1 INFO-VAX	Sat, 09 Dec 2006	Volume 2006 : Issue 676       Contents:! Re: Alpha sales extended 6 months * Can you emulate openvms on x86 windows XP?! Re: IEEE Decimal Float on Itanium ! Re: IEEE Decimal Float on Itanium ! Re: IEEE Decimal Float on Itanium  Re: Novice's questions) Re: OT: Outsourcing to offshore locations  Re: recursive copy in VMS  Re: recursive copy in VMS  Re: recursive copy in VMS  Re: recursive copy in VMS  Saved NVRAM/TOY file Re: SSH Attacks  Re: SSH Attacks  Re: SSH Attacks " Re: Standalone vs a regular backup" Re: Standalone vs a regular backup4 Suggestion: use LAT for updating expired VMS license8 Re: Suggestion: use LAT for updating expired VMS license  F ----------------------------------------------------------------------  # Date: Fri, 08 Dec 2006 20:11:29 GMT % From: Rick Jones <rick.jones2@hp.com> * Subject: Re: Alpha sales extended 6 months2 Message-ID: <RFjeh.3410$ZP4.3235@news.cpqcorp.net>  ) Bill Todd <billtodd@metrocast.net> wrote:  > Rick Jones wrote:    > ...   @ >> HP has started releasing benchmark results with the 11iv3 OS: >>  F >> http://www.tpc.org/tpch/results/tpch_result_detail.asp?id=106120401  C > My somewhat vague impression is that TPC-H may say more about I/O C > throughput than about processor (and OS/compiler) performance (or F > about large-system performance under harder-to-partition workloads).  F Unless the system is woefully I/O challenged I supsect that just aboutF everyone will try to put enough I/O in there to take full advantage of	 the CPUs.   B > Still, coming close to the performance of POWER5+ using the sameB > number of cores isn't bad (though it took the Itanic system over+ > twice as long to *load* the database...).   D I'm afraid I don't know enough about TPC-H to know what factors most affect DB load time.  : >> http://www.sap.com/solutions/benchmark/pdf/cert8306.pdf  A > A 4-socket/8-core rx6600 system.  It does score 40% higher than E > earlier similar hardware (2150 vs. 1523), but it's not at all clear B > how much of that came from moving from Windows and SQL Server to, > HP-UX and Oracle (rather than being due toA > earlier-version-to-current-version improvement in HP-UX and the D > compiler per se).  Even the new score is less than 10% higher thanE > HP's recent 4-socket/8-core 2.8 GHz Opteron submission (running the E > Windows/SQL Server software): the latter ran the 2004 vs.  2005 ERP C > code, which might complicate the comparison - but then so did the C > earlier rx6600 submission which scored only 1523 to the Opteron's  > 1978.   > I will leave it as agreeing there are lots of variables there.  B > Maybe we'll get a better idea of V3's relative improvement if HPC > makes another rx6600 TPC-C submission.  Montecito could certainly < > use a boost there: even without being hobbled by SuperdomeF > internals, it's currently scoring only about 2/3 what POWER5+ shouldA > at that level (if POWER5+ scaled down linearly from its 64-core F > performance it would score over 500K tpmC at 8 cores, but IIRC IBM's9 > Rperf figures suggest that it might be more like 550K).   F Well, that all brings us back to vendors not being allowed to estimateB TPC results in public :) So there isn't all that much I can say in	 response.   
 rick jones --  E The computing industry isn't as much a game of "Follow The Leader" as B it is one of "Ring Around the Rosy" or perhaps "Duck Duck Goose." @                                                     - Rick JonesF these opinions are mine, all mine; HP might not want them anyway... :)D feel free to post, OR email to rick.jones2 in hp.com but NOT BOTH...   ------------------------------   Date: 8 Dec 2006 20:52:49 -0800  From: jfredricks@gmail.com3 Subject: Can you emulate openvms on x86 windows XP? C Message-ID: <1165639969.845284.113570@j72g2000cwa.googlegroups.com>   C Is it possible to emulate openvms on x86 windows XP? If so where is " some free or demo software to use?   ------------------------------  % Date: Fri, 08 Dec 2006 15:06:31 -0500 8 From: Stephen Hoffman <Hoff@HoffmanLabs-RemoveThis-.Org>* Subject: Re: IEEE Decimal Float on Itanium( Message-ID: <elcgkb$3ce$1@pyrite.mv.net>   John Reagan wrote: > Stephen Hoffman wrote: >  >>J >>   Something in particular here that might make these operations useful B >> for a particular task?  I'm guessing you're contemplating some D >> specific application here, and probably something involving PL/I. >> > F > Most definitely.  Currency calculations.  Many (and I do mean MANY) C > applications incorrectly use binary floating point to manipulate  I > currency.  They often find out too late that 1.23 + 4.56 may not equal  I > 5.79 (at least exactly).  COBOL on Alpha and I64 have lots of software  C > to emulate that decimal abilities on the VAX.  Having it back in  H > hardware might allow COBOL applications to see a performance increase.    D    Ayup; switched a billing and management system from its (errant) F floating point implementation over to integer calculations eons back.  (One of my first paying jobs.)  I    The changes resulted in massive speed-ups in the calculations and the  I processing, in addition to getting the accounts receivable and the taxes  A right.  The sales tax calculations were the Achilles heel of the  B floating point in the original application, too -- the particular F floating point chosen for the original application didn't have enough - bits to get that right on even small amounts.   H    The bookkeeping department was amazed that all the columns (finally) H added up, too -- probably once they convinced themselves there wasn't a D "salami" technique skimming off fractions in the new version of the 4 application, but I never asked them about that.  :-)  D    Stuffing a decimal into the displays and stripping it out of the B input (character) data was comparatively easier than dealing with 9 floating point rounding and issues of precision/accuracy.   B    The question should better have been "do you have something in G particular here requiring integer values larger than, say, quadwords?"  G And if so, are the numbers of quadword (or larger) operations required  G going to see a particular benefit from the addition of hardware and/or  	 software?   E There are certainly organizations that are dealing with vast sums of  H money, though most also have the budgets needed for customized hardware.   ------------------------------  % Date: Fri, 08 Dec 2006 15:36:00 -0500 0 From: Chip Coldwell <coldwell@frank.harvard.edu>* Subject: Re: IEEE Decimal Float on Itanium. Message-ID: <m3lkliruq7.fsf@frank.harvard.edu>  : Stephen Hoffman <Hoff@HoffmanLabs-RemoveThis-.Org> writes: > H >   Something in particular here that might make these operations useful@ > for a particular task?  I'm guessing you're contemplating someC > specific application here, and probably something involving PL/I.   B Most (good) hand calculators use a decimal radix in their floatingD point implementations.  Users find it annoying when the numbers they? input are rounded to the nearest binary approximation, esp when 5 numbers like 0.1 have no exact binary representation.   A I believe I read somewhere that IEEE 754 doesn't require a binary A radix, although essentially all computer implementations use one. B Certainly, the standard data types (single, double, extended, quad# precision, etc) use a binary radix.    Chip   --   Charles M. "Chip" Coldwell Turn on, Log in, Tune out    ------------------------------  % Date: Fri, 08 Dec 2006 18:55:18 -0800 * From: "Tom Linden" <tom@kednos-remove.com>* Subject: Re: IEEE Decimal Float on Itanium) Message-ID: <op.tj9uegmftte90l@hyrrokkin>   5 On Fri, 08 Dec 2006 12:06:31 -0800, Stephen Hoffman   ) <Hoff@HoffmanLabs-RemoveThis-.Org> wrote:    > John Reagan wrote: >> Stephen Hoffman wrote:  >> >>> L >>>   Something in particular here that might make these operations useful  D >>> for a particular task?  I'm guessing you're contemplating some  E >>> specific application here, and probably something involving PL/I.  >>> I >>  Most definitely.  Currency calculations.  Many (and I do mean MANY)   E >> applications incorrectly use binary floating point to manipulate   K >> currency.  