1 INFO-VAX	Thu, 21 Dec 2006	Volume 2006 : Issue 700       Contents:  Re: consulting job in raleigh nc  Re: consulting job in raleigh nc# Happy DEC-20 Day to my VMS friends! ' How do I set the unit number on a RA70?   Re: Internet Explorer 7 Released0 Re: Multiple time zones from a single VMS server- Re: Open sourced  VMS as a business concept ? , Re: Open sourced VMS as a business concept ?, Re: Open sourced VMS as a business concept ? SAN disk access  problem...  Re: SAN disk access  problem...  Re: SAN disk access  problem... * Re: Tools to rebuild corrupt indexed files* Re: Tools to rebuild corrupt indexed files* Re: Tools to rebuild corrupt indexed files  F ----------------------------------------------------------------------    Date: 20 Dec 2006 12:42:56 -0800 From: jtower@gmail.com) Subject: Re: consulting job in raleigh nc B Message-ID: <1166647376.444136.321820@73g2000cwn.googlegroups.com>   jtower@gmail.com wrote: H > i need someone with solid openvms experience for a semi-quick contractE > job (openvms on compaq alpha hardware).  if you'd like to make some F > quick holiday cash ping me as jtower@gmail.com and i'll give you theI > details.  must be familiar with block devices on openvms and alpha scsi  > hardware.   A i found someone to help me out, thanks to everyone who responded.    ------------------------------    Date: 20 Dec 2006 12:44:22 -0800 From: jtower@gmail.com) Subject: Re: consulting job in raleigh nc A Message-ID: <1166647462.597190.78130@80g2000cwy.googlegroups.com>   A i found someone to help me out, thanks to everyone who responded.    jason    jtower@gmail.com wrote: H > i need someone with solid openvms experience for a semi-quick contractE > job (openvms on compaq alpha hardware).  if you'd like to make some F > quick holiday cash ping me as jtower@gmail.com and i'll give you theI > details.  must be familiar with block devices on openvms and alpha scsi  > hardware.    ------------------------------    Date: 20 Dec 2006 15:19:50 -05003 From: Rich Alderson <news@alderson.users.panix.com> , Subject: Happy DEC-20 Day to my VMS friends!. Message-ID: <mddwt4m2uc9.fsf@panix5.panix.com>  ? And a happy Winter Solstice Festival of your choice to you all!  --  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: 20 Dec 2006 17:53:05 -0800$ From: "Bob Armstrong" <bob@jfcl.com>0 Subject: How do I set the unit number on a RA70?A Message-ID: <1166665985.155984.62730@48g2000cwx.googlegroups.com>   E   I'm sure this is easy, but I can't figure it out.  How do I set the D unit number on a RA70?  The drive has a set of 8 dip switches on theG front marked "unit number select" and a push button marked "unit number G accept".  So I set the switches, power it up, press the button, and the E unit number doesn't change!  There must be some trick to it, like the 0 drive has to be in a special state or something.  E   BTW, this drive is in a BA213 chassis, and the BA213 OCP panel does D _not_ have a unit number display (unlike the SA7x chassis OCP, which does).  A   And is it just me, or is it impossible to reach the unit number G switches/accept button with the drive mounted in a BA213?  The only way C I can get to them is to remove the drive and power it up while it's  sitting in my lap :-)   G   BTW, the RA70 is unique in this regard.  The RA71-3 drives all have a D small set of DIP switches on the side for selecting unit numbers 0-7 and, AFAIK, no push button.    Thanks, 
 Bob Armstrong    ------------------------------    Date: 20 Dec 2006 19:07:43 -0800' From: "anonymous" <kekemei263@sohu.com> ) Subject: Re: Internet Explorer 7 Released C Message-ID: <1166670463.770612.128970@n67g2000cwd.googlegroups.com>   + "Margrave of Brandenburg =D0=B4=B5=C0=A3=BA  " - > <darknessdevilz@gmail.com> wrote in message = > news:1166572103.292027.47400@a3g2000cwd.googlegroups.com... D > > Well im actualy thinking to change to this version...i'm actualyB > > working with firefox 2.0....and i heard IE 7 was better....hmm" > > :/....what should i choose.... > L > In IE 7, lots of familiar things have changed. The new features are good = .=2E. 0 > but after trying it, I went back to ... Avant. > L > Avant has had for years the tabbing and other features that IE 6 lacked. = But K > its interface is better, I think. And though it's slightly different from E > the familiar IE line of browsers, it's more similar than different.  > L > So, if you question "what should I choose" is a serious (not sarcastic) o= ne > ... I'd recommend Avant.  B No matter where to download it, just  reduces your searching time,	 isn't it?   " Free Christmas Clock screensaver :D http://www.filesforfree.com/freeware/christmas_clock_screensaver.htm  Free Happy New Year Screen SaverD http://www.filesforfree.com/freeware/happy_new_year_screen_saver.htm   ------------------------------  % Date: Wed, 20 Dec 2006 15:34:44 -0500 ( From: Wayne Sewell <wayne@tachysoft.com>9 Subject: Re: Multiple time zones from a single VMS server / Message-ID: <00A607D4.D62F1580.1@tachysoft.com>   ( >From: "mooneydl" <mooneydl@hotmail.com> >X-Newsgroups: comp.os.vms6 >Subject: Multiple time zones from a single VMS server! >Date: 19 Dec 2006 08:27:05 -0800    > G >We have a VMS server located in the U.S. Eastern Time zone (GMT-05:00) G >hosting an application used in the U.S., Switzerland (GMT+01:00),  and G >Japan (GMT+09:00). Some dates and times need to be shown in the user's G >local time zone. What libraries can be used to take a date and time as C >expressed in one time zone and express it in a different time zone C >while accounting for the different starting and ending periods for 0 >daylight saving time in the various time zones? >   F If you don't want to change your code, the DATESIM product can do thisN transparently.  When someone logs in, the entire process can run in the user'sI local timezone, while the rest of the system continues to run in the real I timezone.  As long as your application uses the standard date/time system M services and does not directly access the date/time quadword, it will see the K user's local time and not the actual system time.  This even works with VMS : itself, i.e. SHOW SYSTEM will also display the pseudotime.  L Different users can see different timezones and other users can see the realK time.  All that is required is a SYNCH_WITH_TIMEZONE in the login (or not).    http://www.softwarepartners.com O =============================================================================== N Wayne Sewell, Tachyon Software Consulting  (281)812-0738   wayne@tachysoft.com; http://www.tachysoft.com/www/tachyon.html and wayne.html    O =============================================================================== > Curly:"Hey, I'm no mule."   Moe:"No, your ears are too short!"   ------------------------------  + Date: Wed, 20 Dec 2006 20:08:03 +0000 (UTC) P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)6 Subject: Re: Open sourced  VMS as a business concept ?$ Message-ID: <emc573$kfq$1@online.de>  A In article <d023b$45873c5d$cef8887a$20091@TEKSAVVY.COM>, JF Mezei ( <jfmezei.spamnot@vaxination.ca> writes:   L > With a larger-than-expected number of VMS engineers no longer working for I > the current owner of VMS, but apparently still controlling some of the   > source they contributed,      In a LEGAL sense controlling it?  3 > isn't this a low process of open sourcing VMS to  
 > hobbyists ?   E The moment VMS is open-sourced is the moment it starts quickly dying.    ------------------------------    Date: 20 Dec 2006 12:46:58 -0800( From: "Ian Miller" <gxys@googlemail.com>5 Subject: Re: Open sourced VMS as a business concept ? A Message-ID: <1166647616.605275.89710@80g2000cwy.googlegroups.com>   / Phillip Helbig---remove CLOTHES to reply wrote:  > G > The moment VMS is open-sourced is the moment it starts quickly dying.    Why?   ------------------------------  % Date: Thu, 21 Dec 2006 06:02:27 +0100 / From: Paul Sture <paul.sture.nospam@hispeed.ch> 5 Subject: Re: Open sourced VMS as a business concept ? J Message-ID: <paul.sture.nospam-132408.06022721122006@mac.sture.homeip.net>  $ In article <emcffj$bod$1@online.de>,E  helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to    reply) wrote:  H > In article <1166647616.605275.89710@80g2000cwy.googlegroups.com>, "Ian( > Miller" <gxys@googlemail.com> writes:  > 3 > > Phillip Helbig---remove CLOTHES to reply wrote:  > > > K > > > The moment VMS is open-sourced is the moment it starts quickly dying.  > >  > > Why? > G > Simple.  What differentiates VMS from other operating systems is the  H > fact that it is well planned.  Without meaning any harm to Linus, the G > fact that a student could write a unix kernel in his spare time says  I > more about the lack of sophistication of the latter than the skills of  J > the former.  