In a Cobol-DB2 program, I am fetching rows from 4 tables
using cursor and then based on the a field present in that
table, It processes the information accordingly..for
example stat-c is one digit field..if stat-c is 'D' then
the a row is deleted from table and written those details
in to a file. If the stat-c is 'U' then a row is updated
(hardcoded what to update)in a table and written those
details in to a file. If the stat-c is 'I' then a row is
inserted in a table and written those details in to two
files. The issue is i have to include the intermediate
commits. When an abend occurs, due to commit statement db2
tables will be saved, But there will be lose of file
contents. When we resubmitting the job associated with this
program there will be insert ,update and delete anomolies
to avoid that what measures could be taken?.
The intermediate commit is nothing but issuing
commit after massive inserts, updates and deletes(sum of
500actions)
Answers were Sorted based on User's Feedback
Answer / just helpin
This is a JCL question, not a DB2 one.
You have to make two steps in your JCL. In the first step
preallocate the output file. In the second step run your
program and have DISP=(MOD,KEEP,KEEP) on the output file.
When you rerun after an abend restart the from the second
step.
| Is This Answer Correct ? | 0 Yes | 1 No |
What is the syntax of SELECT statement when embedded in a COBOL program?
Give the COBOL definition of a VARCHAR field.
While unloading huge amount of data from table. Suddenly job failed some error. Imagine 1M data unloading, In that 90% data unloaded only 10% left, So if want to unload the rest 10% what needs to be done? Whether do i need to start from top or anything ?
What is a synonym ?
how to resolve -811 error. give clear explination
we can code COPY DCLGEN or INCLUDE DCLGEN, At which stage of the precompilation , dclgen get expanded if we write 1) copy 2) include one question about dclgen.. Is it mandatory to use declare table in dclgen.. I think no...but it will be used by the precompiler to validate the table name,column name etc.., can one clear my doubt ..is it necessary to include declare table or not?
is it compulsory commitment control in journal?
I have a main program (A) where we delete some rows in table in a cursor, and we commit it in sub program(B). What will happen - will we get an error or not?
What are some characteristics of columns that benefit from indexes?
What do you mean by cursor?
in my table i will update only 100 record among 1000 records i need to display recently updated record'.....can u tell me where those records are stored how can i display those records...?if there is any sql query pls let me know....?
Mention the downsides of page level lock.