1 INFO-VAX	Sun, 24 Dec 2000	Volume 2000 : Issue 717       Contents:/ Re: Any tips and/or tricks for VMS with HSG80s?  BASIC License Question Re: BASIC License Question Data Calculation( Re: Happy Holidays - From Rich Marcelllo( Re: Happy Holidays - From Rich Marcelllo My Hardware WishList OpenVMS Listserv Software  RE: OpenVMS Listserv Software  Re: Opera (was Mozilla 0.6)  Re: Proxies for remote access  Re: system disk space & Re: when is SYS$NODE normally defined?  F ----------------------------------------------------------------------  # Date: Sun, 24 Dec 2000 17:08:31 GMT % From: Uwe Zessin <zessin@my-deja.com> 8 Subject: Re: Any tips and/or tricks for VMS with HSG80s?) Message-ID: <925aie$5c7$1@nnrp1.deja.com>   ' In article <3A45EF6B.6775C6FA@home.nl>, !   Dirk Munk <munk@home.nl> wrote:  > Hi Kerry,  > E > Thanks very much for your suggestions. Further on in this message I  > will reply to them.  > 
 > Regards, >  > Dirk >  > "Main, Kerry" wrote: > >  > > Hello Dirk,  > > ? > > Some good tips, but a few suggestions for consideration  ..  > > E > > >> You have to set the desired path on the HSG80 (I think), or on / > > VMS (you may have 8 paths to the unit !)<<<  > > ? > > In a redundant config (2 HBA's per server, 2 FC switchs and < > > 2 HSG80's), OpenVMS should find 4 connections per device > > automatically.  7 I beleive Dirk was talking about a fully-meshed fabric.   C > > Recommend that you do not set the "preferred path" on the HSG80 B > > as it will be overridden anyway, since OpenVMS will choose theE > > first path it finds as the initial preferred path for the device.  > H > I know about and use the possibility to set the preffered path in VMS,F > and was considering trying it on the HSG80. But if VMS overrides theE > HSG80 settings, then I don't have to try. Furthermore the HSG80 CLI D > manual is a bit confusing about this subject. On page 2-31 it saysC > "This qualifier is effective only in multiple-bus failover mode". C > On page 2-131 it says about the same switch "The controllers only @ > use the PREFERRED_PATH setting if they are in a dual redundant > configuration". & > So the manual contradicts itself....  ? There are many more (e.g. on memory requirements for snapshot).   H > The manual is not very clear about another situation too. It states onH > page 2-29 that if the controllers are in transparent failover mode (soE > not applicable to VMS), units D0 to D99 are assigned to port 1, and C > units D100 to D199 are assigned to port 2. But it does not say on F > which controller, or are the units visible on both controllers? ThatH > would mean that unit D10 is visible on port 1 of both controllers, but > I doubt it. Very confusing.   B No. That's correct. In dual-redundant (=transparent failover) modeC PORT_1 of controller A is active and PORT_2 is standby while PORT_1 C of controller B is standby, PORT_2 is active. Should one controller C fail, the other takes over - WARNING! A bad cable will not cause an 4 automatic failover - at least I have never seen one.  ? You have only 2 visible WWNs (for PORT_1 and PORT_2) and during F failover it will move from PORT_1/Controller_A to PORT_1/Controller_B). or PORT_2/Controller_B to PORT_2/Controller_A.  C Now about the units. The 'trick' is that in dual-redundant mode all @ connections to PORT_2 (remember that a CONNECTION defines a pathD from a host bus adapter to a particular controller/port combination)' automatically get a unit-offset of 100.   B Let's assume all 4 ports are connected to the same switch. 2 PortsA have 'link up' and 2 are 'down'. Also, you will see 2 connections  per HBA, not 4 on your HSG80s.  @ You don't want this mode with OpenVMS (or any other OS) that hasC multi-path capability (or at least some software like Secure Path).   D > > Ideal situation is to leave the default as "nopreferred path" onA > > all devices on the HSG80. However, before the disk drives are @ > > mounted from OpenVMS, run a command procedure which does the@ > > following for each device. This sets the preferred path that( > > you want for optimum load balancing. > > ; > > $ set device $1$DGA34: /path="PKA0.2343-A855-BC99-876F"  > > > actualy: set device $1$DGA34: /path=PGA0.2343-A855-BC99-876F > (device PG, and no "")  > I have seen some text about using or leaving off the '"'s, but I don't remember it right now.  + > > $ help set device /path (for more info) - > > $ help show device /multi (for more info)  > > ! > > Also, see this web reference: F > > <http://www.openvms.compaq.com/openvms/fibre/doc/6318pro_004.html# setup_hsg80_h> > > F > > Suggestion for optimum config, is that you want to try and get allF > > RAID devices created in one group to be mounted via the same HSG80 > > controllerD > > e.g. if S1 is a stripe set of 3 mirror sets M1, M2, M3 and S1 isB > > partitioned into 4 drives $1$DGA34/35/36/37, then ideally, youD > > want the preferred path to be the same for all 4 drives and that$ > > path will be via the same HSG80. > F > And even the same port on that HSG80. ("Partitioned units can not be# > split accross ports" - page 2-29)   G I vaguely recall that partitioned storage sets cannot even split across @ different hosts, although that might be a non-VMS OS limitation.  