1 INFO-VAX	Tue, 08 Nov 2005	Volume 2005 : Issue 623       Contents:> Re: ?= Where did the Alpha layered products go on 8.2 CD dist?) Re: Announcing Vernon version 2.1 (shark) ) Re: Announcing Vernon version 2.1 (shark) ' Re: Digital ethernet device address AA*  Re: Is VMS V8.3 near ?E Re: Microsoft File Format Conversions (was: Re: Posted under request) ( Re: Newbie question about listing files. OpenVMS workstation for sale Posted under request Re: Posted under request" Re: Printing to HP network printer" Re: Printing to HP network printer Re: RCP issue,- Re: Request for feedback - BACKUP enhancement - Re: Request for feedback - BACKUP enhancement - Re: Request for feedback - BACKUP enhancement - Re: Request for feedback - BACKUP enhancement - Re: Request for feedback - BACKUP enhancement - Re: Request for feedback - BACKUP enhancement - Re: Request for feedback - BACKUP enhancement = Re: Should I have DECWindows running when I boot the OpenVMS? = Re: Should I have DECWindows running when I boot the OpenVMS?   Re: TCPDUMP like utility for LAT% Re: USB SUPPORT ON DS10 ALPHA oPENvms  Re: VMS magtape handling Re: VMS magtape handling Re: VMS magtape handling3 Re: ZFS from Sun, what's about a new FS for OpenVMS 3 Re: ZFS from Sun, what's about a new FS for OpenVMS 3 Re: ZFS from Sun, what's about a new FS for OpenVMS 3 Re: ZFS from Sun, what's about a new FS for OpenVMS 3 Re: ZFS from Sun, what's about a new FS for OpenVMS 3 Re: ZFS from Sun, what's about a new FS for OpenVMS   F ----------------------------------------------------------------------   Date: 7 Nov 2005 19:54:34 +0100 6 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)G Subject: Re: ?= Where did the Alpha layered products go on 8.2 CD dist? * Message-ID: <436fb0fa@news.langstoeger.at>  f In article <aus-F3FC4C.15512107112005@news.cis.dfn.de>, "H. M. Aus" <aus@vim.uni-wuerzburg.de> writes:E >Where do I find the Alpha layered products go on the 8.2 CDROM upd ?   N Not there. They are on ConDist CD sets or on http://www1.aclabs.com (user/pwd)  A >Neither the guide nor the v8.2 layered products CD contains the   >compilers. :|  $ And neither the earlier VMS Cds did.  M Maybe @ earlier/VAX times, a "package" did contain both CD sets (VMS,ConDist) L but the compilers haven't been located on the VMS media (tape or CD) so far.   --   Peter "EPLAN" LANGSTOEGER % Network and OpenVMS system specialist  E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  % Date: Mon, 07 Nov 2005 22:29:31 +0100 3 From: Michael Unger <spam.to.unger@spamgourmet.com> 2 Subject: Re: Announcing Vernon version 2.1 (shark)+ Message-ID: <3t9vcgFs1i9jU1@individual.net>   & On 2005-11-07 18:04, "JF Mezei" wrote:  $ > "Fernando M. Roxo da Motta" wrote:9 >>   Perhaps you can distribute it under a license like :  >>  ( >>    The GNU Free Documentation License3 >>    http://www.gnu.org/licenses/licenses.html#FDL  >>    Free Art License) >>    http://artlibre.org/licence/lal/en/  >>    The Approved Licenses ) >>    http://www.opensource.org/licenses/  >>    Creative Commons! >>    http://creativecommons.org/  >  > $ > Thanks, I will keep those in mind.  F Isn't "Vernon" itself based on prior artwork with an undefined/unknown copyright status?    Michael    --  ; Real names enhance the probability of getting real answers. 5 My e-mail account at DECUS Munich is no longer valid.    ------------------------------  % Date: Mon, 07 Nov 2005 16:45:46 -0500 & From: JF Mezei <jfmezei@vaxination.ca>2 Subject: Re: Announcing Vernon version 2.1 (shark)- Message-ID: <436FCAF4.52769F67@vaxination.ca>    Michael Unger wrote:H > Isn't "Vernon" itself based on prior artwork with an undefined/unknown > copyright status?     E It is based on a clipart gallery. But one could argue that sufficient F editing has been done to it to make it a new work of art. (even thoughC in appreance, it may look similar, the structure of the objects has  greatly changed).    ------------------------------  # Date: Tue, 08 Nov 2005 04:22:09 GMT , From: Hoff Hoffman <hoff-remove-this@hp.com>0 Subject: Re: Digital ethernet device address AA*4 Message-ID: <RJVbf.16002$ZK6.14521@news.cpqcorp.net>   Michael Moroney wrote:1 > JF Mezei <jfmezei.spamnot@teksavvy.com> writes:  >  > 2 >>http://www.iana.org/assignments/ethernet-numbers ... I >>What was AA0000 AA0001 and AA0002 used for ? Are they truly obsolete or I >>was this just some guy who figured they are obsolete because Digital no  >>longer exists ?   G    The AA-00-00-00-00-00 through AA-00-04-FF-FF-FF range was issued to  D DIGITAL by Xerox back when the network technology involved was more @ properly known as Ethernet, and that range then fell within the H locally-administered address space per the then-new IEEE 802.3, and the G of (new) addresses within the range was discontinued.  (This was circa  F 1985, as part of the transition from Ethernet to IEEE 802.3.)  During I the transition to IEEE 802.3, the use of and assignment of new addresses  + within this address range was discontinued.   H    I'd not expect to find much of anything in AA-00-00-00-00-00 through F AA-00-02-FF-FF-FF, per the DIGITAL Ethernet registry documents of the E time.  There are various (circa 1983 vintage) devices with addresses  G assigned within the AA-00-03-00-00-00 through AA-00-04-FF-FF-FF range,  C and most particularly the DECnet Phase IV station addresses reside  3 within AA-00-04-00-00-00 through AA-00-04-00-FF-FF.    ------------------------------  # Date: Tue, 08 Nov 2005 04:31:16 GMT , From: Hoff Hoffman <hoff-remove-this@hp.com> Subject: Re: Is VMS V8.3 near ? 4 Message-ID: <oSVbf.16004$JQ6.15998@news.cpqcorp.net>    Peter 'EPLAN' LANGSTOEGER wrote:L > I just saw the release notes of a V7.3-2 ECO and there is (more than once) > the reference  > F >                a.b.c.d  Release Version of OpenVMS that will contain& >                         this change: > 8 >                Next reelase of OpenVMS Alpha afer V8.3 > / > So, my question is, is V8.3 (very) near now ? K > Otherwise, I don't understand why it can't be fixed in (a coming) V8.3...   >    I will pass along the report of the typos within that text.  $ > Or is this just a(nother) typo ;-)  -    H2CY2006, per current roadmap predictions.    ------------------------------  # Date: Tue, 08 Nov 2005 06:49:30 GMT , From: Hoff Hoffman <hoff-remove-this@hp.com>N Subject: Re: Microsoft File Format Conversions (was: Re: Posted under request)4 Message-ID: <_TXbf.16009$_H6.14808@news.cpqcorp.net>   JF Mezei wrote:  > Goh wrote:  H >>  a)I dont have any tool currently to convert Word/Excel to postscript= >>     - can someone point to me where I can get such a tool,   D    This appears to be centrally a Microsoft application file format D question, and not an OpenVMS operating system question.  (This file I format isn't really native anywhere other than on Windows within Office,   obviously.)   F    I am not aware of any Microsoft Office file format specifications, F which makes this conversion all the more interesting, and rather more  unreliable.       That written...  I    The OpenOffice.org suite (Linux and Windows, but not yet OpenVMS) can  G provide direct export into PDF, and there are tools available that can  E extract text from Word documents including catdoc.  (And you can get  H from PDF to Postscript, obviously.)  Windows can also queue directly to I a network-based printer, or (and I generally prefer to use a direct path  H and not a served path) through a printer local to and served by OpenVMS.  H > Digital had it. But under palmer the software was abandonned then soldD > and now I am not sure. It is called the CDA Converter library. TheI > company that did buy it from DEC supposedly did release some updates to F > support more modern versions.(Can't remember the name of the company > that got the cda converters).   >    http://h71000.www7.hp.com/freeware/freeware60/cdacvtlib022/>    http://h71000.www7.hp.com/freeware/freeware60/decwrite_rtf/   ------------------------------  # Date: Tue, 08 Nov 2005 03:22:04 GMT , From: Hoff Hoffman <hoff-remove-this@hp.com>1 Subject: Re: Newbie question about listing files. 3 Message-ID: <wRUbf.15998$eP6.4265@news.cpqcorp.net>    andyg wrote:H > Is there a way of searching directories for a specific sized files ie. > greater than say 1000 blocks? + > I am trying to search for any big files.  F > what I would like to do is *DIR "disk":[000000...]*.*;*/SIZE=>1000.*      The usual approach is:   '    $ DIRECTORY/SELECT=SIZE=MINIMUM=1000   F    Do take a look at HELP DIRECTORY, or at the DCL Dictionary and the D User's Guide in the manuals, for further details on that mechanism. I Depending on the particular OpenVMS version in use here, you can also go  F looking for large file version numbers, for instance; there have been B various enhancements in this area in more recent OpenVMS releases.  E    Depending on what you are up to, the DFU package available on the  * OpenVMS Freeware can also be quite useful.   ------------------------------   Date: 7 Nov 2005 18:51:49 -0800  From: jafir@msn.com % Subject: OpenVMS workstation for sale C Message-ID: <1131418309.516252.191700@g49g2000cwa.googlegroups.com>   = http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=5827953146    ------------------------------  $ Date: Mon, 7 Nov 2005 20:48:41 -0500, From: "Richard Tomkins" <tomkinsr@istop.com> Subject: Posted under request 7 Message-ID: <1131414680_9433@spool6-east.superfeed.net>   
   Hi Richard,   H   Sorry for dropping in like this but it seemed that I cant pose to both2   INFO-VAX@MVB.SAIC.COM, INFO-VAX@LISTSERV.UGA.EDU7   though I can see the postings coming into my mailbox.   =   Appreciate if you can send my email below to the list on my 	   behalf.   ?   In the past my subscription just got cut off after a month or    so & I have to resubscribe.        Thanks   Goh   *   ========================================     Hi,     9   Currently we have a Windows application which generates @   pdf/Word/Excel reports & we'll be using our corporate standardE   tool ConnectDirect to 'binary' transfer the reports to VMS (without @   converting the reports to  to AlphaVMS server. My manager told>   me that I just have to come up with a solution for VMS users>   to  print the pdf/Word/Excel reports on VMS(Ver 7.3-1) printC   queues on an adhoc basis.  IT/MIS dept is not doing the printing.   2   I just cant get the rationale/politics why NOT :6   a)a solution within the Windows platform is not used;      as we have network HP & Xerox laserprinters across our ;      corporate network (which currently VMS spools to) & it >      should not be too difficult to define print queues within      Windows?   b)the pdf/Excel/Word are converted to postscript first before &      they are being transferred to VMS     Current status is :    ------------------------A   I've got a PdfToPs.exe utility which converts pdf to postscript >   file but someone told me pdf to ps conversion can chew quite?   a lot of CPU & sometimes on adhoc basis, up to 30 pdf reports @   can be created/transferred at a time & thus for the particular<   user to login to a VMS command procedure menu to select anC   option to convert his pdf files followed by printing can take him ?   quite a while.  So basically the user have on his PC screen a ?   browser to "generate" reports on the Windows application & on >   the same PC screen a Reflection VMS session to convert/print?   the reports that have been pushed to his VMS login directory.   :   Right now, I've found some models of Xerox could support>   dcps ($print/parameter=.../type=postscript  postscript_file)!   but Xerox DC405 prints garbage.      Types of help needed :    ------------------------------F   a)I dont have any tool currently to convert Word/Excel to postscriptF      - can someone point to me where I can get such a tool, preferablyF        one that doesn't need compilation/linking, ie ready to use justI        like the pdftops.exe that I've got.  So far I've found the pdftops G        tool retains graphics like company logo/tables after conversion; =        hope that the Word to postscript tool will do likewise   C   b)I heard from an acquaintance that on Tru64 Digital Unix, he can C      spool Word documents directly to a print  (using lpr) on Tru64 E      which points to a Windows print server on the network & the Word A      document will just print nicely (no conversion needed) - can E      anyone shed any light on this.  The acquaintance is reluctant to C      reveal more (probably busy...).  I'm hoping to do the same for       VMS  <   c)if someone can give me a complete list of Xerox Document@      Centre printers that can support postscript, appreciate it.>      Or perhaps I've misconfigured the print queue for DC405??     Thanks very much for any help    Vao           L ---------------------------------------------------------------------------- ---- Do you Yahoo!?0 New and Improved Yahoo! Mail - 1GB free storage!   --( No virus found in this incoming message. Checked by AVG Free Edition.J Version: 7.1.362 / Virus Database: 267.12.8/162 - Release Date: 05/11/2005        O ----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==---- S http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups K ----= East and West-Coast Server Farms - Total Privacy via Encryption =----    ------------------------------  % Date: Mon, 07 Nov 2005 23:37:18 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> ! Subject: Re: Posted under request , Message-ID: <43702B4F.FC8B50D3@teksavvy.com>  
