  #5, g^smg$create_menu > .endm CREATION_MENUt >  > and i call it like this: >s > menu .ascid / test / >e1 > CREATION_MENU menu_id, menu, #3, SMG$K_VERTICAL  >. > is it right ?o   ------------------------------  % Date: Tue, 13 May 2003 17:04:43 +0100o0 From: Chris Sharman <chris.sharman@sorry.nospam>> Subject: Re: how to use smg$create_menu in vax macro language?4 Message-ID: <b9r533$spg$1$8302bc10@news.demon.co.uk>   jeams wrote:: > .macro CREATION_MENU menu_id, menu, nb_choice, menu_type > pushal nb_choice > pushl  #0t > pushal menu_type
 > pushal menuW > pushal menu_id > calls  #5, g^smg$create_menu   > .endm CREATION_MENUn >  > and i call it like this: >  > menu .ascid / test / > 1 > CREATION_MENU menu_id, menu, #3, SMG$K_VERTICALe >  > is it right ?l   Nope.o7 iirc you need to pass in a conformant array of strings.rF Very easy in Pascal, but in Macro you'll have to hand-build the array I descriptor - see the descriptor stuff (dsc$) in the manuals & in starlet.g   Chris    ------------------------------  % Date: Tue, 13 May 2003 18:38:53 +0200.% From: "Jeams ZHI" <chnzw@hotmail.com>-> Subject: Re: how to use smg$create_menu in vax macro language?3 Message-ID: <b9r71p$9kt$1@news-reader11.wanadoo.fr>u   but how to define the macro ?l  , "Chris Sharman" <chris.sharman@sorry.nospam>0 ??????:b9r533$spg$1$8302bc10@news.demon.co.uk... > jeams wrote:< > > .macro CREATION_MENU menu_id, menu, nb_choice, menu_type > > pushal nb_choice
 > > pushl  #0b > > pushal menu_type > > pushal menua > > pushal menu_id  > > calls  #5, g^smg$create_menu > > .endm CREATION_MENUo > >  > > and i call it like this: > >e > > menu .ascid / test / > >r3 > > CREATION_MENU menu_id, menu, #3, SMG$K_VERTICAL- > >d > > is it right ?i >i > Nope.b9 > iirc you need to pass in a conformant array of strings.sG > Very easy in Pascal, but in Macro you'll have to hand-build the arraynK > descriptor - see the descriptor stuff (dsc$) in the manuals & in starlet.o >  > Chris  >-   ------------------------------   Date: 13 May 03 10:17:18 PST From: mckinneyj@cpva.saic.com > Subject: Re: how to use smg$create_menu in vax macro language?( Message-ID: <hz1Xkb1g5SO+@cpva.saic.com>  3 In article <b9r1ro$17q$1@news-reader13.wanadoo.fr>,5(  "Jeams ZHI" <chnzw@hotmail.com> writes: > ERROR: > invalide arguments ??? >  > "jeams" <chnzw@yahoo.com.cn>; > ??????:f7e545c0.0305130704.3dcbe616@posting.google.com... ; >> .macro CREATION_MENU menu_id, menu, nb_choice, menu_typeb >> pushal nb_choice  >> pushl  #0 >> pushal menu_type  >> pushal menu >> pushal menu_id  >> calls  #5, g^smg$create_menut >> .endm CREATION_MENU >> >> and i call it like this:e >> >> menu .ascid / test /b >>2 >> CREATION_MENU menu_id, menu, #3, SMG$K_VERTICAL >> >> is it right ? >  >   C You'll need to store the value of SMG$K_VERTICAL somewhere and pass.@ it's location (menu-type is passed by reference). Whether or notA that makes it right depends on the whether menu is the address ofa= a descriptor pointing to a character string array and whether + menu_id is the address of your display-id. V   --   - Jima   ------------------------------  % Date: Tue, 13 May 2003 07:09:03 -0700 . From: "David D Miller" <ddmiller@raytheon.com> Subject: INIT/SHADOWF Message-ID: <OF48E11387.38CF20C1-ON07256D25.004D63C8@rsc.raytheon.com>   Folks:  F I must be blind.  What am I doing wrong?  Why doesn't INIT work?  This" command is right out of HELP INIT.   SYS$HELP? sho dev dk  E Device                  Device           Error    Volume         Freel	 Trans MnthG   Name                   Status           Count     Label        Blocks0	 Count CnttF VAX$DKA0:               Mounted              1  OVMSVAXSYS     1696356 183   1t. VAX$DKA100:             Online               0. VAX$DKA400:             Online wrtlck        0. VAX$DKB200:             Online               0  3 SYS$HELP? init/erase/shadow=(dka100:,dkb200:) spare3E %DCL-W-IVQUAL, unrecognized qualifier - check validity, spelling, and:	 placemente
   \SHADOW\	 SYS$HELP?s  J VAX/VMS 7.3 with level 1, SHAD01 and MOUN01 patches.  VAXstation 3100 M48.   dave.d   ------------------------------    Date: 13 May 2003 09:18:13 -0500+ From: young_r@encompasserve.org (Rob Young)e Subject: Re: INIT/SHADOW3 Message-ID: <nHQdOUfifwkY@eisner.encompasserve.org>v  w In article <OF48E11387.38CF20C1-ON07256D25.004D63C8@rsc.raytheon.com>, "David D Miller" <ddmiller@raytheon.com> writes:m > Folks: > H > I must be blind.  What am I doing wrong?  Why doesn't INIT work?  This$ > command is right out of HELP INIT. >  > SYS$HELP? sho dev dk > G > Device                  Device           Error    Volume         Freeb > Trans Mnt I >   Name                   Status           Count     Label        Blocksp > Count CntbH > VAX$DKA0:               Mounted              1  OVMSVAXSYS     1696356	 > 183   1s0 > VAX$DKA100:             Online               00 > VAX$DKA400:             Online wrtlck        00 > VAX$DKB200:             Online               0 > 5 > SYS$HELP? init/erase/shadow=(dka100:,dkb200:) spare,G > %DCL-W-IVQUAL, unrecognized qualifier - check validity, spelling, and  > placement. >   \SHADOW\ > SYS$HELP?o > L > VAX/VMS 7.3 with level 1, SHAD01 and MOUN01 patches.  VAXstation 3100 M48. >    	You left out the shadowset.  6 	$ init/erase  dsa100: /shadow=(dka100:,dkb200:) spare   			Rob   ------------------------------  % Date: Tue, 13 May 2003 10:49:32 -0700f. From: "David D Miller" <ddmiller@raytheon.com> Subject: Re: INIT/SHADOWF Message-ID: <OFAE977FC4.12323949-ON07256D25.00619387@rsc.raytheon.com>   Rob.  > This example is cut/pasted out of the Volume Shadowing manual:  @ $ INITIALIZE /ERASE /SHADOW=($4$DKA1300, $4$DKA1301) NONVOLATILE> $ MOUN/SYS DSA42 /SHAD=( $4$DKA1300 , $4$DKA1301 ) NONVOLATILE  E The way I read it, the MOUNT, not the INITIALIZE specifies the shadow- device.-  F So I repeat my question. "What am I doing wrong in the example below?"   dave   Rob Young responded.  F In article <OF48E11387.38CF20C1-ON07256D25.004D63C8@rsc.raytheon.com>,0 "David D Miller" <ddmiller@raytheon.com> writes: > Folks: >aH > I must be blind.  What am I doing wrong?  Why doesn't INIT work?  