F > > >> Set the allocation class on the controller, even if VMS doesn't > > use it (yet).<<  > > B > > Not sure if I understand this tip. The main reason for setting? > > a device allocation class is to ensure a device is uniquely  > > identified in a cluster.C > > I have not heard of any future requirement for allocation class D > > to be used by OpenVMS with HSG80's (not saying that there is notC > > a requirement, just that I have never even it being discussed).  > > D > > As you stated, OpenVMS does not use the allocation parameter forC > > HSG80's and SAN's, so I am not sure what this would gain ...the D > > down side could be added confusion when alloc parameters are setB > > on the HSG80's but these are not reflected back in the OpenVMS< > > device numbering scheme, so it might confuse some folks.  C It is confusing! Do a '>>> show device' after you have run WWIDMGR. C I have done some training for a customer, I had just configured the B boot support and his people got a little bit nervous when they saw! all these 'strange device names'.   D > I can imagine this tip looks a bit strange at first. The HSG80 CLIH > manual states op page 2-121 that the allocation class number should beF > set to a unique value for every HSG80 pair. Apparently in an earlierG > stage of development of the SAN infrastructure VMS was suppose to use E > this allocation class number. I don't know if the present situation D > where VMS does not use it is a permanent situation, or that it mayC > change in the future. Furthermore I don't know if other operating H > systems like to see unique allocation class numbers or not. So to keepA > on the save side I give the HSG80's a unique allocation number.  > E > And I would very much like it if VMS could use the allocation class E > number. I have a situation where 4 large computer sites are located H > miles from each other. I would like to use the allocation class numberH > to determine where the HSG80 is located. That way I can design commandG > procedures that will set the appropiate shadow settings for shadowing @ > over 2 sites in disaster tolerant cluster setups with cascaded: >  switches. For now I have designed the following scheme: > D > Every site gets a range of numbers, 2-19, 20-39, 40-59, 60-79, and > 80-99 (spare).? > Every controller pair gets a unique number from these ranges. H > This number is used as allocation class number, CCL identifier, and in* > the first digits of the unit identifier. > 
 > Example:E > the controller pair has allocation class 14, CCL identifier 14, and . > units with identifiers 1401, 1402, 1403 etc. > A > But this way I can 'only' install about 20 HSG80 pairs on every D > computer site, and that is not so much. I was hoping that with theD > proper use of the allocation class numbers I would be able to makeA > a scheme where I can install something like 1000 HSG80 pairs on E > every site (in the scheme that is, so no need to send Rich Marcello F > over here to sign multi-million $$$ contracts -:)) ). Of course thisG > only works if the allocation class is added to the identifier number, : > because identifiers have a maximum value of 'only' 9999.  : I wonder what that would do to you in a DRM environment...   > > Regards, > >  > > Kerry Main > > Senior Consultant  > > Compaq Canada Inc. > > Professional Services  > > Voice: 613-592-4660  > > Fax  :  819-772-7036  > > Email: Kerry.Main@Compaq.com > >  > > -----Original Message-----) > > From: Dirk Munk [mailto:munk@home.nl] # > > Sent: December 23, 2000 3:40 AM  > > To: Info-VAX@Mvb.Saic.Com < > > Subject: Re: Any tips and/or tricks for VMS with HSG80s? > >  > > Some tips: [...]    --
 Uwe Zessin3 (If you want to send mail, please use user "zessin" / who lives at "decus.decus.de", not my-deja.com)    -- --
 Uwe Zessin3 (If you want to send mail, please use user "zessin" / who lives at "decus.decus.de", not my-deja.com)      Sent via Deja.com  http://www.deja.com/   ------------------------------  # Date: Sun, 24 Dec 2000 13:42:43 GMT ) From: "Neil Rieck" <n.rieck@sympatico.ca>  Subject: BASIC License Question < Message-ID: <n3n16.115272$J32.4320000@news20.bellglobal.com>  J I just (00.12.22) did an upgrade from "VAX-BASIC 3.8" to "Compaq-BASIC forG VAX 3.9" and now I get a license too old message when ever the complier H starts. I reinstalled the old compiler and everything is back to normal.  9 Has anyone else ever seen this "license too old" message?   