 Goh wrote:C >   I've got a PdfToPs.exe utility which converts pdf to postscript @ >   file but someone told me pdf to ps conversion can chew quiteA >   a lot of CPU & sometimes on adhoc basis, up to 30 pdf reports   G One way around this is to have the VMS menu simply queue a batch job to  convert and then print the job.   G And depending on how your store documents, you might develop a database G of documents and keep the converted postscript so the next time someone G needs to print that document they'll see the postscript version already  exists and just print it.     < >   Right now, I've found some models of Xerox could support@ >   dcps ($print/parameter=.../type=postscript  postscript_file)# >   but Xerox DC405 prints garbage.   G You'll need to be more specific about the garbage/error messages issued H when you try to print. You need to get postcript to work properly first. Also, the proper syntax is :  8 print/para=data_type=postscript  postscript_file etc etc  H >   a)I dont have any tool currently to convert Word/Excel to postscript= >      - can someone point to me where I can get such a tool,   F Digital had it. But under palmer the software was abandonned then soldB and now I am not sure. It is called the CDA Converter library. TheG company that did buy it from DEC supposedly did release some updates to D support more modern versions.(Can't remember the name of the companyC that got the cda converters).  There was also a company selling the M "Keypack converters". Those would also convert from various document formats.   E >   b)I heard from an acquaintance that on Tru64 Digital Unix, he can E >      spool Word documents directly to a print  (using lpr) on Tru64 G >      which points to a Windows print server on the network & the Word > >      document will just print nicely (no conversion needed)   4 VMS has LPR and LPD support with the TCPIP Services.   ------------------------------  $ Date: Tue, 8 Nov 2005 06:47:18 +1100# From: "Gremlin" <not-here@all.mate> + Subject: Re: Printing to HP network printer / Message-ID: <436faf47$1@duster.adelaide.on.net>   M Thanks for that - had a read through and managed to get completely confused.  ' Oh for the simplicity of something like   8 INIT/QUE/IP=192.168.100.111/PORT=9100/SYMBIONT=HPLJ  etc   sigh  5 "John Sauter" <J_Sauter@Empire.Net> wrote in message  2 news:0porm1h245di7t87luqqtr38d7c62lvkt2@4ax.com...B > On Sun, 6 Nov 2005 21:29:44 +1100, "Gremlin" <not-here@all.mate> > wrote: >  >>Hi All >>J >>My mind has become fuzzy over the years - was there ever a way for a VMSL >>print queue to print to, for example, an HP printer that had an IP networkG >>card?  Without installing third party software?  OpenVMS7.3-2, TCP/IP L >>patched etc - don't remember having created queues on remote IP devices to >>do this......  >> > F > Indeed there was, and is.  TCP/IP services provides a print symbiontD > which can be told to print to a particular IP address and port.  IA > have slowly been moving the printers in our shop to interfacing A > directly to the LAN instead of through a PC, and our VAX prints A > on them just fine.  If you have a recent version of VMS you can 9 > find documentation in chapter 24 of the TCP/IP Services 1 > Management manual, available on-line from HP at M > <http://h71000.www7.hp.com/doc/732final/documentation/pdf/aa-lu50m-te.pdf>.  > & >    John Sauter (J_Sauter@Empire.Net) >  >>TIA  >> >    ------------------------------  $ Date: Tue, 8 Nov 2005 08:19:51 +1100# From: "Gremlin" <not-here@all.mate> + Subject: Re: Printing to HP network printer / Message-ID: <436fc4f9$1@duster.adelaide.on.net>   
 Hello Alan  M Thanks for your info - I was reading too much into the steps required - your  % command works a treat - thanks again.     6 "Alan Greig" <greigaln@netscape.net> wrote in message 4 news:ViPbf.4375$%Q.3259@fe2.news.blueyonder.co.uk... >  >  > Gremlin wrote: > F >> Thanks for that - had a read through and managed to get completely 4 >> confused. Oh for the simplicity of something like >>; >> INIT/QUE/IP=192.168.100.111/PORT=9100/SYMBIONT=HPLJ  etc  > H > It is about that simple! Maybe you are reading the wrong section. The . > imaginary command you gave above is actuallyF > $ INIT/QUEUE/on="192.168.100.11:9100"/processor=tcpip$telnetsym (or L > something close to that anyway from memory -check the docs again). If you ) > need moore flexibility investigate DCPS  >  > --  
 > Alan Greig     ------------------------------  % Date: Mon, 07 Nov 2005 20:45:01 -0600 2 From: David J Dachtera <djesys.nospam@comcast.net> Subject: Re: RCP issue, + Message-ID: <4370112C.2C1F13C5@comcast.net>    Fred Jiang wrote:  >  > Hi, J > I encounter a problem while I use RCP command on UNIX server to copy VMS > file to UNIX server.N > Let's say, Host A is a UNIX server , B is a VMS server which run Multinet to > support TCP/IP protocol.K > on B: I create a file ".rhosts" under the default directory of user "aa".  > with content: / > B.ap.aaa.net #(ap.aaa.net is the domain name) < > add line "B.ap.aaa.net" to the file "multinet:hosts.equiv"> > then on UNIX box: I use command "$rcp aa@B:ops:[aa].rhosts .= > Got information :"$ Set NoOn". But there is no file copied. 3 > Is anyone know what's the problem due to? thanks, - > Pls also advise how to resolve the problem?  > On A: remshd is running  > On B: Portman is running.  > thanks in advance 