This$ > command is right out of HELP INIT. >  > SYS$HELP? sho dev dk >vG > Device                  Device           Error    Volume         Freeg > Trans MntnI >   Name                   Status           Count     Label        Blockse > Count CntcH > VAX$DKA0:               Mounted              1  OVMSVAXSYS     1696356	 > 183   1r0 > VAX$DKA100:             Online               00 > VAX$DKA400:             Online wrtlck        00 > VAX$DKB200:             Online               0 >e5 > SYS$HELP? init/erase/shadow=(dka100:,dkb200:) sparemG > %DCL-W-IVQUAL, unrecognized qualifier - check validity, spelling, andc > placement& >   \SHADOW\ > SYS$HELP?y >nG > VAX/VMS 7.3 with level 1, SHAD01 and MOUN01 patches.  VAXstation 3100g M48. >   (              You left out the shadowset.  B              $ init/erase  dsa100: /shadow=(dka100:,dkb200:) spare  (                                      Rob   ------------------------------    Date: 13 May 2003 09:15:07 -0700* From: ashoffman@comcast.net (Alan Hoffman)( Subject: Re: Java "Unknown host" problem= Message-ID: <838d7cb0.0305130815.7de1d8ee@posting.google.com>   b lewis@PROBE.mitre.org (Keith A. Lewis) wrote in message news:<b9h31s$kkn$1@newslocal.mitre.org>... > ashoffman@comcast.net (Alan Hoffman) writes in article <838d7cb0.0305081255.2e4f3eca@posting.google.com> dated 8 May 2003 13:55:14 -0700:se > >lewis@PROBE.mitre.org (Keith A. Lewis) wrote in message news:<b994av$nbe$1@newslocal.mitre.org>...tQ > >> How's your nameserver configuration?  Try "TCPIP SHOW HOST WSA1" and see howsD > >> long that takes.  (I'm assuming you use Compaq's tcp/ip stack.) > >tH > >    I ran the command and the timeout was 2 minutes (the same as when> > >I run Java). How do you check the nameserver configuration? > H > Which tcp/ip stack are you running, if any?  The person who set up theM > machine should know.  Once you say which stack, the folks here can tell youh% > how to debug the nameserver config.E > G > Did you try the command in the above paragraph?  If so, what were the J > results?  There's a good chance it relates directly to your problem.  IfD > Java is attempting to look up a bad hostname with a bad nameserver- > configuration, it could eat up some time.  b > H > >    These all show WSA1 as a device that is local (which I understandI > >is what it should be). What I don't understand is why Java is treating H > >WSA1 as a host (Unknown host: WSA1) instead of a device (are they theH > >same thing??) and how to get rid of the message all together (and notI > >just mask it). I would also like to find out how to set the timeout so1I > >that it is more like your 10 seconds instead of 2 minutes (as timed bysH > >a watch) so I'm not wasting so much time while I'm debugging my code.H > >    There was an older version of Java on the machine (1.1.8 I think)D > >that didn't produce the Unknown host problem on the same machine. > J > VMS allows the DECW$DISPLAY logical to point to either a WSA device or a1 > Unix-style display name (host:server.screen).  e > N > The VMS implementation of Java is apparantly doing a poor job of finding outI > which one is being used -- attempting to use it as a hostname first ande- > using the device name only if that fails.  e > L > If I'm right about what's causing the delay, the command "def decw$displayL > localhost:0.0" will eliminate it.  But in order for that to work, you mustH > have the TCP/IP transport enabled in DECwindows and authorize incomingB > connections from localhost in your DECwindows security settings. > ? > Obtaining and installing Alpha Linux might be faster.  *sigh*t > - > --Keith Lewis              klewis$mitre.org @ > The above may not (yet) represent the opinions of my employer.    F Our system person wasn't sure what you meant by tcpip stack but we are? running TCPIP version 5.3 on a Decstation DS10. I tried setting B decw$display to localhost:0.0 and got the following message when I tried to run Java:  D Exception in thread "main" java.lang.InternalError: Can't connect toC X11 window server using ':0.0' as the value of the DISPLAY variable   A I assume that this means that something needs to be configured in E TCPIP. We have the security settings all set to "*" so anybody should  be able to connect..  E Also when I ran "TCPIP SHOW HOST WSA1" I got the 2 minute timeout and. then an error message.( %TCPIP-W-NORECORD, information not found -RMS-E-RNF, record not found       Thanks,l         Alan   ------------------------------    Date: 13 May 2003 06:48:10 -0700+ From: n8pvl@yahoo.com (Christopher Francis)iD Subject: Re: MicroVAX and VAX models EOSL list (end of service life)= Message-ID: <7ea798ef.0305130548.4acb20c8@posting.google.com>n  F It is also interesting that the 4000-600 is not going EOSL. It is veryF disturbing that there was not ample notice. I was told from HP that weD would get notified 1 to 2 years in advance of anything going EOSL. IB have not yet received any official notification from HP concerning this.b  N rob.buxton@wcc.spam.govt.nz wrote in message news:<3ec06334.112685803@news>.... > On 10 May 2003 03:28:22 GMT, "Zane H. Healy"% > <healyzh@shell1.aracnet.com> wrote: G > What is also quite interesting is that the 4000-705A has gone EOL butE > the 4000-700A has not.7 > Your comment about parts availability may be spot on.n >  > * > >Rich Jordan <jordan@ccs4vms.com> wrote:J > >> We have some very unhappy customers as a result of this; they've beenJ > >> paying good money for support on top of hefty dollars on the originalE > >> purchases, and do not like the direction HP is taking with theseAD > >> really not _that_ old systems, given the history of longer term > >> support in the past...i > >uM > >While I agree that this stinks, I'm wondering, how much of this is becuasemF > >they no longer have the parts to provide service on these systems?  > >oM > >Of course if that is the case, then the question would be, how much of therN > >reason for that is Compaq dumping the contents of numerous DEC warehouses a > >few years ago?t > >=