I don't see how VMS could continue in an open-source world K > without a lot of coordination, and I don't see anyone with the knowledge  J > and resources willing to do this for free.  A lot of work is going into I > VMS now and this is funded by customers.  Cut off the funding, cut out  J > the development.  Many of the customers would not be willing to pay for  > an open-source product.   F Just to pick one aspect of such an idea, imagine what would happen to  the VMS documentation.   :-(      <snip>   > D > And another danger of going the open-source route is the danger ofJ > approaching the RMS camp, where not only should all software be free butC > people who write software for a living are brandmarked as morally J > inferior.  Read some RMS speeches; it is difficult to find anything moreE > hateful and inhumane in all of literature.  If he wants to make his B > software free, fine.  I've given away software I've written.  NoG > problem.  But I think the concept of someone else telling me that my  K > software should be free is going too far.  I don't want to have anything  H > to do with such hate-mongerers.  If you think this is too extreme, do K > some research into the philosophy, software and otherwise, of the famous  ) > TLA-gurus.  Be afraid.  Be very afraid.  >   > Sorry, what is TLA in this context and who are the "TLA-gurus"   --  
 Paul Sture   ------------------------------    Date: 20 Dec 2006 14:30:43 -0800( From: "rgscomp" <ray@rgscomputing.co.uk>$ Subject: SAN disk access  problem...B Message-ID: <1166653843.325348.67760@n67g2000cwd.googlegroups.com>   Hi All,   F Apologies for lack of detail but the system in question has been builtG by some colleagues and I only got involved for a couple of hours today.   C System is 4-node cluster, ES45 systems with EMC SAN and (I believe)  CISCO SAN switches. VMS V7.3-2.   E Using wwidmgr, we can map the boot disk and all systems boot OK. When F VMS is running, the system disk can be seen as a multipath device, but9 none of the remaining 9 data disks are multipath devices. @ At the >>> prompt, we have added another of the data disks usingF wwidmgr, and at this point there are two paths to both the system disk and this data disk.   D I disconnected the (working) A-path cable to see if the system wouldF boot and mount the data disks via the B-path. The system booted OK viaG PGB0 but the data disks failed to mount and "sho dev dg" only shows the A system disk. "IO AUTO" fails to find any devices down the b-path.   D So, the wwidmgr results seem to indicate that at the console prompt,+ all the EMC presented disks show two paths. E The VMS results seem to say that when VMS is running, only the system D disk is multipath, but all the others are only single pathed via the A-path.   F We will be talking to HP and EMC tomorrow, but wanted to ask if any ofF the group have either seen this activity or can suggest some places to check.  C I'm struggling with the idea that the console can see dual paths to G every disk, but VMS can only see the system disk as dual pathed. As its G the same driver, I can't think how it can be selective over which disks  to "see" dual-pathed. D My current thoughts are that the SAN config is OK, but we may have a* firmware/driver mismatch on the B-channel.B The fact that all 4 systems are affected in the same way, however,B worries me as the chances of all the mismatched HBA's being on the B-path seems remote.  G I'll be logged out until the morning here in the UK, but posted tonight 4 on the off chance someone would be able to help out.   Cheers,    Ray.   ------------------------------  % Date: Wed, 20 Dec 2006 20:30:29 -0600 1 From: Michael Austin <maustin@firstdbasource.com> ( Subject: Re: SAN disk access  problem...< Message-ID: <nlmih.2293$ZT3.1295@newssvr19.news.prodigy.com>   rgscomp wrote:  	 > Hi All,  > H > Apologies for lack of detail but the system in question has been builtI > by some colleagues and I only got involved for a couple of hours today.  > E > System is 4-node cluster, ES45 systems with EMC SAN and (I believe) ! > CISCO SAN switches. VMS V7.3-2.  > G > Using wwidmgr, we can map the boot disk and all systems boot OK. When H > VMS is running, the system disk can be seen as a multipath device, but; > none of the remaining 9 data disks are multipath devices. B > At the >>> prompt, we have added another of the data disks usingH > wwidmgr, and at this point there are two paths to both the system disk > and this data disk.  > F > I disconnected the (working) A-path cable to see if the system wouldH > boot and mount the data disks via the B-path. The system booted OK viaI > PGB0 but the data disks failed to mount and "sho dev dg" only shows the C > system disk. "IO AUTO" fails to find any devices down the b-path.  > F > So, the wwidmgr results seem to indicate that at the console prompt,- > all the EMC presented disks show two paths. G > The VMS results seem to say that when VMS is running, only the system F > disk is multipath, but all the others are only single pathed via the	 > A-path.  > H > We will be talking to HP and EMC tomorrow, but wanted to ask if any ofH > the group have either seen this activity or can suggest some places to > check. > E > I'm struggling with the idea that the console can see dual paths to I > every disk, but VMS can only see the system disk as dual pathed. As its I > the same driver, I can't think how it can be selective over which disks  > to "see" dual-pathed. F > My current thoughts are that the SAN config is OK, but we may have a, > firmware/driver mismatch on the B-channel.D > The fact that all 4 systems are affected in the same way, however,D > worries me as the chances of all the mismatched HBA's being on the > B-path seems remote. > I > I'll be logged out until the morning here in the UK, but posted tonight 6 > on the off chance someone would be able to help out. > 	 > Cheers,  >  > Ray. >   K It may be a cofiguration issue in the EMC storage.  I know that in the HSG  J storage, you had to define each "connection" to the logical device on the 2 controller.  I am not sure how that occurs on EMC.  P Again, on HSG storage with dual controllers each with dual ports you would have  4 connections, not 2. 4                             /---- controllerA Port 1 systemA|     /--FABRIC-A=== 4     hba1|____/              \---- controllerA Port 2	         | 
     hba2|____ 4         |    \              /---- controllerB Port 1               \--FABRIC-B---4                             \---- controllerB Port 2  ; Each LUN would need to be configured for all 4 connections.     What does "show dev/multi" show?   --   Michael Austin. 
 Consultant www.firstdbasource.com   ------------------------------    Date: 21 Dec 2006 00:32:21 -06004 From: cornelius@encompasserve.org (George Cornelius)( Subject: Re: SAN disk access  problem...3 Message-ID: <vSRut$0LkTWb@eisner.encompasserve.org>   p In article <nlmih.2293$ZT3.1295@newssvr19.news.prodigy.com>, Michael Austin <maustin@firstdbasource.com> writes: > rgscomp wrote:R > Again, on HSG storage with dual controllers each with dual ports you would have  > 4 connections, not 2. 6 >                             /---- controllerA Port 1 >  systemA|     /--FABRIC-A===6 >     hba1|____/              \---- controllerA Port 2 >         |  >     hba2|____ 6 >         |    \              /---- controllerB Port 1 >               \--FABRIC-B---6 >                             \---- controllerB Port 2 >   B Hmm. We do it this way, so each HSG controller is on both fabrics:  5                              /---- controllerA Port 1    systemA|     /--FABRIC-A=== 5      hba1|____/              \---- controllerB Port 2 
          |      hba2|____5          |    \              /---- controllerB Port 1                 \--FABRIC-B--- 5                              \---- controllerA Port 2     E And of course each LUN (or, equivalently, id) is configured for which F HBA's are allowed to connect to it (probably not necessary if you have a dedicated fabric).   --@ George Cornelius              cornelius@SPAMNOTencompasserve.org7                               cornelius@SPAMNOTmayo.edu    ------------------------------    Date: 20 Dec 2006 18:29:14 -0800< From: "Hein RMS van den Heuvel" <heinvandenheuvel@gmail.com>3 Subject: Re: Tools to rebuild corrupt indexed files B Message-ID: <1166668154.146237.129530@79g2000cws.googlegroups.com>   JF Mezei wrote: H > What tools exist on VMS to traverse and extract records from a corrupt > indexed file ?   Beer.     7 > TYPE/OUTPUT gave me about 300 records before failing.   F If the key's are somewhat predictable (hint 1: the index buckets! HintF 2: an old backup), then you can use a program to issue  { $GET + RAC=0F untill it breaks then $FIND RAC=KEY, KBF=<next value to try> untill it works } loop while not EOF.   