Re: The POSiX Animal On An MPE Box - Hewlett Packard

This is a discussion on Re: The POSiX Animal On An MPE Box - Hewlett Packard ; Denys says: 1) The MAX REC is 2,147,483,647, yet the build says 147,483,647; it's dropped the MSB of 2GB. Whoops. 2) The REC SIZE is 1 byte, your BUILD says 8192. 3) The file is ASCII, BYTESTREAM, your BUILD says ...

+ Reply to Thread
Results 1 to 3 of 3

Thread: Re: The POSiX Animal On An MPE Box

  1. Re: The POSiX Animal On An MPE Box

    Denys says:

    1) The MAX REC is 2,147,483,647, yet the build says 147,483,647; it's dropped
    the MSB of 2GB. Whoops.

    2) The REC SIZE is 1 byte, your BUILD says 8192.
    3) The file is ASCII, BYTESTREAM, your BUILD says ASCII, VARIABLE.
    4) The FILE CODE is 0, your BUILD says code=2700.
    5) The MAX LABELS is 0, your BUILD says 27.
    >


    Brian retorts:

    1) Unfortunately, the flimit parm on the FOPEN uses a PIC S9(09) comp field
    and that is what is dropping off the 2. Any suggestions on how to get around
    this little snafu?

    2) FLABELINFO is returning the 8192 as the rec size of the bytestream file.
    3) The temp file need not be a bytestream file but it is FLABELINFO that is
    giving
    me all this info. Nothing is hardcoded. It works perfectly on every file
    type
    except the bytestream files which have that enormous flimit. Actually,
    there is nothing to stop the program from makint the eof=flimit.
    That wouldn't be a problem.....

    4) file code=2700 is correct, this tells the users that their input file was
    processed. When they go and do their stuff to put the file back to its
    original state the 2700 is replaced with the original file code.

    5) Program needs 27 user labels in the temp file to store all the attribsof
    the original file so 27 is correct.

    The program works just fine on every file type (flat, KSAM/3000, KSAMXL,
    CIR, RIO etc.) but not bytestream files.

    So, I think I need to have a closer look at the logic when processing
    bytestream files.


    Brian.

    On Sat, 16 Aug 2008 10:00:39 -0500, Denys Beauchemin wrote:

    >Understood. Let me point out that the program doing the interpretation of
    >the file information and then formatting the BUILD command with that
    >information is doing a piss-poor job. If you look closely at your LISTF,2
    >you will notice this:
    >
    >MAX REC: 2147483647
    >REC SIZE: 1(BYTES)
    >FOPTIONS: ASCII,BYTESTREAM,NOCCTL,STD
    >FILE CODE : 0
    >MAX LABELS: 0
    >
    >And from that it creates this BUILD command:
    >
    >BUILD
    >../.sh_history;TEMP;CODE=2700;REC=-8192,0001,V,ASCII;DISC=147483647;ULABEL=0
    >27
    >
    >The MAX REC is 2,147,483,647, yet the build says 147,483,647; it's dropped
    >the MSB of 2GB. Whoops.
    >The REC SIZE is 1 byte, your BUILD says 8192.
    >The file is ASCII, BYTESTREAM, your BUILD says ASCII, VARIABLE.
    >The FILE CODE is 0, your BUILD says code=2700.
    >The MAX LABELS is 0, your BUILD says 27.
    >
    >I think the file name is right.
    >
    >
    >
    >Denys...
    >
    >-----Original Message-----
    >From: HP-3000 Systems Discussion [mailto:HP3000-L@RAVEN.UTC.EDU] On Behalf
    >Of Brian Donaldson
    >Sent: Friday, August 15, 2008 8:03 AM
    >To: HP3000-L@RAVEN.UTC.EDU
    >Subject: Re: [HP3000-L] The POSiX Animal On An MPE Box
    >
    >I'm not hardcoding these values -- I get these values from the input file (a
    >bytestream file) and use it's attributes for the temporary file. And that is
    >what is causing the problem. So I guess I will just build the temp file with
    >a smaller flimit.
    >
    >
    >Brian.
    >
    >
    >On Thu, 14 Aug 2008 21:59:00 -0500, Denys Beauchemin
    >wrote:
    >
    >>I think that if you multiply 8192 by 147,483,647 all will be explained.