 > >			Zane   ------------------------------    Date: 13 May 2003 10:47:20 -0700% From: rbsurfshop@hotmail.com (Roscoe) Y Subject: Re: Not fixed yet but still trying: Problems changing from serial port to DecSer/= Message-ID: <14f75c13.0305130947.1de3a977@posting.google.com>   E Finally - it works!  There were two responses in particular that wereiE of great help.  Mr. Lewis pointed out that I got sidetracked and thatvA my "write" SYS$QIOW didn't look like a "write".  He was correct.  B Also, Mr. Reynolds included Fortran code for his application whichD looks to be nearly identical to ours.  That too was a great help.  IE suspect it was really the fact that I didn't write to the port first,T) however both responses were very helpful.   C Thanks again to all who took the time to write with suggestions and  help.N   Roscoe at ISG Steel Corp      j rbsurfshop@hotmail.com (Roscoe) wrote in message news:<14f75c13.0305090945.38e8d153@posting.google.com>...E > Thanks for all the responses thus far.  I'm not sure if this is the H > proper way to post a follow-up so I appologize if there's another way. > D > I've tried several things as suggested from your guidance however,; > it's still not working.  Here's what's happened thus far.I > A > 1) I tried changing the terminal characteristics to Hostsync aspG > suggested(instead of No Hostsync).  However, this did not make it anyc	 > better.a >  > H > 2) To answer Mr. Sonderegger's question - yes, this is how my LTA port > was setup: >  > $ RUN SYS$SYSTEM:LATCP$ >   CREATE PORT LTA2312 /APPLICATIONE >   SET PORT LTA2312 /NODE=decserver/PORT=prot /APPLICATION /NOQUEUEDe >  > E > 3) To answer Mr. Brown's question - I do not have a service definedaE > for the port.  Just an LTA (LTA2312:).  And yes, I can SET HOST/DTEi0 > LTA2312: and bidirectional comms are possible. >  > D > 4) On a positive note I was able to use the following from DCL andC > receive data from the VT340 (and BTW that's in REMOTE mode on the0 > terminal server).  >  > $ ALLOCATE LTA2312: TERM3 > $ DIR /OUT=TERM (directory listing goes to VT340) A > $ DUMP TERM (keyboard entry is returned after carriage returns)r >  > D > 5) The most prevalent information given on your replies was that aB > "write" must be issued to the port first in order to establish aG > connection.  So, I headed off in that direction.  I also found an oldeC > post entitled "Need help reading data from a decserver port".  Itv= > probably would have helped if I would have told you that myeG > application was written in Fortran.  It was great news when I saw Mr.tB > Moulton's Fortran code describing this exact problem and his fixD > (shown directly below).  However, I tried it and I can't get it toF > work either.  This is of course placed right before my SYS$QIOW call  > that reads data from the port. > 9 >     FUNCTION_CODE = JIOR(IO$_TTY_PORT, IO$M_LT_CONNECT)t > " >     RETURN=SYS$QIOW(,%VAL(CHAN), > $   %VAL(FUNCTION_CODE), > $   IOSB,2. > $   ,,              ! AST address, parameter/ > $   ,,,,,)  ! This function has no parameterse > , >  A read QIO to the channel will then work. >  > C > So, I'm one step closer in that I can read data from the VT340 in @ > order to test the serial port connection.  However, my Fortran) > application does not want to cooperate.  > A > No doubt I'm still doing something wrong.  I'm just hoping that 2 > someone much brighter than me can figure it out. >  > Thanks again,i >  > Roscoe at ISG Steel CompanyB >  > 5 > PS - Here's the terminal settings in case it helps:e > B > Terminal: _LTA2312:   Device_Type: Unknown       Owner: No Owner > D >    Input:    9600     LFfill:  0      Width:  80      Parity: None2 >    Output:   9600     CRfill:  0      Page:   24 >  > Terminal Characteristics:dG >    Interactive        Echo               Type_ahead         No Escape D >    Hostsync           TTsync             Lowercase          No Tab@ >    Wrap               Scope              No Remote          No
 > EightbitE >    Broadcast          No Readsync        No Form            FullduprD >    No Modem           No Local_echo      No Autobaud        HangupG >    No Brdcstmbx       No DMA             No Altypeahd       Set_speedo@ >    No Commsync        Line Editing       Overstrike editing No
 > FallbackH >    No Dialup          No Secure server   No Disconnect      No PasthruH >    No Syspassword     No SIXEL Graphics  No Soft Characters No Printer > Port@ >    Numeric Keypad     No ANSI_CRT        No Regis           No > Block_mode@ >    No Advanced_video  No Edit_mode       No DEC_CRT         No
 > DEC_CRT2@ >    No DEC_CRT3        No DEC_CRT4        No DEC_CRT5        No > Ansi_Color >    VMS Style Input   ------------------------------    Date: 13 May 2003 08:20:33 -07001 From: susan_skonetski@hotmail.com (Sue Skonetski)dB Subject: OpenVMS Pearl - OpenVMS Article on InTech Brazil Magazine= Message-ID: <857e9e41.0305130720.73bf00cb@posting.google.com>a   From: Skonetski, Susan 5$ Sent: Tuesday, May 13, 2003 10:36 AM To: Skonetski, SusanB Subject: OpenVMS Pearl - OpenVMS Article on InTech Brazil Magazine     Folks,  F Renato (OpenVMS Ambassador) has sent me the article referenced which IA have not included since it is in Portuguese.  I you would like it  please let me know.0   OK for external distribution  
 Warm Regards,s Sueh     >  -----Original Message-----  > From: 	Albuquerque, Renato  $ > Sent:	Monday, May 12, 2003 4:15 PM > To:	Skonetski, Susan4 > Subject:	OpenVMS Article on InTech Brazil Magazine >  > Hi Sue >  > Do you remember that Silvia Pereira, from InTech Brazil Magazine, contacted you asking for OpenVMS information? The article on OpenVMS appears on the last issue. I will describe it briefly below.  > C The article talks about VMSs history and quotes two Customers fromiF Petrobras (Brazilian oil company) that were interviewed. They say thatE they run their automation and process control applications on OpenVMSeB since 1990 because it is stable, highly available and secure. TheyB also say that the market trend is to migrate these applications toD Windows, specially because of the OpenVMS costs, but they still have not scheduled any changes.  9> John Egolf was also interviewed. He said that since downtime is too expensive for the industry segment, many Customers (1000) have chosen OpenVMS because of its high availability and scalability. And when the OpenVMS is available on Itanium the OpenVMS solution will be cheaper and more attractive to Customers.e > B The worst thing in the article is the title. It says: "VMS: losing8 space for Windows?". At least there is an question mark!  