They often find out too late that 1.23 + 4.56 may not equal   K >> 5.79 (at least exactly).  COBOL on Alpha and I64 have lots of software   E >> to emulate that decimal abilities on the VAX.  Having it back in   I >> hardware might allow COBOL applications to see a performance increase.  >  > G >    Ayup; switched a billing and management system from its (errant)   I > floating point implementation over to integer calculations eons back.     > (One of my first paying jobs.) > L >    The changes resulted in massive speed-ups in the calculations and the  L > processing, in addition to getting the accounts receivable and the taxes  D > right.  The sales tax calculations were the Achilles heel of the  E > floating point in the original application, too -- the particular   I > floating point chosen for the original application didn't have enough   / > bits to get that right on even small amounts.  > K >    The bookkeeping department was amazed that all the columns (finally)   K > added up, too -- probably once they convinced themselves there wasn't a   G > "salami" technique skimming off fractions in the new version of the   6 > application, but I never asked them about that.  :-) > G >    Stuffing a decimal into the displays and stripping it out of the   E > input (character) data was comparatively easier than dealing with   ; > floating point rounding and issues of precision/accuracy.  > E >    The question should better have been "do you have something in   J > particular here requiring integer values larger than, say, quadwords?"  J > And if so, are the numbers of quadword (or larger) operations required  J > going to see a particular benefit from the addition of hardware and/or   > software?  > H > There are certainly organizations that are dealing with vast sums of  J > money, though most also have the budgets needed for customized hardware.  H We had a bit of a discussion on this topic in comp.lang.pl1  and I thinkF it was somewhat evenly divided on the value of this.  The group I fellI into was of the opinion that scaled fixed decimal (which on VAX and Alpha J gives us 31 digits) was adequated for the task.    The other group that itK was good because it would obviate  FIXED OVER FLOW EXCEPTIONS, FOFL in PL/I  ecxception handling parlance.   F As it turns out, studying the IBM papers on this, they use an encodingH techniques which gives them two additional digits of accuracy to 33, butG when the computations are performed they convert to packed decimal, bcd F format.  My own feeling is that if they add this as a new data type in@ C, for example, it would make C viable to be used in financial  
 applications.   My god, have we sunk that low:-) >  >  >        --  E Using Opera's revolutionary e-mail client: http://www.opera.com/mail/    ------------------------------  % Date: Fri, 08 Dec 2006 13:58:53 -0500 8 From: Stephen Hoffman <Hoff@HoffmanLabs-RemoveThis-.Org> Subject: Re: Novice's questions ( Message-ID: <elcclf$296$1@pyrite.mv.net>   Bob Koehler wrote:[ > In article <elbg7l$lel$1@south.jnrs.ja.net>, "Richard Brodie" <R.Brodie@rl.ac.uk> writes: @ >> That's the point. On recent versions (6.0 and above), AUTOGEN# >> hardcodes PHYSICALPAGES for you.  > C >    I've never seen any version of AUTOGEN do that to me, from its   >    first ship through VMS 8.3.  I    This issue has listed been in the FAQ for some time.  AUTOGEN can and  @ does indeed set PHYSICALPAGES for you.  There have been various G discussions over the years, and this particular feature/misfeature may  I well have been changed.  (At V6?)  But for quite a few releases, AUTOGEN  H can be "helpful" and can set the parameter for you, limiting the amount I of system physical memory recognized by OpenVMS to what was present when  : AUTOGEN was run; to when AUTOGEN reset this value for you.   ------------------------------  % Date: Fri, 08 Dec 2006 14:08:48 -0500 8 From: Stephen Hoffman <Hoff@HoffmanLabs-RemoveThis-.Org>2 Subject: Re: OT: Outsourcing to offshore locations( Message-ID: <elcd80$2f0$1@pyrite.mv.net>   Ian Miller wrote:  > http://www.dnpg.com/ >  > I don't think they do routers         VNswitch.   ------------------------------   Date: 8 Dec 2006 11:32:40 -0800 $ From: "AEF" <spamsink2001@yahoo.com>" Subject: Re: recursive copy in VMSC Message-ID: <1165606360.557074.219220@n67g2000cwd.googlegroups.com>    Bob Koehler wrote:k > In article <1165431442.259664.229210@80g2000cwy.googlegroups.com>, "AEF" <spamsink2001@yahoo.com> writes:  > >  > > G > > I actually saw something very similar to (2.) happen way back in my K > > graduate student days. A user whose name starts with Z lost some files, E > > including a large FORTRAN file because the system manager "moved" K > > everyone's files from one disk to another and it turned out the last so I > > many files didn't make it due to lack of space or some other reason I J > > can't recall. Of course the user, even after having been urged to backJ > > up his own personal files (as all of us were), failed to do so and the; > > large FORTRAN program had to be rewritten from scratch.  > C >    So no one did backups?  Around here people lose jobs for that.   C We were a small physics research group of 5 or 6 professors, 1 or 2 F postdocs, an average of about 6 grad students, and a researcher (and aE secretary and a part-time travel-booking person). The researcher (not C sure what his official title was) doubled as system manager. It was D agreed-upon policy that each user was responsible for backing up hisF own data. The researcher/system manager would back up the system disk,F and I think some other data, but not user data. (I was there from 1985$ thru 1991 and we used 9-track tape.)   AEF    ------------------------------   Date: 8 Dec 2006 11:47:32 -0800 $ From: "AEF" <spamsink2001@yahoo.com>" Subject: Re: recursive copy in VMSB Message-ID: <1165607252.909734.303760@73g2000cwn.googlegroups.com>   Bob Koehler wrote:d > In article <39egn2tkeerp495rhkaq13kmfjto5hj3tb@4ax.com>, jls <jeffls-nospam@sbcglobal.net> writes: > > G > > Maybe you misunderstand my #2.  #2 says that the file on the Mac is E > > deleted - by accident, on purpose (records retention), or it gets , > > corrupted (and possibly goes unnoticed). > 9 >    But that's why you made a backup in the first place.   < I think the point was that JF likes to mark easy-to-replace,C too-large-to-waste-tape-and-time-on files NOBACKUP. Then, a regular D backup saves only their headers to tape while still fully saving theE remaining important, hard-or-impossible-to-replace files. If the disk E dies, a restore would fully restore the important files, but only the C headers of the easy-to-replace files. Then one has to replace these G garbage-filled files with real data from the original source, which was G alledged to be easy. But then the idea is that if you are simply MOVING F the files from one disk to another, you can use /IGNORE=BACKUP to saveF the trouble of replacing garbage files from the original source. SinceD your move is a definite event as opposed to a perhaps-never restore,F and since moving from disk to disk is easier than copying from disk to tape, this scheme makes sense.  E It's still fun to say (NYC-style), "Do you want deez files backed up, 
 or what?"!  E It's a little like call-waiting back in the dial-up days. An incoming = call would screw up your online session. So the phone company E effectively said, "So, you didn't really want that call waiting after ? all. Well, for a little more money, we'll turn it off for you!"   G The only problem with this scheme I say is that if you have more than a @ few files to mark NOBACKUP, it could be (again, depending on the: circumstances!) quite a chore to clean up after a restore.   AEF    ------------------------------   Date: 8 Dec 2006 14:53:26 -0600 4 From: cornelius@encompasserve.org (George Cornelius)" Subject: Re: recursive copy in VMS3 Message-ID: <OMlZ8mdQXmbA@eisner.encompasserve.org>   i In article <1165607252.909734.303760@73g2000cwn.googlegroups.com>, "AEF" <spamsink2001@yahoo.com> writes: I > The only problem with this scheme I say is that if you have more than a B > few files to mark NOBACKUP, it could be (again, depending on the< > circumstances!) quite a chore to clean up after a restore. >  > AEF   D I'm going to weigh in on this on Alan's side.  