JBOB

From HandWiki

JBOB, an acronym for Just a Bunch Of Bytes, is a term is used to describe unstructured data that does not have a fixed format. This is a variation on the term JBOD (Just a Bunch Of Disks) that is used to describe standard hard drives that are used in a storage array.

Many computer files have a defined structure such as fixed length records with the data divided into records that are the same length. Structured data might have records of different lengths but each record is prefixed with a RDW (Record Descriptor Word) that indicates the length of that data as well as other attributes. JBOB data has no structure. Records are defined by the presence of characters in the data. For example, a report might have hundreds of records (or lines) but the length of each record is defined by the presence of a carriage return (and/or line feed). Mainframe computers have traditionally dealt with structured data but unstructured (JBOB) data is much more common in PC environments. The critical difference is that it is difficult, if not impossible, to advance to say, the 100th record without examining every character of the 99 records that precede it. With fixed length records, it is possible to calculate the exact position of a particular record. Even with variable length records, the length of each record is given so navigation is easier.

Since records are determined by the content of the data, metadata is required, such as what is/are the record termination character(s), and is usually stored externally to the actual data or file. The processing of JBOB data is usually more difficult and may require special knowledge by the computer program. Metadata might also be required for structured data, such as the fixed record length or the largest variable length record, but there usually exists standard utility software to read/write structured data since the format is a known structure.