 > Regards, >  > Renato Albuquerque.i   ------------------------------  % Date: Tue, 13 May 2003 16:24:12 +0100a+ From: Rok Vidmar <Rok.Vidmar@NUK.Uni-Lj.Si> & Subject: Re: simh emulator and cluster& Message-ID: <3ec11c2e$1@NUK.Uni-Lj.Si>   Hans Vlems wrote:dI > That will work with the simh versions that support ethernet. I booted a M > microVAX 3100-10e from a simh boothost that ran VMS 7.2. DECnet was used as - > the cluster communication protocol carrier.i > But it is not fast.-  ?   DECnet as the cluster communication protocol carrier? Hahaha!g   -- u Regards,  D Rok Vidmar                       Internet:  rok.vidmar@nuk.uni-lj.si; National and University Library  Phone:     +386 1 421 5461R; Turjaska 1, SI-1000 Ljubljana    Fax:       +386 1 421 5464v Slovenia   ------------------------------  + Date: Tue, 13 May 2003 10:36:57 +0100 (MET) 9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com>I$ Subject: Re: Spamfilter for OpenVMS?; Message-ID: <01KVU8QV4WVWAKVGCS@sysdev.deutsche-boerse.com>h  E > > > Will what you are developing provide for a Teergrubing defense,mF > > > or provide callouts for customers to mount their own Teergrubing > > > defense ?  > > C > > As I understand it, that's a function of the SMTP server, whichdC > > V1 of our product won't have.  I also don't quite see the point A > > of these things, other than to have some fun at the spammer's A > > expense---but unless I'm missing something (which is entirelyi@ > > possible, since I'd never heard of this before), I don't see> > > how it's going to make much difference in the overall spam
 > > issue. > B > See  http://www.iks-jena.de/mitarb/lutz/usenet/teergrube.en.html  E For those for whom it's not obvious, a "Teergrube" is a tarpit, like w where the mammoths got trapped.   I This page looks quite interesting.  If I understand things correctly, my uE local, modified SMTP server will receive mail much more slowly if it ,A comes from a blacklisted IP address.  The idea is that if enough iI Teergruben are out there, this will tie up the spammers.  Normal mail is sE not affected.  The occasional legitimate email from a blacklisted IP g address will just take longer.  I The page above mentions some blocks of IP addresses.  What percentage of gC world-wide spam originate from these or other already known blocks?   B Does anyone else find that not all of the links off the page work?   ------------------------------   Date: 13 May 2003 08:32:15 GMT3 From: gartmann@immunbio.mpg.de (Christoph Gartmann)e$ Subject: Re: Spamfilter for OpenVMS?0 Message-ID: <b9qaif$99g$1@n.ruf.uni-freiburg.de>  g In article <9qVva.8482$gI5.40@fe02.atl2.webusenet.com>, Hunter Goatley <goathunter@goatley.com> writes:r >Larry Kilgallen wrote:a >> oJ >> Will what you are developing allow a Reject in cases where the usernameM >> does not exist ?  I understand the RFCs allow for Bounces to be generated,tG >> but those are counterproductive in this age of generally forged spams >> origins.  >i4 >That's one of the things we're planning to do, yes.  I In such an obvious case the SMTP server should reject the mail before the N complete mail has been transfered (e.g. before the last dot). This would avoidH a bunch of legal issues here in Germany and would even prevent a pile of+ messages to handle at the receiving system.a  B >> Will what you are developing provide for a Teergrubing defense,C >> or provide callouts for customers to mount their own Teergrubingr >> defense ? >a@ >As I understand it, that's a function of the SMTP server, which@ >V1 of our product won't have.  I also don't quite see the point> >of these things, other than to have some fun at the spammer's> >expense---but unless I'm missing something (which is entirely= >possible, since I'd never heard of this before), I don't see ; >how it's going to make much difference in the overall spamo >issue.g  N I think a "teergrube" should not be part of a real mailer. I consider it to beI a separate product (of no commercial value): a very slow mailer that willt& - only receive mail and never send one1 - drop all the mail received into the null device-+ - do all this at the lowest possible speed.hO While it would be relatively simple to derive a teergrube from a real mailer itpL would be difficult to incorporate it into a real mailer. A teergrube is veryO efficient if it runs on a host that is not used for mailing otherwise. Any SMTPtN connection to such a system can be considered as spam and therefore be treatedL accordingly. If you want the real mailserver to distinguish between spam andM normal mail and then have it act as a teergrube in case of spam, this is verya difficult to achieve.n   Regards,    Christoph Gartmanne  H -- --------------------------------------------------------------------+H | Max-Planck-Institut fuer      Phone   : +49-761-5108-464   Fax: -452 |H | Immunbiologie                                                        |H | Postfach 1169                 Internet: gartmann@immunbio.mpg.de     |H | D-79011  Freiburg, Germany                                           |H +------------- http://www.immunbio.mpg.de/home/menue.html -------------+   ------------------------------  + Date: Tue, 13 May 2003 11:40:39 +0100 (MET)o9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com>o$ Subject: Re: Spamfilter for OpenVMS?; Message-ID: <01KVUAGVK8CUAKVGCS@sysdev.deutsche-boerse.com>   M > > > Will what you are developing allow a Reject in cases where the usernameeE > > > does not exist ?  I understand the RFCs allow for Bounces to besI > > > generated, but those are counterproductive in this age of generallyy > > > forged spam origins. t > > 7 > > That's one of the things we're planning to do, yes.l > G > In such an obvious case the SMTP server should reject the mail beforeoH > the complete mail has been transfered (e.g. before the last dot). ThisD > would avoid a bunch of legal issues here in Germany and would even@ > prevent a pile of messages to handle at the receiving system.    I agree.  E > > > Will what you are developing provide for a Teergrubing defense, F > > > or provide callouts for customers to mount their own Teergrubing > > > defense ?  > > C > > As I understand it, that's a function of the SMTP server, whichMC > > V1 of our product won't have.  I also don't quite see the point A > > of these things, other than to have some fun at the spammer'ssA > > expense---but unless I'm missing something (which is entirelya@ > > possible, since I'd never heard of this before), I don't see> > > how it's going to make much difference in the overall spam