 Neil Rieck Bell-ICS, Bell Canada.! http://www3.sympatico.ca/n.rieck/    ------------------------------  % Date: Sun, 24 Dec 2000 16:35:39 +0100   From: Paul Sture <paul@sture.ch># Subject: Re: BASIC License Question + Message-ID: <VA.000001f6.348bd5a0@sture.ch>   H In article <n3n16.115272$J32.4320000@news20.bellglobal.com>, Neil Rieck  wrote:+ > From: "Neil Rieck" <n.rieck@sympatico.ca>  > Newsgroups: comp.os.vms ! > Subject: BASIC License Question % > Date: Sun, 24 Dec 2000 13:42:43 GMT  > L > I just (00.12.22) did an upgrade from "VAX-BASIC 3.8" to "Compaq-BASIC forI > VAX 3.9" and now I get a license too old message when ever the complier J > starts. I reinstalled the old compiler and everything is back to normal. > ; > Has anyone else ever seen this "license too old" message?  >  Yes. From the help:    LICENSE   
   REGISTER       /RELEASE_DATE              /RELEASE_DATE=date  H        Specifies a product release date such that the license authorizesE        use of all product versions released on or before the date. If C        your PAK supplies a product release date, you must enter the G        information exactly as it appears. The date must be presented in 0        the standard OpenVMS format: dd-mmm-yyyy.      ___ 
 Paul Sture Switzerland    ------------------------------  % Date: Sat, 23 Dec 2000 00:07:52 -0600 % From: "David Ash" <jimjj@hotmail.com>  Subject: Data Calculation / Message-ID: <t48g9orsp33889@news.supernews.com>   E Do any of you have a programming example, preferably in DCL or C that B calculates the difference in number of days between any two dates?   ------------------------------   Date: 24 Dec 2000 16:25:45 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon) 1 Subject: Re: Happy Holidays - From Rich Marcelllo , Message-ID: <925829$18lo$2@info.cs.uofs.edu>  0 In article <87y9x6fta1.fsf@k9.prep.synonet.com>,/  Paul Repacholi <prep@prep.synonet.com> writes: 2 |> JF Mezei <jfmezei.spamnot@videotron.ca> writes: |>  N |> > At the next teleconference, you get another Wall Street Casino Analyst toK |> > enquire about a rumour that Compaq might spin off the VMS group into a P |> > separate company that would not have any barriers to marketing its own more# |> > robust alternative to Windows.  |>  A |> Now what would you call you spin-off company that sell digital  |> computers? ;)  $ I've already put in mny vote:    DEC   |>  5 |> But then, aren't all computer Digital computers...   A Not necessarily.  There was a rather lengthy discussion of Analog A Computers in the PDP-11 newsgroup and it turned out the one in my   attic is not really unique.  :-)   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>       ------------------------------   Date: 24 Dec 2000 16:19:11 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon) 1 Subject: Re: Happy Holidays - From Rich Marcelllo , Message-ID: <9257lv$18lo$1@info.cs.uofs.edu>  , In article <3A44EEB6.1081EE7B@videotron.ca>,0  JF Mezei <jfmezei.spamnot@videotron.ca> writes: |>  L |> At the next teleconference, you get another Wall Street Casino Analyst toI |> enquire about a rumour that Compaq might spin off the VMS group into a N |> separate company that would not have any barriers to marketing its own more! |> robust alternative to Windows.   F Hmmmm.  Wonder what the chances are that the spin-off company could be4 called DEC.                                      :-)   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------  % Date: Sun, 24 Dec 2000 12:21:12 -0600 7 From: "David J. Dachtera" <djesys.nospam@earthlink.net>  Subject: My Hardware WishList - Message-ID: <3A463E98.78F5BCF4@earthlink.net>   - (cross-posted to comp.sys.intel, comp.os.vms)   