M > And if you have suggestions for tools, can you provide a hint of where they L > can be found ? (especially if hidden in some non descript directory in the > freeware).  C In the RMS_tools directory you now found (how non descript was that 5 name?) check out RMS_TUNING.PPT slide 39 and onwards.    >  >  > Another question: I > What exactly do the Index Fill Quantity and Data Fill Quantities mean ?   G It's only applicable for the initial load, indicating how many bytes to G fill before declaring a bucket full. The specification is as percentage G of the bucket size. The stored value a byte count. Here 100% = 30 * 512  = 15,000 give or take.  @ If it is truly critical data, then for mere money HP support, or( myself, will resurrect whatever remains.   :-)    Hein.    ------------------------------  % Date: Wed, 20 Dec 2006 23:00:10 -0500 . From: JF Mezei <jfmezei.spamnot@vaxination.ca>3 Subject: Re: Tools to rebuild corrupt indexed files 7 Message-ID: <842a3$458a06f4$cef8887a$5422@TEKSAVVY.COM>    Hein RMS van den Heuvel wrote:I >> What tools exist on VMS to traverse and extract records from a corrupt  >> indexed file ?  >  > Beer.    Thanks ! Needed a good laugh !    H > If the key's are somewhat predictable (hint 1: the index buckets! HintH > 2: an old backup), then you can use a program to issue  { $GET + RAC=0H > untill it breaks then $FIND RAC=KEY, KBF=<next value to try> untill it > works } loop while not EOF.   K Yeah, if the RMS_TOOLS don't turn out to be good enough, I was thinking of  G just writing a program to extrcat records. By RAC=0 I take it you mean   sequential access ?   G By using DUMP, I can pretty much find the first "good" block after the  I overwritten/corrupt blocks, and in there I will see the key (the promary  J key of that file is a filename OA$SHARExx:Z<mumble>.TXT or .WPL, so it is  quite obvious in a dump.  	 Question:   J Say the file is 42000 blocks big. And that blocks 35000 to 39000 are bad. I Is there an easy way to find out if any index blocks are located in this  ? area ? (EG: get a list of VBNs that contain the primary index.)   J If a portion of the index has been zapped by VMS' LDdriver, is it correct I to state that there is no way to traverse the whole index and get to the  L records that would be at the end of the index ?  In such a case, I might be F better off writing a parser that just reads the raw data, looks for a D telltale start of record ("OA$SHARE") and then read the record data.  	 Question:   H Say I find the first byte of a record. Can I just step back 2 bytes and : read the record's length (the records are variable size) ?  K Since these records are likely to span more than one block (they are up to  J 2000 bytes), would I also find some pointer to the next block within that I data, or do I just assume that the next VBN contains the continuation of  
 the data ?    B > If it is truly critical data, then for mere money HP support, or* > myself, will resurrect whatever remains.  H Well, I see this as a learning experience. This is not mission critical F data, it is hobbyist-critical data :-) This whole ordeal is giving me K skills that I feel are of use. It is my first ever disk crash where I have  L an opportunity to recover a large portion of the data.  When I had lost MAC I disks, the utilities to scan drives never really gave me any usable data.    ------------------------------  % Date: Thu, 21 Dec 2006 01:30:41 -0500 ' From: Dave Froble <davef@tsoft-inc.com> 3 Subject: Re: Tools to rebuild corrupt indexed files 9 Message-ID: <OL2dnRgEnuJttBfYnZ2dnUVZ_s7inZ2d@libcom.com>    JF Mezei wrote:  > Dave Froble wrote:? >> Are you converting to an indexed file, or a sequential file?  > J > Trying to convert from an indexed file with bad blocks in the middle to J > ANYTHING ... (I have the FDLs so I can recreate it once I have the data  > out of this corrupted file).  D If the corruption iss in secondary key buckets, then a CONVERT to a E sequential file will recover the data.  A CONVERT back to an indexed   file completes the recovery.  & >> I get paid to do these things.  :-) > I > Do you accept vouchers giving you the right to hit me on the head with   > your baseball bats ?  $ No, I already assume that privilege.   --  4 David Froble                       Tel: 724-529-0450> Dave Froble Enterprises, Inc.      E-Mail: davef@tsoft-inc.com DFE Ultralights, Inc.  170 Grimplin Road  Vanderbilt, PA  15486    ------------------------------   End of INFO-VAX 2006.700 ************************