 > > issue. > J > I think a "teergrube" should not be part of a real mailer. I consider itG > to be a separate product (of no commercial value): a very slow mailere > that will ( > - only receive mail and never send one3 > - drop all the mail received into the null device - > - do all this at the lowest possible speed.dG > While it would be relatively simple to derive a teergrube from a realxF > mailer it would be difficult to incorporate it into a real mailer. AG > teergrube is very efficient if it runs on a host that is not used fore@ > mailing otherwise. Any SMTP connection to such a system can beJ > considered as spam and therefore be treated accordingly. If you want theF > real mailserver to distinguish between spam and normal mail and thenG > have it act as a teergrube in case of spam, this is very difficult tof > achieve. p  F If I understand you correctly, the way to beat spam would be to flood H the internet with email addresses which are in every respect legitimate @ except that the machines which receive mail to them are running I Teergruben instead of normal SMTP mailers.  If this is the case, then as 2' many people as possible should do this.H  I While this will hopefully eventually slow down spam, it won't help folks sI who are flooded by spam at the moment.  These folks probably block stuff  H from certain blocks of IP addresses etc ANYWAY.  Thus, it would be nice G if these same criteria could be used by these folks for turning on the hH teergrube functionality in their SMTP receiver.  Folks who are affected C by spam will probably be even more interested than others in doing o something to combat it.>   ------------------------------    Date: 13 May 2003 05:39:21 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)b$ Subject: Re: Spamfilter for OpenVMS?3 Message-ID: <Ij+w7nlx7vvS@eisner.encompasserve.org>r  w In article <01KVU8QV4WVWAKVGCS@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes:e  K > The page above mentions some blocks of IP addresses.  What percentage of hE > world-wide spam originate from these or other already known blocks?t  E In my experience 95%, since tracking spammy IP addresses is the wholea2 purpose of SpamCop, and it catches 95% of my spam.   ------------------------------    Date: 13 May 2003 05:43:02 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)v$ Subject: Re: Spamfilter for OpenVMS?3 Message-ID: <X6nkMDpg1JAE@eisner.encompasserve.org>e  f In article <b9qaif$99g$1@n.ruf.uni-freiburg.de>, gartmann@immunbio.mpg.de (Christoph Gartmann) writes:  P > I think a "teergrube" should not be part of a real mailer. I consider it to beK > a separate product (of no commercial value): a very slow mailer that willh( > - only receive mail and never send one3 > - drop all the mail received into the null devicen- > - do all this at the lowest possible speed.   D No, only email from specific IP addresses should get the teergrubing
 treatment.  Q > While it would be relatively simple to derive a teergrube from a real mailer itGN > would be difficult to incorporate it into a real mailer. A teergrube is veryQ > efficient if it runs on a host that is not used for mailing otherwise. Any SMTPuP > connection to such a system can be considered as spam and therefore be treatedN > accordingly. If you want the real mailserver to distinguish between spam andO > normal mail and then have it act as a teergrube in case of spam, this is verye > difficult to achieve.t   Not if you believe in DNSbls !  F Of course one must reserve some capacity for processing the real mail,D but that should be easy since the slow response capacity is mainly aC matter of preserving state and using ASTs, rather than using CPU in> great quantities."   ------------------------------   Date: 13 May 2003 10:21:53 GMT3 From: gartmann@immunbio.mpg.de (Christoph Gartmann)o$ Subject: Re: Spamfilter for OpenVMS?0 Message-ID: <b9qh01$brv$1@n.ruf.uni-freiburg.de>  w In article <01KVU8QV4WVWAKVGCS@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes:iJ >This page looks quite interesting.  If I understand things correctly, my F >local, modified SMTP server will receive mail much more slowly if it B >comes from a blacklisted IP address.  The idea is that if enough J >Teergruben are out there, this will tie up the spammers.  Normal mail is F >not affected.  The occasional legitimate email from a blacklisted IP  >address will just take longer.s > J >The page above mentions some blocks of IP addresses.  What percentage of D >world-wide spam originate from these or other already known blocks?  N My impression is that most spam does not originate from blacklisted addresses.L Often spam originates from unprotected relays, from dialup-links, from sitesO that deliver normal mail and spam etc. . Therefore I suggest to put a TeergrubetO onto a machine that has no MX record. Many spammers scan the internet for hosts N that have port 25 (SMTP) open. In this case a Teergrube is easy to set up, hasL no side effects and will force spammers to use real addresses. The real mailJ exchanger should then run an anti-spam software that refuses to accept the' obvious spam and mark the rest as spam.)   Regards,    Christoph Gartmannr  H -- --------------------------------------------------------------------+H | Max-Planck-Institut fuer      Phone   : +49-761-5108-464   Fax: -452 |H | Immunbiologie                                                        |H | Postfach 1169                 Internet: gartmann@immunbio.mpg.de     |H | D-79011  Freiburg, Germany                                           |H +------------- http://www.immunbio.mpg.de/home/menue.html -------------+   ------------------------------   Date: 13 May 2003 10:33:19 GMT3 From: gartmann@immunbio.mpg.de (Christoph Gartmann)-$ Subject: Re: Spamfilter for OpenVMS?0 Message-ID: <b9qhlf$brv$2@n.ruf.uni-freiburg.de>  w In article <01KVUAGVK8CUAKVGCS@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes:   G >If I understand you correctly, the way to beat spam would be to flood  I >the internet with email addresses which are in every respect legitimate :A >except that the machines which receive mail to them are running eJ >Teergruben instead of normal SMTP mailers.  If this is the case, then as ( >many people as possible should do this.  N I didn't think of this possibility, but indeed this is a good idea. Wherever aI web-page asks for an e-mail address and you are not a 100% sure that your.O e-mail address will not be sold or distributed, add an address that points to a 
 Teergrube.  L What I thought about was this: I noticed that we receive a lot of port-scansN for port 25. As soon as you open port 25 and install a mailer you'll reveive aO lot of spam on this port. Thus, spammers scan for smtp-servers and try to placeiN their spam on these mailers, using them as a relay. With a Teergrube installedN you look like an open relay, you look like a spam receiving host but you block the host delivering the spam.   J >While this will hopefully eventually slow down spam, it won't help folks J >who are flooded by spam at the moment.  These folks probably block stuff I >from certain blocks of IP addresses etc ANYWAY.  Thus, it would be nice RH >if these same criteria could be used by these folks for turning on the I >teergrube functionality in their SMTP receiver.  Folks who are affected sD >by spam will probably be even more interested than others in doing  >something to combat it.  O The problem of detecting spam is not trivial. In many cases you need to receivebM the whole message body in order to decide whether the mail is spam or not. InhO case you received the message and confirmed this, you are obliged to deliver itNN (legal reasons). It is quite similar to a standard written letter delivered byL the postal service. So once your Teergrube notices that the incoming mail isO spam, the mail had already been delivered. There is not much use in beeing slowaN sending the last dot. The only possibility would be to reject the mail insteadM of confirming it. This will most likeley cause the spammer to retry it again.-O Now if you have a protocol of the recently rejected spams you could then act as-" a Teergrube from the beginning on.   