 Hi, Folks,  H I normally hang out on comp.os.vms where they keep insisting that VMS on@ Intel won't be possible until these systems provide 255 vectored interrupts.   F At one time, there were Alpha systems that supported NT, VMS and more.  G So, I guess that would be my wish for the New Millenium (now that we're G REALLY on the threshold of it, unlike last year at this time). An Intel D mobo that provides support for Linux, *BSD, Micro$hit *AND* OpenVMS.F Then, the only remaining obstacles would relate to processor modes and> such - and those can be overcome, just as they did with Alpha.   --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/   @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.    ------------------------------  # Date: Sun, 24 Dec 2000 18:19:04 GMT 3 From: skidmore@alternate-venue.org (Barry Skidmore) " Subject: OpenVMS Listserv Software0 Message-ID: <3a463de1.167702301@news.iquest.net>  ? Is anyone aware of listserver software that runs under OpenVMS?    Barry    ------------------------------  % Date: Sun, 24 Dec 2000 12:38:35 -0600 + From: "Main, Kerry" <Kerry.Main@compaq.com> & Subject: RE: OpenVMS Listserv SoftwareN Message-ID: <910612C07BCAD1119AF40000F86AF0D805284B36@kaoexc3.kao.cpqcorp.net>   Barry,  C re: Is anyone aware of listserver software that runs under OpenVMS?   
 Check out: <www.lsoft.com> D <http://www.openvms.compaq.com/openvms/brochures/l-soft/> (Reference	 brochure)    Regards,  
 Kerry Main Senior Consultant  Compaq Canada Inc. Professional Services  Voice: 613-592-4660  Fax  :  819-772-7036 Email: Kerry.Main@Compaq.com     -----Original Message-----H From: skidmore@alternate-venue.org [mailto:skidmore@alternate-venue.org] Sent: December 24, 2000 1:19 PM  To: Info-VAX@Mvb.Saic.Com " Subject: OpenVMS Listserv Software    ? Is anyone aware of listserver software that runs under OpenVMS?    Barry    ------------------------------  % Date: Sun, 24 Dec 2000 10:53:33 +0100   From: Paul Sture <paul@sture.ch>$ Subject: Re: Opera (was Mozilla 0.6)+ Message-ID: <VA.000001f5.3352a0be@sture.ch>   K In article <MPG.14aeead975d8a6c49896ab@news.bellatlantic.net>, John Santos   wrote:- > From: John Santos <john.santos@verizon.net>  > Newsgroups: comp.os.vms & > Subject: Re: Opera (was Mozilla 0.6)% > Date: Sat, 23 Dec 2000 22:00:44 GMTr > C > In article <VA.000001f2.2e25de81@sture.ch>, paul@sture.ch says...s > [...]lF > > Sorry if that confused. Here's what typically happens in Netscape: > >  > > o Visit a site's home page! > > o Follow a link off that pageyE > > o Get engrossed in the article (or go off and do something else),e< > >   so that my ISDN inactivity timer comes in and hangs up& > > o Click on the browser Back button > > J > > On the last step, Opera will get it back from cache. Netscape wants to8 > > refresh the home page and so initiates another call. > F > In Netscape Edit/Preferences/Advanced/Cache, there are radio buttons= > for "Document in cache is compared to document on network:"oE > with possible values of "Once per session", "Every time" or "Never"s > B > I have no idea how it compares the cache to the network without C > downloading the network copy, but I think setting it to "Once per - > session" or "Never" might cure the problem.m > K It was at "Once per session", and I was assuming that a "session" meant an - invocation of Netscape.0  + Now set to "Never". We'll see what happens.p   <rant>K I first got online from home in 1995, via Compuserve. It suited me fine as sG I could just go online to collect unread messages and then hang up the 8G phone ASAP. Unfortunately, within a year to 18 months, large companies wH started abandoning Compuserve for the WWW. Fine for folks in the US who F could pay a few cents for a connection and then leave it open forever.  J I was then in the UK, and a simple bit of maths told me that to keep open K a local call for 40 working hours per week, even with discount, would have  8 cost me in the order of 5,000 GBP (>8,000 USD) per year.  I In 1997 I read that the beta version of what was to eventually arrive as cK Windows 98 was "severely dysfunctional unless connected permanently to the rI net" *. This is what made me grab the phone bill and spreadsheet to come jG out with the above figure, and I duly fired off complaints to relevant   bodies.(  H Although I have a much cheaper solution where I am now, it is still far E from free, which is why I get upset about the abuse of my phone bill.t   </rant>n  K * Well I now _know_ that W98 is severely dysfunctional, and refuse to give sF it hard drive space, even though it means my choice of peripherals is  limited.   ___E