 > Fred Krond.   B You may want to try posting to the vmsnet.networks.tcp-ip.multinet. newsgroup (cross-posted for your convenience).  C Not sure, but you may need to recompile and install the hosts.equiv 7 table just like the hosts.local table (Multinet quirk).    --   David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------   Date: 7 Nov 2005 10:53:58 -0800  From: "R Boyd" <bob@hax.com>6 Subject: Re: Request for feedback - BACKUP enhancementB Message-ID: <1131389638.825283.23500@o13g2000cwo.googlegroups.com>  E Well how silly of the news software that I am using -- I assumed that D the relevant references were being included when I selected REPLY to Antonio's message.  & Here is the message I was replying to:   24. Antoniov 	Nov 7, 12:16 pm   show options  Newsgroups: comp.os.vms D From: "Antoniov" <anton...@shs-av.it> - Find messages by this author Date: 7 Nov 2005 09:16:37 -0800  Local: Mon, Nov 7 2005 12:16 pm 6 Subject: Re: Request for feedback - BACKUP enhancementE Reply | Reply to Author | Forward | Print | Individual Message | Show  original | Report Abuse    Hi, F I think both features may be interesting for VMS users. But in my mindF it's more important backup should make a sort of snapshot before start it. B Any system administrator 24x7 would like to backup while system is= running and everyone knows how dangerous is /IGNORE=INTERLOCK @ qualifier. I even heard, next OpenVMS version don't support this
 qualifier.   Antonio  http://it.openvms.org  Larry Kilgallen wrote:c > In article <1131384044.504645.52550@g49g2000cwa.googlegroups.com>, "R Boyd" <bob@hax.com> writes: E > > If /IGNORE=INTERLOCK won't be supported, and there is no snapshot C > > service available, VMS backups will become nearly impossible to ! > > accomplish in 24x7x365 shops.  > > " > > Where did you hear this rumor? > E > Your post quoted no context showing anyone talking about any rumor. B > Do not assume that everybody's newsreading software is identical< > to your own, or that they receive posts in the same order.   ------------------------------  $ Date: Mon, 7 Nov 2005 20:59:00 +02007 From: "Guy Peleg" <guy.peleg@remove_this_header@hp.com> 6 Subject: Re: Request for feedback - BACKUP enhancement* Message-ID: <436fa3f9@usenet01.boi.hp.com>  + <briggs@encompasserve.org> wrote in message - news:$3cBWYc7ztd+@eisner.encompasserve.org... 4 > In article <1131381453.181142@smirk>, Alan Frisbie) <Usenet02_REMOVE@Flying-Disk.com> writes:  > > Guy Peleg wrote: > > 8 > >> We are toying with the idea of modifying VMS BACKUP; > >> to support tape striping and tape shadowing, what does 
 > >> it mean?  > > ; > >> Backup of large data stores can be substaintially sped 8 > >> up by writing the save set to multiple tape drives. > >>< > >> Restoring the saveset will require all tapes making the( > >> strip to be loaded into the drives. > > > > > I am concerned about the possible increased risk caused by? > > a damaged tape.   If by "striping", you mean writing pieces > > > of each file on different tapes (like disk striping), then? > > a single bad tape means (I think) that you lose everything.  > > = > > For example, let's say that a disk backup takes two tapes < > > to hold everything.   Half of the files are on the firstB > > tape and half are on the second (ignoring boundary conditions,? > > etc.).   If the second tape is damaged, but my desired file - > > is on the first one, I can restore it OK.  > > < > > If striping causes pieces of my file to be spread across= > > two tapes, then no files from either tape can be restored  > > if one tape is bad.  > >  > > Can you address this fear? > H > Clearly, it can be addressed.  RAID 0+1 or RAID 5 are both applicable. > G > Just write one or more parity tapes in parallel with your data tapes.  > I > Or one could simply realize that this is a niche product and the people F > who need performance may be willing to live with the increased risk.J > If I save my backup on a save set that spans 5 reels, and I lose a reel,G > the backup is, for many purposes, just as useless whether those are 5 ' > reels in sequence or 5 reels striped.  >   H I agree that striping will require (or at least allow) using parity tape   ------------------------------  $ Date: Mon, 7 Nov 2005 21:01:17 +02007 From: "Guy Peleg" <guy.peleg@remove_this_header@hp.com> 6 Subject: Re: Request for feedback - BACKUP enhancement* Message-ID: <436fa481@usenet01.boi.hp.com>  ' "R Boyd" <bob@hax.com> wrote in message < news:1131384044.504645.52550@g49g2000cwa.googlegroups.com...C > If /IGNORE=INTERLOCK won't be supported, and there is no snapshot A > service available, VMS backups will become nearly impossible to  > accomplish in 24x7x365 shops.  > A > Where did you hear this rumor?  This is a dangerous rumor to be 7 > spreading without any verification from good sources.  >  > Robert >   6 I can assure you that next version of VMS will support /ignore=interlock !!!    ------------------------------   Date: 7 NOV 2005 11:19:48 GMT 4 From: karcher@thuria.waisman.wisc.edu (Carl Karcher)6 Subject: Re: Request for feedback - BACKUP enhancement5 Message-ID: <7NOV05.11194864@thuria.waisman.wisc.edu>   O In a previous article, "Guy Peleg" <guy.peleg@remove_this_header@hp.com> wrote:    ->Tape striping  ->------------------------ ->  7 ->Backup of large data stores can be substantially sped 5 ->up by writing the save set to multiple tape drives.  ->  9 ->Restoring the saveset will require all tapes making the % ->strip to be loaded into the drives.  ->   ->What do you think?  ? This one sounds very useful for cutting the backup window time. I As our SAN grows that's always the issue - will the full backups complete L before work starts tomorrow. Cutting that window in half would be a big win.  @ I know that some EBS products do this but we'd rather stick with VMS backup.    --G -- Carl Karcher, Waisman Computing Services, Waisman Center, UW-Madison 7 --                karcher.nomorespam@waisman.wisc.edu      ------------------------------  % Date: Mon, 07 Nov 2005 11:46:01 -0800 # From: "Tom Linden" <tom@kednos.com> 6 Subject: Re: Request for feedback - BACKUP enhancement( Message-ID: <opszvyizd6zgicya@hyrrokkin>  ; I would prefer the effort be put into other media, e.g. DVD  with similar capability.  . On Mon, 7 Nov 2005 12:22:13 +0200, Guy Peleg  , <guy.peleg@remove_this_header@hp.com> wrote:   > Hi folks,  > 5 > We are toying with the idea of modifying VMS BACKUP 8 > to support tape striping and tape shadowing, what does
