|
View previous topic :: View next topic
|
| Author |
Message |
cvishu
Active User
Joined: 31 Jul 2007
Posts: 136
Location: india
|
|
|
|
Hi ,
I am reading a KSDS and if that record matches some criteria i need to delete it fro the KSDS and then continue reading the KSDS till EOF.
My logic:
I read the ksds seuentially after a start brws , and then if i have to delete a record , i store the KEY in a working storage variable , and then delete the record followed by an end browse . Agian i give a start browse with the stored Key andd GTEQ option and cotinue my process till EOF .
Problem :
It works fine , now the problem is for the last record, when i reach the last record , i deleted and gave a start browse again , instead of getting a CICS response 20 (EOF) i am getting 13 (row not found) .
Could explain me the funda why it happens so ? (KSDS is intiallized to low values ) |
|
| Back to top |
|
 |
Binop B
Active User
Joined: 18 Jun 2009
Posts: 407
Location: Nashville, TN
|
|
|
|
Hi,
Please refer the manual for STARTBR command
As per the manual, it clearly says the given below are the error conditions for STARTBR
| Quote: |
Conditions: DISABLED, FILENOTFOUND, ILLOGIC, INVREQ,
IOERR, ISCINVREQ, LOADING, NOTAUTH, NOTFND, NOTOPEN, SYSIDERR |
The ENDFILE condition that you are expecting will never be returned by the STARTBR command... |
|
| Back to top |
|
|
Bill O'Boyle
CICS Moderator
Joined: 14 Jan 2008
Posts: 2501
Location: Atlanta, Georgia, USA
|
|
|
|
This is bad practice. A STARTBR, READNEXT, ENDBR, DELETE, STARTBR could cause a deadlock when the file is defined to LSR (Local Shared Resource) but, if it is not LSR defined (known as NSR - Non Shared Resource) the VSAM CRP (Current Record Pointer, under the covers), at the very least, has the potential to lose track as to where it's at in the file.
In other words, don't mix Random access with Sequential access unless you fully understand the pitfalls and consequences.
A different method is recommended....
Bill |
|
| Back to top |
|
|
cvishu
Active User
Joined: 31 Jul 2007
Posts: 136
Location: india
|
|
|
|
| Thanks Binop |
|
| Back to top |
|
|
cvishu
Active User
Joined: 31 Jul 2007
Posts: 136
Location: india
|
|
|
|
Bill ,
Thanks for the response , but am not sure i understand why you recomend this method not to be used. I am not sure i understand the harm , could you please ellobrate with an example / if you could suggest an alternative it would be of great help .
Thanks |
|
| Back to top |
|
|
Binop B
Active User
Joined: 18 Jun 2009
Posts: 407
Location: Nashville, TN
|
|
|
|
| Use READNEXT statement with UPDATE option, and DELETE the record then and there .... You will not have to ENDBR and do random access every time ... |
|
| Back to top |
|
|
Robert Sample
Global Moderator
Joined: 06 Jun 2008
Posts: 8700
Location: Dubuque, Iowa, USA
|
|
|
|
| Terminology note: terminology is critical in IT, where similar words may mean very different things. Your statement that a response code 13 is "row not found" is not only incorrect, it has negative implications. Files have records, not rows. Databases have rows. Stating that you didn't get any rows would send a support person looking towards a database problem, not a code issue or VSAM problem. |
|
| Back to top |
|
|
Bill O'Boyle
CICS Moderator
Joined: 14 Jan 2008
Posts: 2501
Location: Atlanta, Georgia, USA
|
|
|
|
Binop,
There isn't a READNEXT command with UPDATE. The OP would need to ENDBR, save the current key, READ for UPDATE, DELETE and issue another STARTBR, using the saved-key (GTEQ) in the RIDFLD. Because the saved-key record had been deleted, VSAM should point the CRP at the record following the deleted record (EIBRESP=ZERO) or raise a NOTFND condition (effectively EOF).
This is the safest method to avoid deadlock's and/or messing up the CRP when combining Random and Sequential access together.
Bill |
|
| Back to top |
|
|
Binop B
Active User
Joined: 18 Jun 2009
Posts: 407
Location: Nashville, TN
|
|
| Back to top |
|
|
Bill O'Boyle
CICS Moderator
Joined: 14 Jan 2008
Posts: 2501
Location: Atlanta, Georgia, USA
|
|
|
|
Binop,
Yes, I read the LINK and (much to my surprise), a READNEXT with UPDATE is now offered. However, it is only available under RLS (Record Level Sharing).
I don't get to code too much anymore, so my apologies.
I think the OP needs to review the usage of this command together with or without RLS. The following link may help -
VSAM Demystified -
www.redbooks.ibm.com/redbooks/pdfs/sg246105.pdf
Bill |
|
| Back to top |
|
|
Bill O'Boyle
CICS Moderator
Joined: 14 Jan 2008
Posts: 2501
Location: Atlanta, Georgia, USA
|
|
|
|
| Quote: |
| "Files have records, not rows. Databases have rows. Stating that you didn't get any rows would send a support person looking towards a database problem, not a code issue or VSAM problem". |
Robert,
Can't count the number of times you've explained and posted this.
Seems like nobody pays any attention and posters continue to perpetuate this mismatch of terminology....