If you dig a big holeC in your back yard, someday someone's going to fall into it.  If you E mark it carefully, someone like a neighbor child who can't (or won't) B read will fall into it.  Conclusion: don't dig the hole, of if you do, seal it up.   F I have done what JF does many times, and it is always a minefield. I'mI known as a bit of a packrat for old data, so I have BACKUP savesets lying H about all over the place.  When someone asks for room on a disk I searchD for big files and if one of my savesets is there I move it someplace. that does have space (or copy it off to tape).  E But there's nothing worse than thinking, "Oh, yeah, I have that saved E off on volume XXXX" and then look on XXXX and find someone has had to 6 restore it from a backup and the savesets are useless.  D So over the years I have put the NOBACKUP idea into the list of poor3 practices that should be avoided whenever possible.   I To put things in perspective, my main cluster has nearly 100 shadow sets. F NOBACKUP makes no sense for me. But even in more modest environments I  would avoid it wherever I could.   --9 George Cornelius              cornelius@encompasserve.org 7                               cornelius@SPAMNOTmayo.edu    ------------------------------  % Date: Fri, 08 Dec 2006 21:40:39 +0100 / From: Paul Sture <paul.sture.nospam@hispeed.ch> " Subject: Re: recursive copy in VMSJ Message-ID: <paul.sture.nospam-26CC26.21403908122006@mac.sture.homeip.net>  8 In article <4f8a6$4579b9fe$cef8887a$29412@TEKSAVVY.COM>,/  JF Mezei <jfmezei.spamnot@teksavvy.com> wrote:    > Bob Koehler wrote:E > >    So no one did backups?  Around here people lose jobs for that.  >  > 4 > The question is whether you do backups of backups. > K > If you backup a mac onto a VMS system, do you really need to include the  B > MAC files (which are a backup) when you backup that VMS system ?  I Yes if you want to put them onto tape then free up disk space on the VMS  F system. I did this with my first PC. I didn't have enough space on my F VAXstation 2000 for a full backup of the PC, so did it in chunks. The E speed of a serial connection nicely matched the speed of the TK50 :-)    --  
 Paul Sture   ------------------------------   Date: 8 Dec 2006 12:40:21 -0800 $ From: "Camiel" <iamcamiel@gmail.com> Subject: Saved NVRAM/TOY file B Message-ID: <1165610421.673483.27050@n67g2000cwd.googlegroups.com>   Hello everyone,   B I'm trying to figure out the layout of data in the NVRAM / last 50@ bytes of TOY memory on a DS20. Could anyone help me by using the@ "save_nvram srm" and "save_nvram toy"  commands, and mailing theD resulting files to me? This data from a DS20 would be most useful toG me, but data from another system would be welcome to, as I don't expect  this to have changed a lot.   2 If anyone could help me, I would be most thankful!   Camiel.    ------------------------------   Date: 8 Dec 2006 13:03:31 -0600 - From: Kilgallen@SpamCop.net (Larry Kilgallen)  Subject: Re: SSH Attacks3 Message-ID: <SLqyOTjx53uB@eisner.encompasserve.org>   q In article <G0zBMromwMuX@eisner.encompasserve.org>, koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes: _ > In article <1165582811.782531.97080@80g2000cwy.googlegroups.com>, bob@instantwhip.com writes:  >>  2 >> why worry?  intrusion will take care of him ... >>   > B >    For the lurking newbies, I think he means intrustion evasion.  C I believe the official terms are "intrusion detection" to determine B that attempts are being made and "breakin evasion" to prevent them from being successful.  J >    Yes, it works.  I use it to tell the official scanner folks what theyH >    tried against my systems.  You should see the looks on thier faces.   ------------------------------  % Date: Fri, 08 Dec 2006 14:51:01 -0500 8 From: Stephen Hoffman <Hoff@HoffmanLabs-RemoveThis-.Org> Subject: Re: SSH Attacks( Message-ID: <elcfn9$35u$1@pyrite.mv.net>   Tom Linden wrote: C > Occasionally I see intrusion attempts via ssh.  Knowing the IP of A > offender is there a way to prevent him from trying to log in?       ,    What's the configuration of your network?  D    Do you have the whole of the Internet bouncing directly off your ? OpenVMS host?  (Or any other host system you might care about?)        If so, considered a firewall?     > A teergrube would be great!   H    I'd rather show the electronic equivalent of a hunk of network cable H with nothing connected to it; a so-called smooth wall.   I'd rather let I the nice botnet folks try somewhere else, rather than creating and using  B a tarpit/teergrube.  But if you want a tarpit, here's the classic:  1    http://labrea.sourceforge.net/labrea-info.html   C    LaBrea has various prerequisites.  libpcap and other pieces are  9 around and available for OpenVMS.  I didn't look further.   G    TinyHoneyPot (thp) is written in Perl, and may well work for you --  3 particularly if you have an exposed OpenVMS system.   (    There's a Perl module LaBrea::Tarpit.  G    Most of these tarpits are dedicated network boxes.  All discussions  F of a native port of LaBrea or thp or such aside, it would seem rather F more reasonable to tie up the equivalent of a WRT54GS box as a tarpit G than a VAX, Alpha or Integrity box.  And a tarpitting box tends to run  2 contrary to normal IP network operations on a box.  G    Or to plug (most) access into your SMTP and related services with a  D decent firewall, and a network server.  There are various available H boxes that -- looking strictly at out-of-the-box configurations -- make F better/simpler network servers than does an equivalent out-of-the-box * OpenVMS system.  (Is that heresy?  Donno.)  D    There's also the discussion of whether or not you really want to ? tarpit -- what's the advantage to doing it for you?  Are there  I disadvantages of potentially annoying individuals and organizations that  G have massive collections of computers and bandwidth at their disposal?  F   Per some of the discussions I've read, various of the botnets might F well rival what the NSA has for computes and bandwidth.  Well, unless H the NSA already has its own backdoor into the command and control links  for these same botnets.  :-)      Here are some tools:   # http://www.securityfocus.com/tools/   H    And do take a look around for information on the Blue Security folks " and Blue Frog, and related events:  < http://www.wired.com/news/technology/internet/0,70798-0.html   ------------------------------  * Date: Sat, 9 Dec 2006 03:06:16 +0000 (UTC)7 From: moroney@world.std.spaamtrap.com (Michael Moroney)  Subject: Re: SSH Attacks( Message-ID: <eld978$vqk$1@pcls6.std.com>  : Stephen Hoffman <Hoff@HoffmanLabs-RemoveThis-.Org> writes:   >Tom Linden wrote: >> A teergrube would be great!  I >   I'd rather show the electronic equivalent of a hunk of network cable  I >with nothing connected to it; a so-called smooth wall.   I'd rather let  J >the nice botnet folks try somewhere else, rather than creating and using C >a tarpit/teergrube.  But if you want a tarpit, here's the classic:   2 >   http://labrea.sourceforge.net/labrea-info.html  C I simplified and (heavily) modified Labrea for my own use long ago.   H >   Most of these tarpits are dedicated network boxes.  All discussions G >of a native port of LaBrea or thp or such aside, it would seem rather  G >more reasonable to tie up the equivalent of a WRT54GS box as a tarpit  H >than a VAX, Alpha or Integrity box.  And a tarpitting box tends to run 3 >contrary to normal IP network operations on a box.   H My solution was to use a VMS box with two network adapters.  One was forG everything and ran the normal TCP/IP software, the other was the tarpit I with its builtin TCP/IP stack.  Both were behind a NAT router, with ports K I 'wanted' forwarded to the first network adapter, and the default incoming < (what many NAT documentation calls the 'DMZ') to the tarpit.  