    >The
    >>message is misleading, you are in fact trying to create a 1,208,186,036,244
    >>byte file. I think that's rather large. If I read that right, and it's
    >>very late where I am, that's a 1.2 terabyte file. You will need a bigger
    >>disk. :-)
    >>
    >>The other file you show is a 2GB byte stream file, which I believe is the
    >>default maximum size. Change your REC specifications.
    >>
    >>Denys
    >>
    >>-----Original Message-----
    >>From: HP-3000 Systems Discussion [mailto:HP3000-L@RAVEN.UTC.EDU] On Behalf
    >>Of Brian Donaldson
    >>Sent: Thursday, August 14, 2008 9:17 PM
    >>To: HP3000-L@RAVEN.UTC.EDU
    >>Subject: Re: [HP3000-L] The POSiX Animal On An MPE Box
    >>
    >>Now I see what my error is --
    >>
    >>BUILD /DEV/PUB/mama/papa/baby/bear/porridge/.sh_history;TEMP;CODE=2
    >>700;REC=-8192,0001,V,ASCII;DISC=147483647;ULABEL=027
    >>
    >>====>>> EXTENT SIZE EXCEEDS MAXIMUM (FSERR 106) ====<<<
    >>
    >>Build of file "/DEV/PUB/mama/papa/baby/bear/porridge/.sh_history" failed.
    >>(CIERR
    >> 279)
    >>
    >>However, I am not specifying extents on the build command.
    >>
    >>The temp file is being built with the same FLIMIT as the perm file and that
    >>is causing the extents size error.
    >>
    >> FILE: /DEV/PUB/mama/papa/baby/bear/porridge/.sh_history
    >>
    >> FILE CODE : 0 FOPTIONS: ASCII,BYTESTREAM,NOCCTL,STD
    >> BLK FACTOR: 1 OWNER : MGR.DEV
    >> REC SIZE: 1(BYTES) GROUP ID: DEV
    >> BLK SIZE: 1(BYTES) SECURITY--READ :
    >> EXT SIZE: 0(SECT) WRITE :
    >> NUM REC: 52 APPEND :
    >> NUM SEC: 16 LOCK :
    >> NUM EXT: 1 EXECUTE :
    >> MAX REC: 2147483647 **SECURITY IS ON
    >> FLAGS : NO ACCESSORS
    >> NUM LABELS: 0 CREATED : WED, AUG 13, 2008, 7:34 PM
    >> MAX LABELS: 0 MODIFIED: WED, AUG 13, 2008, 7:35 PM
    >> DISC DEV #: 2 ACCESSED: THU, AUG 14, 2008, 7:13 PM
    >> SEC OFFSET: 0 LABEL ADDR: $0000005E.$005A9920
    >> VOLCLASS : MPEXL_SYSTEM_VOLUME_SETISC
    >>
    >>Apart from building the temp file with a smaller flimit, how do I get out

    >of
    >>this one?
    >>
    >>
    >>TIA,
    >>Brian.
    >>
    >>On Thu, 14 Aug 2008 20:20:40 -0500, Denys Beauchemin
    >>wrote:
    >>
    >>>I'm sorry, I mistyped. I meant CIERROR 279.
    >>>
    >>>:build ./denys/test1;temp
    >>>COMPONENT OF PATH DOES NOT EXIST (FSERR 457)
    >>>Build of file "./denys/test1" failed. (CIERR 279)
    >>>:
    >>>
    >>>
    >>>-----Original Message-----
    >>>From: Brian Donaldson [mailto:bmdinsocal@AOL.COM]
    >>>Sent: Thursday, August 14, 2008 8:17 PM
    >>>To: HP3000 List; Denys Beauchemin
    >>>Subject: Re: The POSiX Animal On An MPE Box
    >>>
    >>><>
    >>>
    >>>A cierror 273 on a build means the file was not built, for whatever

    >reason.
    >>>What is important to get is the FSERROR. I my case, I got a CIERROR 273
    >>>when I created a temp POSIX file in a dummy directory, but the FSERRORis
    >>>457, which is critical.
    >>>
    >>><>
    >>>
    >>>Your 273 = 273 Expected record format of F, V, or U. (CIERR 273)
    >>>Your 457 = 457 COMPONENT OF PATH DOES NOT EXIST (FSERR 457)
    >>>
    >>>(I got both of these errors from catalog.pub.sys)
    >>>
    >>>However, the error I am getting is a 279 (Build of file "!" failed. (CIERR
    >>>279))
    >>>
    >>>The file actually exists even though the prior PURGE command gave me a