Regards,    Christoph Gartmanng  H -- --------------------------------------------------------------------+H | Max-Planck-Institut fuer      Phone   : +49-761-5108-464   Fax: -452 |H | Immunbiologie                                                        |H | Postfach 1169                 Internet: gartmann@immunbio.mpg.de     |H | D-79011  Freiburg, Germany                                           |H +------------- http://www.immunbio.mpg.de/home/menue.html -------------+   ------------------------------  + Date: Tue, 13 May 2003 13:29:37 +0100 (MET)S9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com>i$ Subject: Re: Spamfilter for OpenVMS?; Message-ID: <01KVUEKIRUMUAKVGCS@sysdev.deutsche-boerse.com>h  J > > If I understand you correctly, the way to beat spam would be to flood L > > the internet with email addresses which are in every respect legitimate D > > except that the machines which receive mail to them are running M > > Teergruben instead of normal SMTP mailers.  If this is the case, then as k+ > > many people as possible should do this.1 > E > I didn't think of this possibility, but indeed this is a good idea. G > Wherever a web-page asks for an e-mail address and you are not a 100% G > sure that your e-mail address will not be sold or distributed, add anA& > address that points to a Teergrube.   G However, this would be bad for folks LEGITITMATELY requesting an email  G address.  If I understand the Teergrube concept, spamming will stop as TG soon as there are enough Teergruben to trap most of the spammers.  And  I this can be achieved by addresses which are good for nothing except this ?H purpose.  Thus, I would put them on some otherwise useless (but easy to H find) web page, post articles to newsgroups using them etc.  (These two F sources seem to be the main sources of email addresses---perhaps even A more from newsgroups than from the web.  I think that legitimate  7 addresses sold to spammers account for just a minority.-  C > What I thought about was this: I noticed that we receive a lot of J > port-scans for port 25. As soon as you open port 25 and install a mailerD > you'll reveive a lot of spam on this port. Thus, spammers scan forJ > smtp-servers and try to place their spam on these mailers, using them asF > a relay. With a Teergrube installed you look like an open relay, youG > look like a spam receiving host but you block the host delivering then > spam.   C I think someone posted here recently about this idea: turn on SMTP n@ relaying (it is off by default in VMS TCPIP) but stop the queue.  I > The problem of detecting spam is not trivial. In many cases you need toKG > receive the whole message body in order to decide whether the mail is G > spam or not. In case you received the message and confirmed this, youDE > are obliged to deliver it (legal reasons). It is quite similar to alG > standard written letter delivered by the postal service. So once yourMH > Teergrube notices that the incoming mail is spam, the mail had alreadyG > been delivered. There is not much use in beeing slow sending the lastrB > dot. The only possibility would be to reject the mail instead ofE > confirming it. This will most likeley cause the spammer to retry it5F > again. Now if you have a protocol of the recently rejected spams you7 > could then act as a Teergrube from the beginning on.    G What is the problem with rejecting email which comes from IP addresses  E known to be mis-used by spammers?  One of my providers (1&1) doesn't  G relay email from customers unless SMTP authentification is used or the  I From: address is the 1&1 username (something for various reasons I don't  I want to use as an email address).  (I thus started PAYING FOR the use of  F an SMTP relay with dynaccess.de, one of my dynamic-dns providers (the F other is dyndns.org; the former is slightly more expensive but a) can @ host .de domains, b) has an SMTP relay, c) can register domains C (dyndns.org just started doing this very recently) and d) offers a  I technically more advanced service (allowing updates MUCH more often than  E the IP address changes with the goal being to redirect the name to a hG neutral "offline IP" if an expected update doesn't come)---I recommend  G both of these services highly.)  I thus tried to send SMTP mail myself 3E without a relay, but found that I had an IP which was blacklisted at IH various places simply because it was dynamic and COULD be used for spam.G I'm sure that any legitimate user with such an IP address would notice gG this soon (as I did) and take measures (as I did) to allow him to send I mail again.S  H A bit of research on the net seems to indicate that over 90% of spam is E received from blocks of IP addresses which are well known already or pE which become well known very rapidly after abuse sets in.  Much SMTP  I software (I believe even HP TCP/IP Services for OpenVMS) can make use of CD these.  Thus, blocking blocks of IP addresses---via a Teergrube, by C rejecting the mail, by deleting it, whatever---would seem to be an ' effective measure.   ------------------------------  + Date: Tue, 13 May 2003 13:32:47 +0100 (MET) 9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com>g$ Subject: Re: Spamfilter for OpenVMS?; Message-ID: <01KVUF1OQIOCAKVGCS@sysdev.deutsche-boerse.com>   E > My impression is that most spam does not originate from blacklisted9@ > addresses. Often spam originates from unprotected relays, fromD > dialup-links, from sites that deliver normal mail and spam etc. .   I But many of these ARE blacklisted, such as the dynamic IP addresses used b by 1&1.r  F > Therefore I suggest to put a Teergrube onto a machine that has no MXE > record. Many spammers scan the internet for hosts that have port 25v > (SMTP) open. M  ! I can confirm this port scanning..  B > In this case a Teergrube is easy to set up, has no side effects    Right.  
 > and will( > force spammers to use real addresses.   I What do you mean by that?  Force them to send from real addresses (or at 2# least real IPs) rather than relays?o   ------------------------------    Date: 13 May 2003 08:49:57 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen).$ Subject: Re: Spamfilter for OpenVMS?3 Message-ID: <$Di5UibGN4MR@eisner.encompasserve.org>n  f In article <b9qh01$brv$1@n.ruf.uni-freiburg.de>, gartmann@immunbio.mpg.de (Christoph Gartmann) writes:  P > My impression is that most spam does not originate from blacklisted addresses.N > Often spam originates from unprotected relays, from dialup-links, from sitesQ > that deliver normal mail and spam etc. . Therefore I suggest to put a Teergrube Q > onto a machine that has no MX record. Many spammers scan the internet for hostsIP > that have port 25 (SMTP) open. In this case a Teergrube is easy to set up, hasN > no side effects and will force spammers to use real addresses. The real mailL > exchanger should then run an anti-spam software that refuses to accept the) > obvious spam and mark the rest as spam.o  0 In that case, I believe it is called a honeypot.   ------------------------------    Date: 13 May 2003 08:54:05 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)r$ Subject: Re: Spamfilter for OpenVMS?3 Message-ID: <8P9bkOP9PyLi@eisner.encompasserve.org>e  w In article <01KVUEKIRUMUAKVGCS@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes:e  J > A bit of research on the net seems to indicate that over 90% of spam is G > received from blocks of IP addresses which are well known already or yG > which become well known very rapidly after abuse sets in.  Much SMTP dK > software (I believe even HP TCP/IP Services for OpenVMS) can make use of sF > these.  Thus, blocking blocks of IP addresses---via a Teergrube, by E > rejecting the mail, by deleting it, whatever---would seem to be an   > effective measure.  E But of all these, Teergrubing does the best job of tying up resourcesnD used by the spammer, either on their own machine or on an open relay" they usurp for their own purposes.   ------------------------------  % Date: Tue, 13 May 2003 07:27:25 -0700e# From: "Tom Linden" <tom@kednos.com>o$ Subject: RE: Spamfilter for OpenVMS?9 Message-ID: <CIEJLCMNHNNDLLOOGNJICEOMHCAA.tom@kednos.com>u   >-----Original Message----- 5 >From: Larry Kilgallen [mailto:Kilgallen@SpamCop.net]f$ >Sent: Tuesday, May 13, 2003 6:54 AM >To: Info-VAX@Mvb.Saic.Com% >Subject: Re: Spamfilter for OpenVMS?  >  >t< >In article <01KVUEKIRUMUAKVGCS@sysdev.deutsche-boerse.com>,< >Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes: >rJ >> A bit of research on the net seems to indicate that over 90% of spam isG >> received from blocks of IP addresses which are well known already or G >> which become well known very rapidly after abuse sets in.  Much SMTPoK >> software (I believe even HP TCP/IP Services for OpenVMS) can make use ofcF >> these.  Thus, blocking blocks of IP addresses---via a Teergrube, byE >> rejecting the mail, by deleting it, whatever---would seem to be ann >> effective measure.   K From the FAQ at http://www.iks-jena.de/mitarb/lutz/usenet/teergrube.en.htmld  ) How does a teergrube recognize a spammer?SH Currently, the IP address of the remote host is matched against a fixed,I configurable table. A standard entry for AGIS (All you Get Is Spam = ApexsK Global Information Systems/Service) is derived from the Internic ressourcestH containing 204.137.128/18, 204.137.192/19, 205.137.48/18, 205.164.64/17,6 205.254.160/22, 205.254.176/21, 207.142/16, 209.14/16.  , You could also put these IPs in the firewall   >-F >But of all these, Teergrubing does the best job of tying up resourcesE >used by the spammer, either on their own machine or on an open relayi# >they usurp for their own purposes.3  5 It effectively amounts to a reverse denial of servicei >t >---' >Incoming mail is certified Virus Free. ; >Checked by AVG anti-virus system (http://www.grisoft.com).PA >Version: 6.0.480 / Virus Database: 276 - Release Date: 5/12/2003i >E ---I& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).@ Version: 6.0.480 / Virus Database: 276 - Release Date: 5/12/2003   ------------------------------  % Date: Tue, 13 May 2003 10:40:09 -0400o* From: Brian Hechinger <wonko@4amlunch.net>  Subject: Re: Spamfilter for VMS?, Message-ID: <20030513104009.B22751@zill.net>  ; On Tue, May 13, 2003 at 07:27:25AM -0700, Tom Linden wrote:< > + > How does a teergrube recognize a spammer?yJ > Currently, the IP address of the remote host is matched against a fixed,K > configurable table. A standard entry for AGIS (All you Get Is Spam = ApexnM > Global Information Systems/Service) is derived from the Internic ressourceseJ > containing 204.137.128/18, 204.137.192/19, 205.137.48/18, 205.164.64/17,8 > 205.254.160/22, 205.254.176/21, 207.142/16, 209.14/16. > . > You could also put these IPs in the firewall  M the teergrube concept is a great one IMHO.  i am not entirely hot on the ideahM of using a database driven list of IPs to run it however, since these systemsg( have been repeatedly abused in the past.  E check out the following link:   http://www.benzedrine.cx/relaydb.html   I while i don't run OpenBSD to do something like this, it could very easilyuH be modified into powering the teergrube.  this way *YOU* control who you5 block, and you don't have to maintain an list of IPs.7  J just something to consider.  i wouldn't mind seeing the two systems mergedJ so that those of us who don't use OpenBSD/pf can very easily set somethingL like this up (i was going to try and integrate it into my firewall, but it's5 very hard to get to my firewall from the mail server)a  H when i have some free time i'll look into getting a teergrube powered byK something like this.  first round will probably be on unix, but i'd like tor get it working on VMS as well.   -brian --  P "I'd say don't quit your day job, but you're pretty lousy at that too."  -Bender   ------------------------------    Date: 13 May 2003 05:44:56 -0700( From: bob@instantwhip.com (Bob Ceculski)< Subject: Re: Sparky is losing the race Andrew ... and badly!= Message-ID: <d7791aa1.0305130444.68284d37@posting.google.com>f  c "rob kas" <rob@paychoice.com> wrote in message news:<TmTva.131$oA2.65073272@news.netcarrier.net>...r > Bobc > 5 >   Do you get paid to regurgitate Inquirer Articles?e > $ >                                Rob >  >  > 7 > "Bob Ceculski" <bob@instantwhip.com> wrote in message 9 > news:d7791aa1.0305121116.5a682f00@posting.google.com... : > > Sun might as well forget it Andrew ... you should have6 > > bought alpha/vms/tru64 when you had the chance ...; > > NIH (not invented here) syndrome will hurt the company!L > >2, > > http://www.theinquirer.net/?article=9437  1 no, I only regurgitate when I read your posts ...    ------------------------------  % Date: Tue, 13 May 2003 10:33:44 -0400n# From: "rob kas" <bob@paychoice.com>:< Subject: Re: Sparky is losing the race Andrew ... and badly!/ Message-ID: <vc211m846ttpb1@corp.supernews.com>L  4 "> no, I only regurgitate when I read your posts ...  1    You must have been Reading about the IA64 thenO    L     Intel has acknowledged a bug in its Itanium 2 processor. The error showsG up in some rare data sets when a particular sequence of instructions isTK executed. Intel tells us it has shared a test with OEMs and as a workaroundtJ advises concerned Itanium customers to knock the frequency down to 800MHz.  < Taken from http://www.theregister.co.uk/content/3/30669.html   ------------------------------  % Date: Tue, 13 May 2003 16:37:06 -0000h- From: wspencer@ap.nospam.org (Warren Spencer) < Subject: Re: Sparky is losing the race Andrew ... and badly!5 Message-ID: <937A8CA9Awarrenspencer1977@216.168.3.30>t  I bob@paychoice.com (rob kas) wrote in <vc211m846ttpb1@corp.supernews.com>:n  G >    Intel has acknowledged a bug in its Itanium 2 processor. The errorQ >    shows hH >up in some rare data sets when a particular sequence of instructions isA >executed. Intel tells us it has shared a test with OEMs and as aVF >workaround advises concerned Itanium customers to knock the frequency >down to 800MHz. , > = >Taken from http://www.theregister.co.uk/content/3/30669.htmla  J Open question, but likely to Bill Todd:  We've seen a few Intel processor K bugs over the years, but I don't recall ever hearing of an Alpha processor yJ bug.  Is this a hearing problem on my end, or has the Alpha been clean in  all it's iterations?   ws   --     Warren Spencer' Senior Software Engineer (not a writer)  The Associated Press  + ** What's brown and sticky?    A stick.  **    ------------------------------  % Date: Tue, 13 May 2003 11:10:21 -0600L% From: Dan O'Reilly <dano@process.com>n< Subject: Re: Sparky is losing the race Andrew ... and badly!A Message-ID: <5.2.0.9.2.20030513110842.00b92200@raptor.psccos.com>a  , At 10:37 AM 5/13/2003, Warren Spencer wrote:J >bob@paychoice.com (rob kas) wrote in <vc211m846ttpb1@corp.supernews.com>: >sI > >    Intel has acknowledged a bug in its Itanium 2 processor. The error  > >    showsJ > >up in some rare data sets when a particular sequence of instructions isC > >executed. Intel tells us it has shared a test with OEMs and as aoH > >workaround advises concerned Itanium customers to knock the frequency > >down to 800MHz. > > ? > >Taken from http://www.theregister.co.uk/content/3/30669.htmls >kJ >Open question, but likely to Bill Todd:  We've seen a few Intel processorK >bugs over the years, but I don't recall ever hearing of an Alpha processoraJ >bug.  