Bill |
|
| Back to top |
|
|
Robert Sample
Global Moderator
Joined: 06 Jun 2008
Posts: 8700
Location: Dubuque, Iowa, USA
|
|
|
|
Bill:
And I can't count the number of times terms have been misused on this forum, causing responders to look in one direction, when the actual problem was in a different area that the correct terminology would have helped identify earlier. Yet the posters get mad when told their lack of correct terminology is causing the delay in answering! |
|
| Back to top |
|
|
Earl Haigh
Active User
Joined: 25 Jul 2006
Posts: 475
|
|
|
|
| Quote: |
The OP would need to ENDBR, save the current key, READ for UPDATE, DELETE and issue another STARTBR, using the saved-key (GTEQ) in the RIDFLD. Because the saved-key record had been deleted, VSAM should point the CRP at the record following the deleted record (EIBRESP=ZERO) or raise a NOTFND condition (effectively EOF).
|
Bill's recommendation is the industry standard.
Just do it, and move on with your work.. |
|
| Back to top |
|
|
CICS Guy
Senior Member
Joined: 18 Jul 2007
Posts: 2146
Location: At my coffee table
|
|
|
|
| Robert Sample wrote: |
| And I can't count the number of times terms have been misused on this forum, causing responders to look in one direction, when the actual problem was in a different area that the correct terminology would have helped identify earlier. Yet the posters get mad when told their lack of correct terminology is causing the delay in answering! |
Robert, I sometimes think you are out on a crusade, a modern day IT Don Quixote, the man of La Mancha....... 
FWIW, I don't know what other 'responders' read into some of the clues we get, but row/record and column/field (and to some extent, sort/jcl) are fairly much interchangeable in my mind, depending upon the data structure. |
|
| Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006
Posts: 19243
Location: Inside the Matrix
|
|
|
|
If our "senior" people do not insist on proper terminology, who will.
If we let slop get by, we're "aiding and abetting". . .
From what i can tell, most of the new people would prefer to "say it right", but if their teachers, managers, and the seniors they interact with are not going to insist, they may go a long time not even knowing that their choice of words makes them look ignorant, lazy, and sloppy.
Quite often "wrong wording" causes confusion and makes a topic get ugly because people are not on the same page. Using improper terms is also quite unprofessional. It has also cost people a job they were being considered for.
When i mess up, i'm thankful when someone tells me. . . Many/most new people are also.
d |
|
| Back to top |
|
|
Robert Sample
Global Moderator
Joined: 06 Jun 2008
Posts: 8700
Location: Dubuque, Iowa, USA
|
|
|
|
| I was working one place where someone who interviewed claimed he could code in "COBALT". Whether or not that was true, the boss decided if the person was unable to properly identify the language being used, he could not hire the person. So I have personal experience that improper terminology keeps people out of jobs. |
|
| Back to top |
|
|
CICS Guy
Senior Member
Joined: 18 Jul 2007
Posts: 2146
Location: At my coffee table
|
|
|
|
| dick scherrer wrote: |
| If our "senior" people do not insist on proper terminology, who will. |
IIRC, some of those 'senior people' had their heads up their **es, I learned in that environment for a while, and it did promote a feeling of failure.
| Quote: |
| If we let slop get by, we're "aiding and abetting". . . |
Why let 'slop' get by, an insistence to proper terminology does not always require a slap on the head.
| Quote: |
| From what i can tell, most of the new people would prefer to "say it right", but if their teachers, managers, and the seniors they interact with are not going to insist, they may go a long time not even knowing that their choice of words makes them look ignorant, lazy, and sloppy. |
Saying it right is a good thing, but understanding (in this forum environment) what is being said and asked is on us 'senior people' to get past the minor affronts of the misuse of terminology and provide pointers, solutions and advice for solutions and correct terminology.
| Quote: |
| Quite often "wrong wording" causes confusion and makes a topic get ugly because people are not on the same page. Using improper terms mis also quite unprofessional. It has also cost people a job they were being considered for. |
Common sense is the biggest tool we 'senior people' can bring to this forum. Any 'wrong wording' I've seen has only caused confusion among the less experienced people, and that is where Moderators come into play. The Moderators duty is to correct the 'wrong wording' in a way that does not bit**-slap the poster. I have mentored a number of (what turned out to be) good programmer/analysts and ridicule and scorn was not one of my teaching tools.
| Quote: |
| When i mess up, i'm thankful when someone tells me. . . Many/most new people are also. |
Dropped into a three man VSE shop as a COBOL/Assembler junior programmer and handed the systems responsibility the next month when he quit, my PSR taught me more through guidance and example than through scorn and I still maintain those standards. |
|
| Back to top |
|
|
William Thompson
Global Moderator
Joined: 18 Nov 2006
Posts: 3156
Location: Tucson AZ
|
|
|
|
| Robert Sample wrote: |
| I was working one place where someone who interviewed claimed he could code in "COBALT". Whether or not that was true, the boss decided if the person was unable to properly identify the language being used, he could not hire the person. So I have personal experience that improper terminology keeps people out of jobs. |
Some 30 or 40 years ago, I had a neighbor, a teenage high school senior, tell me that he thought the 'via' (Telstar) satellite was really neat, he really thought its name was 'via' because that is what he saw in a corner of his TV screen. I wondered then what is the future going to be like? |
|
| Back to top |
|
|
dbzTHEdinosauer
Global Moderator
Joined: 20 Oct 2006
Posts: 6966
Location: porcelain throne
|
|
|
|
well one benefit of knowing correct terminology is that you
can find things when you search thru manuals. |
|
| Back to top |
|
|
Earl Haigh
Active User
Joined: 25 Jul 2006
Posts: 475
|
|
|
|
| Quote: |
IIRC, some of those 'senior people' had their heads up their **es, I learned in that environment for a while, and it did promote a feeling of failure.
|
I've worked with a lot of 'senior people' , but I must admit, I have never seen any of them with their head up their **es.
That must have been quite a sight and could definitely lead to a sense of failure if one was to attempt emulation. |
|
| Back to top |
|
|
|
|
