Hierarchical DBMS

Details

The typical representative (the most known and widespread) is Information Management System (IMS) of IBM company. The first version appeared in 1968. Still there are bases which are supported by this DBMS. Hierarchical models have a tree structure where to each node there corresponds one segment representing the named linear train of data fields. One corresponds to each segment input and several output segments. Each structure member lies on the only hierarchical way beginning from root. Hierarchical databases are most suitable for modeling of the structures which inherently are hierarchical. It is possible to give the military divisions or difficult mechanisms consisting of simpler nodes which in turn too can be subjected to decomposition as examples. Nevertheless there is a significant amount of the organizations which are not coming down to simple hierarchy. In this model the inquiry sent down hierarchy is simple, however the inquiry sent hierarchy is more difficult. Effective use of computer memory and quite good indicators of runtime of the main transactions over data belong to advantages of a hierarchical data model. The hierarchical database is the file system consisting of a root directory in which there is a tree structure of subdirectories and files.

Hierarchical DB

Hierarchical databases — the earliest model of representation of a difficult data structure. Information in hierarchical base is organized by the principle of a tree structure, in the form of the relations "ancestor descendant". Each record can have no more than one parent record and several subordinates. Communications of records are implemented in the form of physical pointers from one record on another. The main lack of a hierarchical structure of the database  — impossibility to implement the relations "much to - much" and also situations when record has several ancestors.

Graphically such structure can be represented in the form of the tree consisting of objects of different levels. The top level occupies one object, the second — objects of the second level and so on.

Between objects there are communications, each object can include several objects of lower level. Such objects are in the ancestor's relation (an object closer to a root) to the descendant (an object of lower level), at the same time it is possible that an object ancestor had no descendants or had them a little whereas at a descendant object surely only one ancestor. The objects having the general ancestor are called twins.

Shortcomings

It is necessary to carry to the main shortcomings of hierarchical models: inefficiency, slow access to data segments of the bottom levels of hierarchy, accurate orientation to certain types of requests, etc. Also a lack of hierarchical model is its bulkiness for information processing with rather complex logical relationships and also complexity of understanding for the normal user. Hierarchical DBMS quickly passed the peak of popularity which was caused by their early emergence in the market. Then their shortcomings made them noncompetitive, and now the hierarchical model is of exclusively historical interest.

Links