After you know how many records will fit on a track and the maximum number of records you expect to create, you can determine how many tracks you will need for your data.

   Number of tracks required = (Maximum number of blocks / Blocks per track) + 1

Example: Assume that a 200,000 record parts-of-speech dictionary is stored on an IBM 3380 Disk Storage as an indexed sequential data set. Each record in the dictionary has a 12-byte key that contains the word itself and an 8-byte data area that contains a parts-of-speech code and control information. Each block contains 50 records; LRECL=20 and BLKSIZE=1000. Using the following formula, you can calculate that each track can contain 26 blocks, or 1300 records, and a total of 155 tracks is required for the dictionary.

Blocks = 47968/(256+((12+267)/32)(32)+((1000+267)/32)(32))
       = 47968/1824 = 26
 
Records per track = (26 blocks)(50 records per block) = 1300
 
Prime data tracks required = (200000 records / 1300 records per track) + 1 = 155