I Anyway, after not running the tarpit for quite some time, this thread had I me fire it up to see what's up.  My tarpit saw SSH access attempts from 5 I IP addresses.  Two tried to do something for a little while and just went D away (seemed to just give up, no FIN packet received), two tried to H disconect right away (sent a few repeated FIN packets before giving up),H and one repeatedly connected, gave up, immediately connected again, gaveA up again, for quite some time.  I'm not sure if the last one was  J semi-tarpitted by repeatedly trying to get in, or if it was a try-a-bunch-G of-passwords script, which when one attempt failed, it just went to the & next login/password pair to try again.  ; Anyway, I'm not really seeing any real "stuck" connections.    ------------------------------   Date: 8 Dec 2006 13:42:51 -0800 $ From: "AEF" <spamsink2001@yahoo.com>+ Subject: Re: Standalone vs a regular backup C Message-ID: <1165614171.081141.134430@l12g2000cwl.googlegroups.com>   / Phillip Helbig---remove CLOTHES to reply wrote: H > In article <GL-dnfVivsc1FuTYnZ2dnUVZ_vudnZ2d@comcast.com>, "Richard B.+ > Gilbert" <rgilbert88@comcast.net> writes:  > L > > Your daily or weekly backups of the system disk should also be made withI > > /IMAGE if you expect to be able to boot the restored disk.  /IMAGE is ' > > not necessary for non-system disks.  > I > I would say that /IMAGE is needed if you want an image.  While problems H > might be noticed less quickly on a non-system disk, they will still beH > there, in particular the way alias entries are handled.  OK, you mightC > not have them on a non-system disk, but it is certainly possible.   A You should definitely use /IMAGE for full backups to tape. Why do D anything else? You risk missing things with [*...], [000000...], and such.    ------------------------------   Date: 8 Dec 2006 15:35:31 -0800 # From: "H Vlems" <hvlems@freenet.de> + Subject: Re: Standalone vs a regular backup B Message-ID: <1165620931.512366.308610@79g2000cws.googlegroups.com>   Richard B. Gilbert schreef:   ! > tomarsin2015@comcast.net wrote:  > > > > I thought that a standalone backup is a image copy where aH > > "weekly/daily" backup is just a snap-shot of the disk. I recall thatK > > the queue manager and some other files will not work.correctly when you C > > restore from a weekly. Anyway I will give it a try and see what  > > happens. > > phil > >  > > JF Mezei wrote:  > <snip> > H > A standalone backup is just a backup made without the O/S running!  ItF > is an image backup only if you include the /IMAGE qualifier.   You'dD > bloody well better include it if you expect to be able to boot the > restored disk!!!!  > J > Your daily or weekly backups of the system disk should also be made withG > /IMAGE if you expect to be able to boot the restored disk.  /IMAGE is % > not necessary for non-system disks.  > 4 > As a matter of habit I make all my backups /IMAGE.  F An image backup saves (and restores) the file- and directory structureF of a disk as well as a couple of parameters of the volume itself (e.g.F the disk's label). More importantly, security information is stored asA well as backlinks (remember SET FILE/ENTER). You wouldn't want to  repair that manually, right?G Whether a disk is a systemdisk or a datadisk is imho irrelevant. If you C want to replace and restore a bad disk then you'd obviously want to D have the old label back as well as other attributest that might haveE been altered while INIT-ing the disk or had been subsequently changed  with the SET VOLUME command.  G Standalone backup makes IMAGE backups, can't remember whether /PHYSICAL @ is even an option with SAB but those who feel they really need aG physical backup do that for a very specific purpose, usually just once.   @ If the " week backup"  is not an image saveset then it cannot beC restored with SAB. Then again, chances are slim that such a saveset F would build a functional systemdisk. No owner information nor securityG attributes will be restored so the resulting system is unlikely to be a  copy of  what was lost.    