 Paul Sture SwitzerlandT   ------------------------------  # Date: Sun, 24 Dec 2000 06:51:14 GMTdL From: winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr")& Subject: Re: Proxies for remote access8 Message-ID: <009F50BB.5119D0DC@SSRL04.SLAC.STANFORD.EDU>  n In article <3oe16.124136$IP1.4308700@news1.giganews.com>, "William Hymen" <t18_pilot@hotmail.spam.com> writes:< >If I have two nodes A and B, and I have user id Bravo on B.5 >Can I simply create a proxy on A without having user5 >bravo also on A?  ie:> >$mcr authorize add/proxy b::bravo local_valid_A_user /default >t    9 >I need user bravo on B to be able to push save-sets overr? >to A (decnet)  Does it matter if I use user system or sysbatcheF >as a proxy account on A? (I know these accounts will always be there)  F Yes, no problem with that, works just fine (assuming DECnet phase IV).M You do have to consider whether you want Mr. Bravo to have all the privilegedgE access to node A that proxying to a privileged account will give him..  L Also, it works just fine to do your backup locally and then copy the savesetI over DECnet, but I seem to recall (probably in an earlier version of VMS;eJ haven't tried it lately) that there are sometimes issues with creating theK saveset remotely, eg backup dkwhatever/image a::[directory]dkbck.bck/save  h! didn't always work exactly right.a   -- Alan   O ===============================================================================u0  Alan Winston --- WINSTON@SSRL.SLAC.STANFORD.EDUM  Disclaimer: I speak only for myself, not SLAC or SSRL   Phone:  650/926-3056oM  Physical mail to: SSRL -- SLAC BIN 69, PO BOX 4349, STANFORD, CA  94309-0210sO ===============================================================================n   ------------------------------  % Date: Sun, 24 Dec 2000 12:35:54 -0500 3 From: "Walker Bennett" <Wabennett@gw.total-web.net>s Subject: Re: system disk space. Message-ID: <t4ccukgv61m75@corp.supernews.com>  F Check OPERATOR.LOG, ACCOUNTNG.DAT, and the security log - these have aL tendancy to get out of hand fast.  I'd turn off accounting and security, andK put a line in systartup to create a new operator log and purge the old one.l  H Also, if you run UCX, you'll have a pile of ftp logs left around - purge	 them off.e   Walker Bennett wabennett@gw.total-web.net w_bennett@msn.comd wbennett@caldwellspartin.com        9 Jim Binder <james.binder@pss.boeing.com> wrote in messages( news:3A42A72A.5DD26E87@pss.boeing.com...F > I've inherited an Alpha 4100 w/openVMS 7.2 installed.  We do not add
 > softwareJ > yet, the system disk is getting smaller.  Is there any location I should > look1 > for logs or temp files which should be deleted?  >l > Thanks a lot.    ------------------------------   Date: 24 Dec 2000 13:37:15 GMT* From: helbig@astro.rug.nl (Phillip Helbig)/ Subject: Re: when is SYS$NODE normally defined? . Message-ID: <924u6b$hlb$1@info.service.rug.nl>  G >   This logical name is reserved to DECnet.  I would tend to recommendlE >   against defining reserved logical names, unless expressly told to0 >   do so...  E Agreed.  IIRC, it was something involving DECwindows and/or announce/ G welcome messages which wanted SYS$NODE to denote the machine.  Nothing o self-written....   ------------------------------   End of INFO-VAX 2000.717 ************************