    >-383
    >>>error code.
    >>>
    >>> FILE: /DEV/PUB/mama/papa/baby/bear/porridge/.sh_history
    >>>
    >>> FILE CODE : 0 FOPTIONS: ASCII,BYTESTREAM,NOCCTL,STD
    >>> BLK FACTOR: 1 OWNER : MGR.DEV
    >>> REC SIZE: 1(BYTES) GROUP ID: DEV
    >>> BLK SIZE: 1(BYTES) SECURITY--READ :
    >>> EXT SIZE: 0(SECT) WRITE :
    >>> NUM REC: 52 APPEND :
    >>> NUM SEC: 16 LOCK :
    >>> NUM EXT: 1 EXECUTE :
    >>> MAX REC: 2147483647 **SECURITY IS ON
    >>> FLAGS : NO ACCESSORS
    >>> NUM LABELS: 0 CREATED : WED, AUG 13, 2008, 7:34 PM
    >>> MAX LABELS: 0 MODIFIED: WED, AUG 13, 2008, 7:35 PM
    >>> DISC DEV #: 2 ACCESSED: THU, AUG 14, 2008, 6:07 PM
    >>> SEC OFFSET: 0 LABEL ADDR: $0000005E.$005A9920
    >>> VOLCLASS : MPEXL_SYSTEM_VOLUME_SETISC
    >>>
    >>>Brian.
    >>>
    >>>
    >>>On Thu, 14 Aug 2008 19:53:49 -0500, Denys Beauchemin
    >>>wrote:
    >>>
    >>>>I'm a little confused. I can certainly create a temporary POSIX fileand
    >>>>view it with listfile ;temp, as donna explained.
    >>>>
    >>>>But I cannot create a temporary POSIX file in a dummy directory. CIERROR
    >>>>-383, a warning, simply means the temp file was not found, for whatever
    >>>>reason.
    >>>>
    >>>>A cierror 273 on a build means the file was not built, for whatever

    >>reason.
    >>>>What is important to get is the FSERROR. I my case, I got a CIERROR 273
    >>>>when I created a temp POSIX file in a dummy directory, but the FSERROR is
    >>>>457, which is critical.
    >>>>
    >>>>Just do a showjcw and see what the fserror is and let us know.
    >>>>
    >>>>Denys
    >>>>
    >>>>-----Original Message-----
    >>>>From: HP-3000 Systems Discussion [mailto:HP3000-L@RAVEN.UTC.EDU] On

    >Behalf
    >>>>Of Brian Donaldson
    >>>>Sent: Thursday, August 14, 2008 7:29 PM
    >>>>To: HP3000-L@RAVEN.UTC.EDU
    >>>>Subject: Re: [HP3000-L] The POSiX Animal On An MPE Box
    >>>>
    >>>>Then why is my code not working? --
    >>>>
    >>>>
    >>>>Back to my original posting --
    >>>>
    >>>><< In my program I am purging a file in the temp domain and then building
    >>>it
    >>>>again in the temp domain (possibly has different parms on the build
    >>>>statement, that is why I am purging the old one first)
    >>>>
    >>>>>-0383>-0007>PURGE
    >>>>/DEV/PUB/mama/papa/baby/bear/porridge/.sh_history,TEMP;NOCONFIRM
    >>>>
    >>>>The -383 is telling me the purge failed because the file does not exist

    >in
    >>>>the temp domain, which is just fine.
    >>>>
    >>>>So further down in the logic I am building the file in the temp domain.
    >>>>The build command fails.
    >>>>
    >>>>Works just fine on files in the MPE file space but obviously isn't

    >working
    >>>>in the HFS POSiX space:
    >>>>
    >>>>
    >>>>
    >>>> <
    >>>>>+0279>+0106>BUILD
    >>>>/DEV/PUB/mama/papa/baby/bear/porridge/.sh_history;TEMP;CODE=2700;REC=-819

    >2
    >>,
    >>>0
    >>>>001,V,ASCII
    >>>>;DISC=147483647;ULABEL=027
    >>>>
    >>>>
    >>>> <
    >>>>>>Build of file "!" failed. (CIERR 279)
    >>>>
    >>>> <<
    >>>>
    >>>>The +279 is the code returned when the file already exists in the temp
    >>>>domain.
    >>>>
    >>>>Can anyone tell me why this isn't working in the POSiX area but works