Is this a hearing problem on my end, or has the Alpha been clean in >all it's iterations?c  H If memory serves, early VAX 8600's had a problem where they were missingN an addressing mode for one specific instruction, and if memory serves further,M it was one commonly used by code generated by the COBOL compiler of the time.y   ------J +-------------------------------+----------------------------------------+J | Dan O'Reilly                  |  "There are 10 types of people in this |J | Principal Engineer            |   world: those who understand binary   |J | Process Software              |   and those who don't."                |J | http://www.process.com        |                                        |J +-------------------------------+----------------------------------------+   ------------------------------  % Date: Tue, 13 May 2003 11:07:05 +0300K- From: "Kari Keronen" <kari.keronen@digita.fi>u) Subject: Re: Status of Commserver productr; Message-ID: <VI1wa.2565$M8.1925854@reader1.news.jippii.net>    This was the last version:J DEC COMMserver for OpenVMS        VA    3.0A    6.1   6.2    GCQAA RETIRED  E May work with VMS/Alpha V7.3. I haven't tried. I don't think that theo4 product was sold to any other company, just retired.   -Kari-  * "David Bodey" <dbodey@yahoo.com> kirjoitti' viestiss:3EC0535B.3000901@yahoo.com...G > Does anyone know what happen to the old DEC Commserver software whichyJ > allows a VMS systems to communicate with a Commserver gateway (e.g. to aE > PLC network). Did HP (or a previous incarnation) sell it to anotherg company? > I > We have Commserver V2.0 running on VMS/VAX V7.1 using X25 HDLC LAPB anda' > we want to upgrade to VMS/Alpha V7.3.h >p   ------------------------------  % Date: Tue, 13 May 2003 11:26:43 -0400f! From: Jim Agnew <jpagnew@vcu.edu> 2 Subject: Re: Stopping a que other then a print que' Message-ID: <3EC10EB3.415AF65E@vcu.edu>p  E our perfect disk has a pda command that manages the perfectdisk runs,VG and shut 'em down nicely.. you may really, really, really want to do ita4 that a way instead of hammering the life outta it...   DigiDemon wrote: >  > Hello all! > E > Real quick....we have PerfectDisk running here..and there have been,K > sometimes when I wan't to stop the scheduled que that it starts...and for1H > the life of me I just haven't been able to shut it down.  Anyway I can > nuke this que?  Thanks!a >  > Jamese   -- DF "4,000 years ago I made a mistake."  Elrond Half-Elven, in "Fellowship of the Ring"   ------------------------------  + Date: Tue, 13 May 2003 10:07:34 +0100 (MET)t9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com>n@ Subject: Re: What is the schedule for the DII COE certification?; Message-ID: <01KVU7XG40QQAM3M2Q@sysdev.deutsche-boerse.com>   L > > > Eventually, a 64-bit integer will become too small for some uses. Just > > > trying to plan ahead.r > > D > > I don't think 1280K of RAM would have made that much difference.6 > > Planning ahead requires more than a factor of two. >  > Indeed. To add on to that: >  > 2^32 > 4294967296 > 2^64 > 18446744073709551616 > K > Unsigned 64 bit address space is about 16.77 million terabytes, or 16,384Q > petabytes.  H I think the point was that 1 more bit (64 instead of 63) gives one just  a factor of 2.   ------------------------------  + Date: Tue, 13 May 2003 08:19:57 +0000 (UTC) ) From: Dan Foster <dsf@globalcrossing.net> @ Subject: Re: What is the schedule for the DII COE certification?3 Message-ID: <slrnbc1alk.n6s.dsf@gaia.roc2.gblx.net>   v In article <01KVU7XG40QQAM3M2Q@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> wrote:J > I think the point was that 1 more bit (64 instead of 63) gives one just  > a factor of 2.  B That's technically true, but is meaningless in the overall scheme.  K Assuming future growth is the same linear rate as the last 20 years, you're I still going to end up at about 2 TB for a desktop machine, a far cry fromsH 16.77 million TB. Halve that with a signed integer, that's still 2 TB vs ~8.4 million TB.   -Dan   ------------------------------  % Date: Tue, 13 May 2003 05:58:36 -0400m* From: JF Mezei <jfmezei.spamnot@istop.com>@ Subject: Re: What is the schedule for the DII COE certification?) Message-ID: <3EC0C1AB.C8DFEBBD@istop.com>o   Phillip Helbig wrote:lI > I think the point was that 1 more bit (64 instead of 63) gives one justm > a factor of 2.  K And one more bit in a byte gives the same factor of 2. But make it a signedo0 byte, and the "" is no longer greater than "a".  N If you use a 64 bit register to store 8 characters for some string comparison,M manipulation, then you really don't want the high order bit to be interpreted  as a signed bit.  @ (ok, change "" with "" or "" if you speak germanic language.)   ------------------------------   Date: 13 May 2003 12:53:51 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)@ Subject: Re: What is the schedule for the DII COE certification?5 Message-ID: <b9qpsv$m44id$1@ID-135708.news.dfncis.de>>  3 In article <slrnbc1alk.n6s.dsf@gaia.roc2.gblx.net>,c, 	Dan Foster <dsf@globalcrossing.net> writes: >  > G > Assuming future growth is the same linear rate as the last 20 years, f  G I don't know that it was particularly linear.  I went from 128M to 256MsI to 512M to 1GB in a lot shorter period of time than I went from 4K to 16K G or 16K to 64K.  And then it seemed like ages till I hit 128K. (and thate was only by bank switching!)   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   n   ------------------------------  % Date: Tue, 13 May 2003 08:13:29 -0700i' From: David Mathog <mathog@caltech.edu> @ Subject: Re: What is the schedule for the DII COE certification?8 Message-ID: <20030513081329.050900fb.mathog@caltech.edu>    On Mon, 12 May 2003 16:50:48 GMT: "Fred Kleinsorge" <my-last-name@stardotzko.dec.com> wrote:   > M > The COE work added new functionality to the CRTL to start filling out a lotnG > of the unimplemented, or partially implemented (or worse, incorrectlym# > implemented) UNIX library calls.    B Functionality is one thing - performance is another.  Does DII COE2 (or a separate fix) do anything about the terrible9 performance of pipes on VMS? On a unix system one can do:a  +   program1 | program2 | program4 | program4r  A and it runs very quickly.  On VMS, at least the last time I triedoD this on an Alpha at 7.2-1, the throughput fell off VERY rapidly withB increasing numbers of pipes.  The pipe length -> time(seconds) forB the test case (just an echo program: text in from stdin copied out to stdout) were    0 ->   1 1 ->   1 2 ->  13 3 -> 524  " This was discussed here previously+ three years ago, it's still in google, lookr% for "Pipe throughput, sigh, slow... "-   Regards,   David Mathog mathog@caltech.edu> Manager, Sequence Analysis Facility, Biology Division, Caltech   ------------------------------  % Date: Tue, 13 May 2003 06:16:45 -0400 * From: JF Mezei <jfmezei.spamnot@istop.com>0 Subject: X-windows: XtAppSetExitFlag() missing ?) Message-ID: <3EC0C60D.2AC728AB@istop.com>    VAX VMS 7.2, DECwindows 1.2.5:  M I have been told about the XtAppSetExitFlag() routine which allows a callback-L routine to signal to the XtAppMainLoop() routine that it should exit. (thinkD of it as a $WAKE in an AST to wake up the mainline stuck in $HIBER).  H However, that routine is absent from both the include files and from the" libraries against which one links.  G Does anyone know why this is missing ? Is there another way to tell theeG MainLoop that it can stop ? Or is the only way truly to do all the exitd% processing from inside the callback ?    ------------------------------   End of INFO-VAX 2003.264 ************************