 > it mean? >  > Tape shadowing:  > ---------------------- > 5 > The ability to write two (or more) copies of a tape  > saveset concurrently.  >  > Tape striping  > ------------------------ > 8 > Backup of large data stores can be substaintially sped5 > up by writing the save set to multiple tape drives.  > 9 > Restoring the saveset will require all tapes making the % > strip to be loaded into the drives.  >  > What do you think? > : > I'm looking for feedback like yes this is a good idea or7 > no it is a waste of time. Please do not ask questions 3 > about implementation as we are in a VERY!!! early : > stage of planning and obviously this is not a commitment > that we'll actually do it....  >  > Thank you for your feedback. >  > Guy Peleg  > OpenVMS engineering  >  >    ------------------------------  $ Date: Mon, 7 Nov 2005 16:28:10 -0500( From: Wayne Sewell <wayne@tachysoft.com>6 Subject: Re: Request for feedback - BACKUP enhancement/ Message-ID: <00A4C740.B4C3E770.3@tachysoft.com>   5 >From: karcher@thuria.waisman.wisc.edu (Carl Karcher)  >X-Newsgroups: comp.os.vms7 >Subject: Re: Request for feedback - BACKUP enhancement  >Date: 7 NOV 2005 11:19:48 GMT> >Organization: Waisman Center, University of Wisconsin-Madison    P >In a previous article, "Guy Peleg" <guy.peleg@remove_this_header@hp.com> wrote: >  >->Tape striping >->------------------------  >-> 8 >->Backup of large data stores can be substantially sped6 >->up by writing the save set to multiple tape drives. >-> : >->Restoring the saveset will require all tapes making the& >->strip to be loaded into the drives. >->  >->What do you think?  > @ >This one sounds very useful for cutting the backup window time.J >As our SAN grows that's always the issue - will the full backups completeM >before work starts tomorrow. Cutting that window in half would be a big win.  > A >I know that some EBS products do this but we'd rather stick with  >VMS backup. >     M My answer to the tape striping issue is the same as my earlier response about L the tape shadowing issue: get tapesys to handle it for you.  Available right& now, and currently running on itanium.  N True, tapesys doesn't have actual tape striping as described above, but it hasN another feature that can increase your backup throughput in a similar fashion: parallel backups.   K Sure, tape striping would help when backing up a single large saveset.  But ? most sites have quite a few savesets that they want to back up.   O The basic problem for a system manager to solve is: "I have a certain number of N gigabytes on a certain number of disks that must all be backed up to tape on aK certain schedule.  I want all these backups to be performed in the shortest  possible time."     I As others have mentioned, and the original poster admitted, fanning out a I single backup saveset to multiple drives greatly complicates the recovery 5 process.  And only one saveset can be done at a time.   M If you have multiple tape drives available, the goal is simply to keep all of > them spinning at full speed until all of the backups are done.  O Having a number of *instances* of vms backup equal to the number of tape drives O will accomplish this as easily as tape striping, and the resultant savesets are M exactly like the savesets we are now using.  When restoring, you only have to H mount the one tape containing the beginning of the saveset.  If it spans* volumes, backup will ask for the next one.  K You have a list of disks to back up.  You break them up into a set of batch * jobs equal to the number of tape drives.    N Just writing the command procedures to do this will allow you to increase your thoughput time.   - The disavantages to this manual approach are:   L 1.  you must keep track of which disks are in which backup job and make sure all of them are accounted for   J 2.  you must do manual load balancing; if one backup job contains only bigM disks and another only small ones, the tape drive assigned to the latter will L finish way early and sit idle for hours while the other keeps grinding; each? backup job should roughly contain about the same amount of data   M 3.  if the backup load for a disk drastically changes (many files added, many 0 files deleted) load rebalancing may be required.    J You can do all this manually, or you can get tapesys and activate parallelN backups.  Say you have 40 disks of varying sizes to back up and 5 tape drives.  H You simply create a system backup file (.sbk) to define the backup job. 8 Specify all 40 disks in whatever order you want and add:   $ parallel_backups == 1  $ n_drives == 5   L Set it to kick off at a particular time via the builtin backup job scheduler of tapesys and that's it.     K When the backup job starts, sysbak will notice that parallel backups are in M effect, and kick off 5 clones of itself, one for each tape drive.  The master I process will then begin assigning disks to back up, using the list in the H parameter file.  Each clone will accept the next assignment, perform theO backup, and then ask for the next in the queue.  All five clones are doing this N in parallel.  The tapes spin continuously.  If a clone gets a short backup, itN is quickly back to ask for another.  If it gets a long backup, it grinds untilK it is finished.  There is no idle time, until *all* backups have ether been L done or are in progress.  As each clone runs out of backups, it terminates. + When all clones are gone, so is the master.     K To me, this seems just as efficient as processing each saveset linearly and A splitting it out to multiple drives.  And a lot less complicated.    Wayne O =============================================================================== N Wayne Sewell, Tachyon Software Consulting  (281)812-0738   wayne@tachysoft.com; http://www.tachysoft.com/www/tachyon.html and wayne.html    O =============================================================================== P Jake Blues:"You traded the Caddy for a microphone? ...... Okay, I can buy that."   ------------------------------  % Date: Mon, 07 Nov 2005 16:21:33 -0600 / From: Chris Scheers <chris@applied-synergy.com> 6 Subject: Re: Request for feedback - BACKUP enhancement2 Message-ID: <436FD36D.1020805@applied-synergy.com>   Guy Peleg wrote: >  > Tape striping  > ------------------------ > 8 > Backup of large data stores can be substaintially sped5 > up by writing the save set to multiple tape drives.  > 9 > Restoring the saveset will require all tapes making the % > strip to be loaded into the drives.  >  > What do you think?  C I find the requirement that all tapes be loaded simultaneously for    restoration to be too dangerous.  F A bad tape drive or a different tape configuration at a DR site could  make the backup useless.  B If the restoration can be done with a subset of the original tape & drives, this feature would be usefull.    > Personally, I would much rather see BACKUP support host based D compression of the data.  That way, time does not need to be wasted - sending blocks of zeros to the backup device.    --  G ----------------------------------------------------------------------- $ Chris Scheers, Applied Synergy, Inc.  B Voice: 817-237-3360            Internet: chris@applied-synergy.com    Fax: 817-237-3074   ------------------------------   Date: 7 Nov 2005 14:20:44 -0800 ( From: "pbritto" <britto.paulo@gmail.com>F Subject: Re: Should I have DECWindows running when I boot the OpenVMS?C Message-ID: <1131402044.872434.225860@g49g2000cwa.googlegroups.com>   C The seller told me it's SN-PBXGI-AA which is a PowerStorm 4D40T PCI D SXGA. I am using a SVGA video, he said he heard rumors of this boardD doesn't work with SVGA videos? bottomline? Is anything wrong between% OpenVMS 7.1-H2 and this graphic card?    ------------------------------  * Date: Mon, 7 Nov 2005 17:01:14 -0600 (CST)* From: sms@antinode.org (Steven M. Schweda)F Subject: Re: Should I have DECWindows running when I boot the OpenVMS?2 Message-ID: <05110717011423_202AAB19@antinode.org>  ( From: "pbritto" <britto.paulo@gmail.com>  E > The seller told me it's SN-PBXGI-AA which is a PowerStorm 4D40T PCI F > SXGA. I am using a SVGA video, he said he heard rumors of this boardF > doesn't work with SVGA videos? bottomline? Is anything wrong between' > OpenVMS 7.1-H2 and this graphic card?   E    It's not one of the supported cards.  Tru64 or Windows, maybe, but 5 not VMS.  Impressive looking thing, though, isn't it?   G    It may not be an official "SN-PBXGI-AA/AB/AC".  Mine has a prominent @ "digital" label on the back with that notation.  It also has theE daughter card near the flange, and memory modules in their slots, but B there are still spots where more chips could have been placed.  It. might be a nice card if it worked (under VMS).  H ------------------------------------------------------------------------  4    Steven M. Schweda               (+1) 651-699-98183    382 South Warwick Street        sms@antinode-org     Saint Paul  MN  55105-2547    ------------------------------  % Date: Mon, 07 Nov 2005 12:25:44 -0800 # From: Joe Bloggs <JBloggs@acme.com> ) Subject: Re: TCPDUMP like utility for LAT 8 Message-ID: <85dvm112c2tulqkoe98p3bltbd17nmeril@4ax.com>  C On Fri, 04 Nov 2005 16:09:53 GMT, VAXman-  @SendSpamHere.ORG wrote:   K >I need to capture some LAT activity at a customer site.  I could have them K >do this for me if there was a TCPDUMP like utility for LAT.  I am not sure J >it will be possible to install any software on these systems, however, to+ >gather this data due to security concerns.  > % >Any "brain storming" ideas welcome.    4 not a brainstorm answer, and likely not easily done:  < but the way I've seen this done at sites, with the corporate4 no-freeware rule, is to put the problem system onto / "mirrored" (monitored) ethernet switch port(s).   9 basically, the (cisco, 3com, etc) switch can be config'ed = copy/duplicate the in/out traffic to from the monitored port  ' to some other unused (watching) port..    9 the idea being that one puts a collector on the watching  @ port.   the collector could be a Network General sniffer ($$$$),* or some other host.  (Alpha, laptop, what)  ; One has to take care, that switch has ample free resources, 6 and/or that the traffic to monitored is not too heavy.  E main trouble with all that, it's likely that a sites w/ a no-freeware A rule, is also quite likely to have similar rules (and even larger 9 reservations against any mucking with network configs ...   A Could be a bit easier,  if the problem system was plugged into a  8 hub (vs a switch) but that's not so common these days.     ------------------------------  # Date: Tue, 08 Nov 2005 03:17:28 GMT , From: Hoff Hoffman <hoff-remove-this@hp.com>. Subject: Re: USB SUPPORT ON DS10 ALPHA oPENvms3 Message-ID: <cNUbf.15997$3Q6.8940@news.cpqcorp.net>     Peter 'EPLAN' LANGSTOEGER wrote:\ > In article <nzb8f.15465$X04.14288@news.cpqcorp.net>, hoff@hp.nospam (Hoff Hoffman) writes: ... e >>In article <435fe5b1@news.langstoeger.at>, peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) writes:  ... = >>:But only on Itanics. No EFI* on Alphas (at least on mine).  >>B >> Load the OpenVMS I64 cross-tools kit, or (when it arrives) load >> the OpenVMS V8.3 release. > 5 > Thanks Hoff. A pointer/URL would have been nice ;-)   G    The OpenVMS Alpha cross-tools kit is a component of the OpenVMS I64  G distribution kit DVD media (though then obviously transfered over onto  E and subsequently installed onto an OpenVMS Alpha configuration), and  G that package and that media is available through the usual HP software  	 channels.    ------------------------------  % Date: Mon, 07 Nov 2005 10:44:23 -0800  From: Z <Z@no.spam> ! Subject: Re: VMS magtape handling ) Message-ID: <7gNbf.2483$4i6.760@fe06.lga>    Jeff Cameron wrote: L > The ANSI specification does state that two consecutive file marks found inB > place of the header file does indeed indicate the end of volume.  H Only means EOT when in place of the header? Ooops. I just realized some + code I wrote 20 years ago at DEC has a bug.    My bad.    ------------------------------    Date: 07 Nov 2005 13:50:14 -05003 From: Rich Alderson <news@alderson.users.panix.com> ! Subject: Re: VMS magtape handling . Message-ID: <mddpspc2syh.fsf@panix5.panix.com>  * Jeff Cameron <roktsci@comcast.net> writes:  O > In an ANSI tape volume, after the Volume header files, each named file stored N > on the tape is actually made up of two physical tape files; first the headerO > file (which describes the attributes of the file), followed by the data file.   > Three:  The header records, the data, and the trailer records.   --  L Rich Alderson                                       | /"\ ASCII ribbon     |L news@alderson.users.panix.com                       | \ / campaign against |L "You get what anybody gets. You get a lifetime."    |  x  HTML mail and    |L                          --Death, of the Endless    | / \ postings         |   ------------------------------  % Date: Mon, 07 Nov 2005 20:41:43 -0600 2 From: David J Dachtera <djesys.nospam@comcast.net>! Subject: Re: VMS magtape handling + Message-ID: <43701067.832BEB37@comcast.net>    Alan Frisbie wrote:  >  > David J Dachtera wrote:  > L > > Two successive tape marks with no interventing data records signify "end > > of tape  > < > Not always true!   If you have an ANSI-formatted mag tape,? > and one of the files is empty (has no records), you will have ; > two tape marks with no intervening data records.   We got = > caught by this when making copies of an old RT-11 DECUS SIG > > tape with exactly this situation.   The "standard" tape copy> > program of the time, TPC, had to be modified (with an option > switch) to handle this.   G Depend on how you're reading it. If you're treating it as file-oriented ? (FILES-11), then your logical flow should be able to handle the " situation if written to expect it.  G If you're treating it as /FOREIGN ("raw") or otherwise unlabelled, then  this would be a problem.   --   David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Mon, 07 Nov 2005 15:56:50 -0500 ( From: Bill Todd <billtodd@metrocast.net>< Subject: Re: ZFS from Sun, what's about a new FS for OpenVMS= Message-ID: <FNKdnYkxd7EKIvLeRVn-iQ@metrocastcablevision.com>    Rudolf Wingert wrote:  > Hello,I > Suns new file system (Zettabyte File System) is a great new innovation.   C Yes, it is - it certainly scooped some of the stuff that I've been  A working on (primarily involving the end-to-end checksums and the  L creation of no-overwrite storage that is not conventionally log-structured).  D > I am missing something like that within OpenVMS. Is there any file* > system greater then 32bit under OpenVMS?  D Yes:  AFAIK the only 32-bit limits in VMS relate to low-level block E interfaces, where 32 bits yields a volume limit of 2 TB (1 TB if the  C value is treated as signed).  Whether multiple such volumes can be  = combined to produce a larger single file system I don't know.   @ RMS-32 has been 64-bit for a decade or so, IIRC - but until the F underlying constraints are lifted it will remain only 40 - 41 bits in 	 practice.      What's the future of OpenVMS, 7 > if we don't get anything like ZFS in the near future?   H Better you should just ask what's the future of VMS, because there's no D chance in hell that VMS will acquire anything like ZFS 'in the near C future'.  Sun has been working hard on ZFS since 2001 (with design  H starting even earlier) - and even if they open-source it along with the I rest of Solaris porting it to VMS would take some considerable amount of  C time (especially since it's currently a local-only implementation,  6 though they plan to extend it to clusters eventually).   - bill   ------------------------------  # Date: Tue, 08 Nov 2005 04:46:57 GMT , From: Hoff Hoffman <hoff-remove-this@hp.com>< Subject: Re: ZFS from Sun, what's about a new FS for OpenVMS3 Message-ID: <55Wbf.16005$QD6.8146@news.cpqcorp.net>    Rudolf Wingert wrote:   I > Suns new file system (Zettabyte File System) is a great new innovation. D > I am missing something like that within OpenVMS. Is there any fileH > system greater then 32bit under OpenVMS? What's the future of OpenVMS,7 > if we don't get anything like ZFS in the near future?  > Best regards R. Wingert   H    OpenVMS Engineering has been asking if you, our customers, want this I over-a-terabyte file system support, and (much to my and various others'  C great surprise) we have not gotten very much interest in this area.   G    OpenVMS Engineering is very well aware of the capacities of current  D and upcoming disk volumes and of the current terabyte limits within G ODS-2 and ODS-5.  If you have an interest in this area and/or in these  E larger disk volumes, I can most certainly forward the requests along.   D    As discussed in the OpenVMS FAQ, the current OpenVMS disk volume & addressing limit is a terabyte volume.   ------------------------------  # Date: Tue, 08 Nov 2005 04:51:57 GMT , From: Hoff Hoffman <hoff-remove-this@hp.com>< Subject: Re: ZFS from Sun, what's about a new FS for OpenVMS3 Message-ID: <N9Wbf.16006$ZQ6.5392@news.cpqcorp.net>    Alan Greig wrote:  > H > A better link is http://www.hpl.hp.com/hpjournal/dtj/vol8num2/toc.htm B > which is direct link to Digital Technical Journal Volume 8 No 2  > discussing Spiralog.]  > H > The official story is that it was cancelled because of insurmountable D > performance problems - the reason the developers gave is that the H > project was cancelled to save money and all staff on the project made K > redundant. All cited performance issues were already being addressed and  I > were expected in the first release. At least that's what the team told  " > me just before they were sacked.  %    Spiralog shipped various releases.   F    The limitations look to be inherent within log file system designs.   ------------------------------  % Date: Mon, 07 Nov 2005 23:45:27 -0600 % From: Dan Foster <usenet@evilphb.org> < Subject: Re: ZFS from Sun, what's about a new FS for OpenVMS5 Message-ID: <slrndn0ern.s0c.usenet@zappy.catbert.org>   a In article <55Wbf.16005$QD6.8146@news.cpqcorp.net>, Hoff Hoffman <hoff-remove-this@hp.com> wrote:  > Rudolf Wingert wrote:  > J >> Suns new file system (Zettabyte File System) is a great new innovation.E >> I am missing something like that within OpenVMS. Is there any file I >> system greater then 32bit under OpenVMS? What's the future of OpenVMS, 8 >> if we don't get anything like ZFS in the near future? >> Best regards R. Wingert > J >    OpenVMS Engineering has been asking if you, our customers, want this K > over-a-terabyte file system support, and (much to my and various others'  E > great surprise) we have not gotten very much interest in this area.  > I >    OpenVMS Engineering is very well aware of the capacities of current  F > and upcoming disk volumes and of the current terabyte limits within I > ODS-2 and ODS-5.  