    >fine
    >>>>in the MPE area?
    >>>>
    >>>>>>
    >>>>
    >>>>
    >>>>Brian.
    >>>>
    >>>>
    >>>>On Thu, 14 Aug 2008 15:12:11 -0700, donna hofmeister
    >>>>wrote:
    >>>>
    >>>>>temporary posix files are quite possible:
    >>>>>
    >>>>>: build ./foo;temp
    >>>>>: listfile ./@,2;temp
    >>>>>
    >>>>> TEMPORARY FILES FOR MGR.DONNA
    >>>>>
    >>>>> PATH= /DONNA/PUB/
    >>>>>
    >>>>> CODE ------------LOGICAL RECORD----------- ----SPACE---- FILENAME
    >>>>> SIZE TYP EOF LIMIT R/B SECTORS #X MX
    >>>>>
    >>>>> 128W FB 0 1023 1 0 0 * foo
    >>>>>
    >>>>>(I do believe, I do believe)
    >>>>>
    >>>>> - donna
    >>>>>
    >>>>>---
    >>>>>Donna Hofmeister
    >>>>>Allegro Consultants, Inc.
    >>>>>408-252-2330
    >>>>>
    >>>>>
    >>>>>> -----Original Message-----
    >>>>>> From: HP-3000 Systems Discussion [mailto:HP3000-L@RAVEN.UTC.EDU] On
    >>>>>> Behalf Of Johnson, Tracy
    >>>>>> Sent: Thursday, August 14, 2008 6:36 AM
    >>>>>> To: HP3000-L@RAVEN.UTC.EDU
    >>>>>> Subject: Re: [HP3000-L] The POSiX Animal On An MPE Box
    >>>>>>
    >>>>>> I always believed you cannot have TEMP files in the POSIX space, they
    >>>>>> only exist in MPE space.
    >>>>>>
    >>>>>>
    >>>>>> Tracy Johnson
    >>>>>> Measurement Specialties Inc.
    >>>>>> BT
    >>>>>>
    >>>>>>
    >>>>>>
    >>>>>>
    >>>>>>
    >>>>>>
    >>>>>> NNNN
    >>>>>>
    >>>>>> > -----Original Message-----
    >>>>>> > From: HP-3000 Systems Discussion
    >>>>>> > [mailto:HP3000-L@RAVEN.UTC.EDU] On Behalf Of Brian Donaldson
    >>>>>> > Sent: Thursday, August 14, 2008 2:10 AM
    >>>>>> > To: HP3000-L@RAVEN.UTC.EDU
    >>>>>> > Subject: Re: [HP3000-L] The POSiX Animal On An MPE Box
    >>>>>> >
    >>>>>> > The POSiX saga continues.
    >>>>>> >
    >>>>>> > In my program I am purging a file in the temp domain and then
    >>>>>> > building it
    >>>>>> > again in the temp domain (possibly has different parms on the build
    >>>>>> > statement, that is why I am purging the old one first)
    >>>>>> >
    >>>>>> >
    >>>>>> >
    >>>>>> > >-0383>-0007>PURGE
    >>>>>> > /DEV/PUB/mama/papa/baby/bear/porridge/.sh_history,TEMP;NOCONFI
    >>>>>> > RM
    >>>>>> >
    >>>>>> >
    >>>>>> >
    >>>>>> > The -383 is telling me the purge failed because the file does
    >>>>>> > not exist in
    >>>>>> > the temp domain, which is just fine.
    >>>>>> >
    >>>>>> > So further down in the logic I am building the file in the
    >>>>>> > temp domain.
    >>>>>> > The build command fails.
    >>>>>> >
    >>>>>> > Works just fine on files in the MPE file space but obviously
    >>>>>> > isn't working
    >>>>>> > in the HFS POSiX space:
    >>>>>> >
    >>>>>> >
    >>>>>> >
    >>>>>> > <
    >>>>>> > >+0279>+0106>BUILD
    >>>>>> > /DEV/PUB/mama/papa/baby/bear/porridge/.sh_history;TEMP;CODE=27
    >>>>>> > 00;REC=-8192,0001,V,ASCII
    >>>>>> > ;DISC=1474
    >>>>>> > 83647;ULABEL=027
    >>>>>> >
    >>>>>> >
    >>>>>> > <
    >>>>>> > >>Build of file "!" failed. (CIERR 279)
    >>>>>> >
    >>>>>> > <<
    >>>>>> >
    >>>>>> > The +279 is the code returned when the file already exists in
    >>>>>> > the temp domain.
    >>>>>> >
    >>>>>> > Can anyone tell me why this isn't working in the POSiX area
    >>>>>> > but works fine
    >>>>>> > in the MPE area?
    >>>>>> >
    >>>>>> > Something I am doing wrong, I know but I am clueless as to
    >>>>>> > what my error
    >>>>>> > is........
    >>>>>> >
    >>>>>> > TIA,
    >>>>>> > Brian.
    >>>>>> >
    >>>>>> > * To join/leave the list, search archives, change list settings,*
    >>>>>> > * etc., please visit http://raven.utc.edu/archives/hp3000-l.html*
    >>>>>> >
    >>>>>>
    >>>>>> * To join/leave the list, search archives, change list settings, *
    >>>>>> * etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
    >>>>>
    >>>>>* To join/leave the list, search archives, change list settings, *
    >>>>>* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
    >>>>
    >>>>* To join/leave the list, search archives, change list settings, *
    >>>>* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
    >>>>No virus found in this incoming message.
    >>>>Checked by AVG - http://www.avg.com
    >>>>Version: 8.0.138 / Virus Database: 270.6.2/1609 - Release Date: 8/13/2008
    >>>>6:43 AM
    >>>>
    >>>>* To join/leave the list, search archives, change list settings, *
    >>>>* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
    >>>
    >>>
    >>>No virus found in this incoming message.
    >>>Checked by AVG - http://www.avg.com
    >>>Version: 8.0.138 / Virus Database: 270.6.2/1609 - Release Date: 8/13/2008
    >>>6:43 AM
    >>>
    >>>* To join/leave the list, search archives, change list settings, *
    >>>* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *

    >>
    >>* To join/leave the list, search archives, change list settings, *
    >>* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
    >>No virus found in this incoming message.
    >>Checked by AVG - http://www.avg.com
    >>Version: 8.0.138 / Virus Database: 270.6.2/1609 - Release Date: 8/13/2008
    >>6:43 AM
    >>
    >>* To join/leave the list, search archives, change list settings, *
    >>* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *

    >
    >* To join/leave the list, search archives, change list settings, *
    >* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
    >No virus found in this incoming message.
    >Checked by AVG - http://www.avg.com
    >Version: 8.0.138 / Virus Database: 270.6.3/1613 - Release Date: 8/15/2008
    >5:58 AM
    >
    >* To join/leave the list, search archives, change list settings, *
    >* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *


    * To join/leave the list, search archives, change list settings, *
    * etc., please visit http://raven.utc.edu/archives/hp3000-l.html *


  2. Re: The POSiX Animal On An MPE Box

    In the past I've used the Cobol Compute statement to get around this issue.

    Compute my-16-bit-nbr = 0 + 65.535

    --
    Michael Anderson,
    J3k Solutions
    Sr.Systems Programmer/Analyst
    832.515.3868




    Gilles Schipper wrote:
    > At 07:14 PM 08-08-17, Brian Donaldson wrote:
    >
    >> Brian retorts:
    >>
    >> 1) Unfortunately, the flimit parm on the FOPEN uses a PIC S9(09) comp
    >> field
    >> and that is what is dropping off the 2. Any suggestions on how to get
    >> around
    >> this little snafu?

    >
    > Yes.
    >
    > One of the problems with COBOL is how it handles binary data.
    >
    > For example, an item defined as PIC S9(04) COMP occupies 2 bytes of
    > real-estate and can have a maximum value of 9999 - even though the 16
    > bits represented by the 2-byte field can actually contain a value as
    > high as 2 to the power of 16, less 1, which equals 65,535.
    >
    > So, if you really want to account for this fact, you need to move the
    > field non-numerically to a larger COBOL field, examine its true
    > contents (allowing its true maximum value to be revealed), perform
    > some operation on it (if necessary) then move it back (again
    > non-numerically) back to its originally-sized field if required to be
    > again utilized in its smaller form.
    >
    >


    * To join/leave the list, search archives, change list settings, *
    * etc., please visit http://raven.utc.edu/archives/hp3000-l.html *


  3. Re: The POSiX Animal On An MPE Box

    Actually signed 16 bit should be Compute my-signed-16-bit-nbr = (0 +
    (16384 * 2)) - 1, unsigned ((0 + (16384 * 4) - 1.

    In any case 16 bit 32 bit and so on MOVE looks at the four digit
    limit in s9(4) comp, where compute looks at the 16 bit limit.

    It's Ok to talk to yourself, as long as you don't answer your own
    questions :-P

    I wrote:
    > In the past I've used the Cobol Compute statement to get around this
    > issue.
    >
    > Compute my-16-bit-nbr = 0 + 65.535
    >


    * To join/leave the list, search archives, change list settings, *
    * etc., please visit http://raven.utc.edu/archives/hp3000-l.html *


+ Reply to Thread