Scheduling random I/O for data storage tape

In the operation of a serpentine data storage tape, the improvement comprising the following steps:

• a) obtaining a list of data blocks to read from the tape;
• and
• b) using data which indicates respective times for reaching each block from the beginning of the tape, scheduling the data blocks in an order which requires less time than reading the data blocks in the order listed.

In the operation of a serpentine data storage tape, the improvement comprising:

• a) maintaining several algorithms for scheduling read operations from the tape;
• and
• b) when a list of data blocks is to be read, selecting an algorithm, based on the number of data blocks in the list, and executing it.

A data storage system, comprising:

• a) a serpentine tape system;
• and
• b) means for predicting total times to execute lists containing less than 512 read requests, within two percent accuracy, for over 90 percent of the lists.

A method of reading data blocks, identified in a list, from a serpentine tape in which the data blocks are organized into sections, comprising the following steps:

• a) identifying data blocks contained in common sections;
• and
• b) presenting a list of said common sections to an algorithm for scheduling an order in which to read the common sections.

In a serpentine tape drive system wherein (A) data is stored in segments and (B) executing a list of read requests in order of listing results in an average data-read time T, the improvement comprising:

• a) means for re-ordering the list into an order which reduces T by at least 66 percent.