HansG Omit /IMAGE and you need to specify the directory or directorytree that 0 needs backing up. Owner information is not store   ------------------------------   Date: 8 Dec 2006 13:16:47 -0600 4 From: cornelius@encompasserve.org (George Cornelius)= Subject: Suggestion: use LAT for updating expired VMS license 3 Message-ID: <CCxbl3iqMo3h@eisner.encompasserve.org>   ? An earlier thread discussed getting licenses into a system with B expired licenses without setting the date back, but I see that one7 approach that has worked well for me was not mentioned.   E My trick for loading licenses across the network from a VMS system to C another system whose licenses (hobbyist, CSLG, etc.) are expired is 8 to use $ SET HOST/LAT, which does not require a license.  F Steps (local node needs PAKs; target node has them installed already):  @  1. Log in from console (allowed without licenses) on local node  G  2. Start LAT on both nodes (may require editing LAT$SYSTARTUP.COM). Be D     sure to enable "LAT master" (outgoing LAT to remote host) on the     local node.   4  3. Using $ SET HOST/LAT/LOG , log in to target node  N  4. $ LICENSE ISSUE VMS-USER/PROC/OUT=VMS-USER.PAK ! Or whatever PAK is needed?     $ LICENSE ENABLE VMS-USER ! Reenable - prior op disables it      $ TYPE VMS-USER.PAK   >  5. Repeat for other PAKs as needed (DECnet, TCPIP, whatever).  A  6. Logout, then edit logfile SETHOST_LAT.LOG to extract just the       output of the TYPE commands.  G  7. Execute to install the PAKs.  Remaining PAKs can be installed after      starting the network.    --9 George Cornelius              cornelius@encompasserve.org 7                               cornelius@SPAMNOTmayo.edu    ------------------------------  % Date: Fri, 08 Dec 2006 21:55:18 +0100 / From: Paul Sture <paul.sture.nospam@hispeed.ch> A Subject: Re: Suggestion: use LAT for updating expired VMS license J Message-ID: <paul.sture.nospam-859041.21551808122006@mac.sture.homeip.net>  3 In article <CCxbl3iqMo3h@eisner.encompasserve.org>, 6  cornelius@encompasserve.org (George Cornelius) wrote:  A > An earlier thread discussed getting licenses into a system with D > expired licenses without setting the date back, but I see that one9 > approach that has worked well for me was not mentioned.  > G > My trick for loading licenses across the network from a VMS system to E > another system whose licenses (hobbyist, CSLG, etc.) are expired is : > to use $ SET HOST/LAT, which does not require a license. > H > Steps (local node needs PAKs; target node has them installed already): > B >  1. Log in from console (allowed without licenses) on local node > I >  2. Start LAT on both nodes (may require editing LAT$SYSTARTUP.COM). Be F >     sure to enable "LAT master" (outgoing LAT to remote host) on the >     local node.  > 6 >  3. Using $ SET HOST/LAT/LOG , log in to target node > P >  4. $ LICENSE ISSUE VMS-USER/PROC/OUT=VMS-USER.PAK ! Or whatever PAK is neededA >     $ LICENSE ENABLE VMS-USER ! Reenable - prior op disables it  >     $ TYPE VMS-USER.PAK  > @ >  5. Repeat for other PAKs as needed (DECnet, TCPIP, whatever). > C >  6. Logout, then edit logfile SETHOST_LAT.LOG to extract just the " >     output of the TYPE commands. > I >  7. Execute to install the PAKs.  Remaining PAKs can be installed after  >     starting the network.  >   E An interesting approach thanks, especially since LAT is available on   Linux:  , http://linux-decnet.sourceforge.net/lat.html   and Windows via Kermit95:   ' http://www.columbia.edu/kermit/k95.html    --  
 Paul Sture   ------------------------------   End of INFO-VAX 2006.676 ************************                                                                                                                                                                                                                                                                                                                                                                                                                                      