Interested to Buy Any Domain ? << Click Here >> for more details...
Answer Posted / stu
Data is read and written to files in blocks. The reading
and writing of blocks is I/O. I/O is overhead (increases
runtime and cost) and you want to keep it to a minimum.
When reading a file, a whole block of records will be read,
but the records in the block will be fed to the program one
at a time until the block is exhausted and the next block is
read. The operating system does this behind the
scenes--it's transparent to the programmer.
When writing to a file, the records are not actually written
to the file until a block is full (or the program ends--at
which time the last block is written if it contains at least
1 record.) Again, this is transparent to the programmer.
Obviously, to keep I/O low, you want the cram as many
records into a block as possible.
Let's say your FB file's record length (LRECL) is 252 bytes
and your facility allows block sizes up to 27000 bytes.
Your blocksize would be ((integer (27000 / 250)) * 252) or
26964 (107 records).
| Is This Answer Correct ? | 0 Yes | 0 No |
Post New Answer View All Answers
I have many files which i am receiving from client everyday. I have one step for every file to check for empty or not. Here client gives 30 files i need to check for every file for empty or not and i need to perform 30 steps. Can I do it in single step. Dynamically i need to change the File name in my step.
What is jcl in mainframe, and how many types of jcl statements are there for a job?
I have a job called careerride and some steps in it as step1,step2. But I want to execute step2 before step1 ? How do I do that?
what is use of dsn parameter in dd statement?
which utility is used to run a cobol-db2 program?
Explain the purpose of dd * statement in jcl?
How jcl work to handle various input output file operations?
Is it possible to left uncode disp? If yes, how?
write a jcl to execute a job by 7:00 am on jan 20,1986?
I want to join to input files and write the matching fields on to an output file but i dont want the output file a sorted one
How can the submitting users racf authority be overridden in a job stream?
What statement can be used to send data to another mvs jes3 node?
How does the jcl specify the job to os?
how to do automated restart when a job abends?
A chkpt is the parameter coded for multi-volume qsam datasets in a dd statement. When a chkpt is coded as chkpt=eov, a checkpoint is written to the dataset specified in the sysckeov statement at the end of each volume of the input/output multi-volume dataset. State whether true or false?