If you have an interest in this area and/or in these  G > larger disk volumes, I can most certainly forward the requests along.  > F >    As discussed in the OpenVMS FAQ, the current OpenVMS disk volume ( > addressing limit is a terabyte volume.  C I might gently suggest that it may be a good idea to start thinking F about concrete steps to deal with these issues now or soon, regardless: of level of direct customer feedback at this present time.  E Reason? Disk capacity is steadily growing, and I believe that OpenVMS C will continue to be, assuming no unforeseen events, a presence 5-10  years from now.   F By then, disk capacity is sure to have reached a size where users will= commonly come up against these limitations with new machines.   H Users may not necessarily fully appreciate the lead time needed to plan,/ design, modify, and qualify such major changes.   H For SCSI, there are '300' GB drives available now. For Serial ATA, thereH are 500 GB drives available now. And, naturally, you easily exceed theseA today with certain setups, such as a stripe-set or RAID-5 volume.    -Dan   ------------------------------  % Date: Tue, 08 Nov 2005 00:48:12 -0500 ( From: Bill Todd <billtodd@metrocast.net>< Subject: Re: ZFS from Sun, what's about a new FS for OpenVMS= Message-ID: <XJ-dnb-Kybq8oe3eRVn-qw@metrocastcablevision.com>    Hoff Hoffman wrote:  > Alan Greig wrote:  >  >>I >> A better link is http://www.hpl.hp.com/hpjournal/dtj/vol8num2/toc.htm  C >> which is direct link to Digital Technical Journal Volume 8 No 2   >> discussing Spiralog.] >>I >> The official story is that it was cancelled because of insurmountable  E >> performance problems - the reason the developers gave is that the  I >> project was cancelled to save money and all staff on the project made  H >> redundant. All cited performance issues were already being addressed I >> and were expected in the first release. At least that's what the team  ( >> told me just before they were sacked. >  > & >   Spiralog shipped various releases. > G >   The limitations look to be inherent within log file system designs.   # That seems just a bit presumptuous.   F IIRC Spiralog was not really a 'log file system' but more like a 'log H block system' transparently shuffled into place beneath ODS-2 (which of A course is not a 'log file system' in any sense of the term).  So  E whatever limitations it had (even if they were indeed 'inherent', an  F interesting point to discuss in its own right if you feel free to get H into details which should no longer have any real competitive relevance @ by now) would not be generic to other potential implementations.  G And to dismiss the more general difficulties of log file system design  ? before even a small percentage of the kinds of effort put into  E conventional file system design have been applied to it seems rather  
 parochial.  F I'm not sold on the log-structured approach, but I certainly wouldn't & dismiss it in such a cavalier fashion.   - bill   ------------------------------  % Date: Tue, 08 Nov 2005 01:30:07 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> < Subject: Re: ZFS from Sun, what's about a new FS for OpenVMS, Message-ID: <437045B9.210E4094@teksavvy.com>   Dan Foster wrote: E > I might gently suggest that it may be a good idea to start thinking H > about concrete steps to deal with these issues now or soon, regardless< > of level of direct customer feedback at this present time.  G I concurr. VMS management must show leadership as well as responding to  cutsomer requests.  B The way it is starting to look now is that VMS has become a customC product tailored to the needs of a relatively small active customer C base. And while it is necessary at this point in time to do what it F takes to retain your remaining customers to keep VMS alive, there mustD also be vision to allow VMS to eventually break out of its remaining niches and be successful again.   D Lack of foresight results in very long term limitations, such as the8 infamous 1gb disk size limit in the ROMs of vaxstations.  D Since ODS5 was designed and implemented for 64 bit platforms and notH ported to VAX, how come it wasn't made fully 64 bit file system from day 1 ?   E Is it because customers weren't yet asking for 64 bit file systems at E the time and VMS management decided there was no extra budget to make  ODS5 fully 64 bit ?   E And now that ODS5 is out, you can't radically change it. So customers H have to wait for ODS-6 or whatever for 64 bit support on a platform that' has existed  for 10 years (64 bit VMS).   D Interesting that VMS was one of the pionneers of 64 bit computing toG allow very large memory, but has yet to adopt a file system that allows  very large files.     G Another example: there might be a HUGE market clamouring for VMS on the G 8086 and HUGE market potential, but VMS management don't dare thread on F this because none of their remaining customers are saying they want to move to the 8086.     H VMS management need VISION. They need to lead.  They are not contractorsE responding to customer requests and only making changes when customer G promises enough revenus to pay for the changes.  They are engineers and D product management whose goal is to position the product to maximise3 sales, product succes and visibility in the market.       D A new file system is no easy task. Microsoft hasn't exactly had much? success with all its hype around a new file system. But the VMS F engineers have far more experience and less "marketing" pressure and IG trust them to design a rock solid file system that can be trusted.  But G if VMS management doesn't give them the go ahead, their hands are tied.   E And it isn't just a question of large files. It is also a question of G handling modern data sources. Basically, VMS should understand URLs the @ same way it understands DECNET file specifications or the way itH understood IBM MVS dataset specifications. And you shoudl be able to SET> FILE/ENTER to an URL. So if you try to type the file, it wouldL automatically invoke the URL protocol (http, ftp, etc) to retrieve the data.  H I am sure the VMS engineers have plenty of ideas of what could be done. ( They just need to be given the go ahead.   ------------------------------   End of INFO-VAX 2005.623 ************************