Loading...
Select Version
The user program processes the file block-by-block (e.g. with the access method UPAM). Blocking and deblocking of records is thus not necessary. The user can speed up processing by using:
exchange buffer operation or overlapping input/output
While the program is processing the contents of one buffer, DMS executes the next I/O operation. If, for example, the program is reading buffer A, DMS can transfer the next data block to buffer B; when the program has finished processing buffer A, it can simply switch to buffer B, and so on.
chained input/output
DMS transfers up to 16 PAM pages or blocks between the peripheral device and the main memory at one time; time is saved by the reduced number of input/output operations needed for this.